Component Description Format User

Guide
Product Version 5.1.41
December 2005
 1990-2005 Cadence Design Systems, Inc. All rights reserved.
Printed in the United States of America.
Cadence Design Systems, Inc., 555 River Oaks Parkway, San Jose, CA 95134, USA
Trademarks: Trademarks and service marks of Cadence Design Systems, Inc. (Cadence) contained in
this document are attributed to Cadence with the appropriate symbol. For queries regarding Cadence’s
trademarks, contact the corporate legal department at the address shown above or call 800.862.4522.
Open SystemC, Open SystemC Initiative, OSCI, SystemC, and SystemC Initiative are trademarks or
registered trademarks of Open SystemC Initiative, Inc. in the United States and other countries and are
used with permission.
All other trademarks are the property of their respective holders.
Restricted Print Permission: This publication is protected by copyright and any unauthorized use of this
publication may violate copyright, trademark, and other laws. Except as specified in this permission
statement, this publication may not be copied, reproduced, modified, published, uploaded, posted,
transmitted, or distributed in any way, without prior written permission from Cadence. This statement grants
you permission to print one (1) hard copy of this publication subject to the following conditions:
1. The publication may be used solely for personal, informational, and noncommercial purposes;
2. The publication may not be modified in any way;
3. Any copy of the publication or portion thereof must include all original copyright, trademark, and other
proprietary notices and this permission statement; and
4. Cadence reserves the right to revoke this authorization at any time, and any such use shall be
discontinued immediately upon written notice from Cadence.
Disclaimer: Information in this publication is subject to change without notice and does not represent a
commitment on the part of Cadence. The information contained herein is the proprietary and confidential
information of Cadence or its licensors, and is supplied subject to, and may be used only by Cadence’s
customer in accordance with, a written agreement between Cadence and its customer. Except as may be
explicitly set forth in such agreement, Cadence does not make, and expressly disclaims, any
representations or warranties as to the completeness, accuracy or usefulness of the information contained
in this document. Cadence does not warrant that use of such information will not infringe any third party
rights, nor does Cadence assume any liability for damages or costs of any kind that may result from use of
such information.
Restricted Rights: Use, duplication, or disclosure by the Government is subject to restrictions as set forth
in FAR52.227-14 and DFAR252.227-7013 et seq. or its successor.
 Component Description Format User Guide

Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Related Documents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Typographic and Syntax Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Typographic Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Syntax Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Data Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

1
What Is CDF? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
A Typical Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Levels of CDF Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
CDF Worksheet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

2
CDF Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
CIW CDF Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Edit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
CDF Editor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Copy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Delete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Scale Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
cdfEditScaleFactors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
SKILL CDF Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Saving User-Level CDF Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Using SKILL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Complex Pole Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Using the CDF Editor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
NFET Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Using the CDF Editor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

December 2005 3 Product Version 5.1.41

 Component Description Format User Guide

Using SKILL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

3
Defining Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Component Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Edit CDF Parameter Form . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Attributes and Parameter Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Editing Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
Using the CDF Editor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
Using SKILL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
Passing Parameters in a Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
pPar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
iPar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
atPar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
dotPar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
Inherited Parameters in Callbacks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Parameterized Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Complex Pole Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Using the Edit CDF Parameter Form . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
NFET Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
Adding New Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Setting Sample Callbacks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Completing Parameter Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

4
Modifying Simulation Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Edit Simulation Information Form . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Simulator Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Editing Simulator Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
Using the CDF Editor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
Using SKILL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Complex Pole Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
NFET Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90

December 2005 4 Product Version 5.1.41

 Component Description Format User Guide

5
Specifying Label Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Interpreted Labels Information Section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Library Label Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Cell Label Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Creating Labels Using the CDF Editor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
The Edit Component Display Options Form . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
Creating Labels in the Schematic Editor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Creating Labels with SKILL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
Complex Pole Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
NFET Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104

6
Other CDF Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
Other Information Section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Setting Up Fields and Prompts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
Adjusting Button Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
Adjusting Field Heights . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
Adjusting Field Widths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Adjusting Prompt Widths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
Initialization Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
Postprocessing Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113

7
Writing Callbacks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
Loading Callbacks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
NFET Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
mosLayout Callback . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
pcMOSw Callback . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
pcMOSl Callback . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122

December 2005 5 Product Version 5.1.41

 Component Description Format User Guide

8
CDF SKILL Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
CDF SKILL Function Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
Cell and Library Data IDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
Data Objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
Expressions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
Global Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
CDF SKILL Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
Create Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
cdfCreateBaseLibCDF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
cdfCreateUserLibCDF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
cdfCreateBaseCellCDF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
cdfCreateUserCellCDF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
cdfCreateParam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
Queries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
Saving Descriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
Dumping and Editing Descriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
Deleting Descriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
Copying, Finding, and Updating Data and Parameters . . . . . . . . . . . . . . . . . . . . . . 149
cdfCopyCDF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
cdfCopyParam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
Setting Scale Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
cdfGetUnitScaleFactor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
cdfSetUnitScaleFactor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
Other SKILL Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
cdfParseFloatString . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
cdfFormatFloatString . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
cdfSyncInstParamValue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
cdfUpdateInstSingleParam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
Invoking the CDF Editor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156

December 2005 6 Product Version 5.1.41

 Component Description Format User Guide

9
Verifying CDF Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
Testing the Component . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
Test Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
Virtuoso XL Layout Editor Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
LVS Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
Making Changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
Finishing the NFET Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
Varying the Layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
Changing Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164

A
Advice and Warnings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

B
Accessing Subcircuit Simulation Data . . . . . . . . . . . . . . . . . . . . . . . . 173
Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
Accessing Subcircuit Port Currents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
Accessing Subcircuit Model Parameter and
Operating Point Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
Accessing Schematic Primitive Model and
Operating Point Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
Editing a CDF Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
Writable Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
Sample CDF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
Subcircuit Simulation Data for Socket Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
Accessing Subcircuit Port Currents for Socket Simulation . . . . . . . . . . . . . . . . . . . . 182
Sample CDF for spectreS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182

December 2005 7 Product Version 5.1.41

 Component Description Format User Guide

C
NBSIM Transistor CDF SKILL Description . . . . . . . . . . . . . . . . . . . 185

Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191

Index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197

December 2005 8 Product Version 5.1.41

 Component Description Format User Guide
Preface

Preface

This Component Description Format User Guide assumes that you are a computer-aided
design librarian or a circuit designer and that you are familiar with designing and developing
electronic components with Cadence® design and simulation software.

Related Documents
Component Description Format (CDF) descriptions are often used with other Cadence
products during the design process. The following give you more information about the use
of CDF descriptions.
■ Virtuoso Parameterized Cell Reference describes using CDF parameters with
parameterized cells.
■ Composer: Design Entry User Guide describes the Edit Component Display Options
form and label specifications.
■ Analog Expression Language Reference Manual describes the analog expression
language syntax frequently used in CDF descriptions.

Typographic and Syntax Conventions

This book uses two kinds of conventions, one to help you use menu commands and the other
to help you use Cadence SKILL language functions.

Typographic Conventions
Boxes and arrows in a sequence like the ones below show you the order in which you select
a command from a menu.

Tools ➾ CDF ➾ Copy ...

December 2005 9 Product Version 5.1.41

 Component Description Format User Guide
Preface

Syntax Conventions
This list describes the syntax conventions used for the SKILL functions specified in this
manual.

literal (LITERAL) Nonitalic (UPPERCASE) words indicate keywords that you

must enter literally. These keywords represent command
(function, routine) or option names.
argument Words in italics indicate user-defined arguments for which you
(z_argument) must substitute a name or a value. (The characters before the
underscore (_) in the word indicate the data types that this
argument can take. Names are case sensitive. Do not type the
underscore (z_) before your arguments.)
[ ] Brackets denote optional arguments.

… Three dots (...) indicate that you can repeat the previous
argument. If you use them with brackets, you can specify zero
or more arguments. If they are used without brackets, you must
specify at least one argument, but you can specify more.
argument...: specify at least one, but more are possible
[argument]...: you can specify zero or more

,… A comma and three dots together indicate that if you specify

more than one argument, you must separate those arguments
by commas.
{ } Used with vertical bars and encloses a list of choices from
which you must choose one.
| Separates a choice of options.

If a command line or SKILL expression is too long to fit inside the paragraph margins of this
document, the remainder of the expression is put on the next line, indented. For example:
?callback "cdfgData->pName1->value = cdfgData->pName2->value")

When writing the code, you should always put a backslash (\) at the end of any line that
continues on to the next line.

December 2005 10 Product Version 5.1.41

 Component Description Format User Guide
Preface

Data Types
Data Types by Type

Prefix Internal Name Data Type

a array array
b ddUserType Boolean
C opfcontext OPF context
d dbobject Cadence database object (CDBA)
e envobj environment
f flonum floating-point number
F opffile OPF file ID
g general any data type
G gdmSpecIlUserType gdm spec
h hdbobject hierarchical database configuration object
l list linked list
m nmpIlUserType nmpIl user type
M cdsEvalObject —
n number integer or floating-point number
o userType user-defined type (other)
p port I/O port
q gdmspecListIlUserType gdm spec list
r defstruct defstruct
R rodObj relative object design (ROD) object
s symbol symbol
S stringSymbol symbol or character string
t string character string (text)
u function function object, either the name of a function (symbol) or
a lambda function body (list)
U funobj function object
v hdbpath —

December 2005 11 Product Version 5.1.41

 Component Description Format User Guide
Preface

Data Types by Type, continued

Prefix Internal Name Data Type

w wtype window type
x integer integer number
y binary binary function
& pointer pointer type

December 2005 12 Product Version 5.1.41

 Component Description Format User Guide
What Is CDF?

1
What Is CDF?

■ Overview on page 14
■ A Typical Application on page 14
■ Levels of CDF Data on page 17
■ CDF Worksheet on page 19

December 2005 13 Product Version 5.1.41

 Component Description Format User Guide
What Is CDF?

Overview
The Component Description Format (CDF) describes the parameters and the attributes of
parameters of individual components and libraries of components. The CDF lets you create
and describe your own components. A CDF description assigns parameters and parameter
attributes to libraries and cells for many purposes:
■ Assigning parameter names and values
■ Allocating units and default values
■ Checking that values lie within specified ranges
■ Dynamically changing how parameters are displayed depending on predefined
conditions
■ Executing a SKILL callback function whenever certain information is changed
Note: Callbacks are called only from the Add Component and Edit Properties forms. As
a consequence, many problems are associated with callbacks and their use is
discouraged.

The CDF gives you:

■ a single location for storing all component information so that applications can share and
reuse parts of the component description.
■ independence from applications and cellviews.
■ a graphical user interface (the Edit Component CDF form) for entering and editing
component information.
■ strict checking so that tools do not need to check component data before using it.
■ basic units and scale factors that are common in electrical design.
■ a modular format, so that you can edit and update parts of a description.
■ a command to copy the CDF description of one component into another component.

A Typical Application
The CDF encompasses all levels of components, including discrete and full-custom
components, and all cellviews of components. For example, the resistor cell res can take on
two forms: an ideal resistor or a thin film resistor. If you want an ideal resistor, you specify the
value of resistance desired. If you want a thin-film resistor, which models discontinuities as

December 2005 14 Product Version 5.1.41

 Component Description Format User Guide
What Is CDF?

well as skin and substrate effects, you specify the length and width of the resistor and the
resistivity of the material in which the resistor body is built.

You can choose how to specify the thin film resistivity by

■ Specifying the resistivity as a sheet resistance (ohms per square)
■ Specifying the bulk resistivity of the material
In this case, you need to also specify the height of the resistor body.

To determine what parameters the resistor needs, follow the decision tree in the following
figure:

ideal
resistance
resistor
type
width
thin film

length
sheet sheet
rho
resistivity
type
bulk
rho
bulk

height

This pattern of resistor parameters illustrates how you can construct a CDF description for
this resistor that lets you automatically switch between different choices.

To create a complete CDF for this resistor, you create eight parameters.

The first parameter is the radio button field, resistorType. It is always displayed and editable
and has only two values, ideal and thin_film.

The second parameter, resistance, uses a cdfgData variable expression to test the value of
resistorType. If resistorType = ideal, then resistance is displayed and editable. If
resistorType = thin_film, then resistance is not available. resistance has a floating-point
numerical value.

December 2005 15 Product Version 5.1.41

 Component Description Format User Guide
What Is CDF?

The third, fourth, and fifth parameters behave in the opposite manner from resistance.
Through another cdfgData expression for each parameter, width, length, and resistType
become active if resistorType = thin_film, and they become unavailable if resistorType =
ideal. The width and length parameters have floating-point numerical values. The
resistType parameter is a string with only two values, sheet and bulk.

The last three parameters, srho, brho, and height, have floating-point numerical values.
srho becomes active when resistType = sheet, while brho and height become active when
resistType = bulk.

Additionally, you can use CDF to display prompts that spell out the complete name of each
variable (bulk rho instead of brho), and you can also set the CDF of this resistor to accept
expressions (such as those in the Analog Expression Language [AEL] for the values of these
parameters.

When a designer accesses this resistor through the Add Component, Create Instance, or
Edit Properties command, the form prompts the designer for the necessary values. The
form changes based on the values entered.

The CDF description of the resistance parameters for this resistor is stored in the following
SKILL code.
libId = ddGetObj("base")
cellId = ddGetObj(libId "res")
cdfDataId = cdfCreateBaseCellCDF(cellId)
cdfParamId = cdfCreateParam(cdfDataId
?name "resistorType"
?type "radio"
?prompt "Type:"
?defValue "ideal"
?choices list("ideal" "thin film")
?callback "myModelTypeCB()")
cdfParamId = cdfCreateParam(cdfDataId
?name "resistance"
?type "float"
?prompt "Resistance:"
?defValue 100.0
?callback "myResCB()")
cdfParamId = cdfCreateParam(cdfDataId
?name "width"
?type "float"
?prompt "Width:"
?defValue 20e-6
?callback "myWidthCB()")
cdfParamId = cdfCreateParam(cdfDataId
?name "length"
?type "float"
?prompt "Length:"
?defValue 100e-6
?callback "myLengthCB()")

December 2005 16 Product Version 5.1.41

 Component Description Format User Guide
What Is CDF?

cdfParamId = cdfCreateParam(cdfDataId
?name "resistType"
?type "radio"
?prompt "Resistivity Type:"
?defValue "sheet"
?choices list("sheet" "bulk")
?use "cdfgData->modelType->value ==\"thin film\""
?callback "myResTypeCB()")
cdfParamId = cdfCreateParam(cdfDataId
?name "srho"
?type "float"
?prompt "Sheet Rho:"
?defValue 25.0
?use "cdfgData->modelType->value == \"thin
film\" && cdfgData->resType->value == \"sheet\""
?callback "mySheetRhoCB()")
cdfParamId = cdfCreateParam(cdfDataId
?name "brho"
?type "float"
?prompt "Bulk Rho:"
?defValue 25e-10
?use "cdfgData->modelType->value == \"thin
film\" && cdfgData->resType->value == \"bulk\""
?callback "myBulkRhoCB()")
cdfParamId = cdfCreateParam(cdfDataId
?name "height"
?type "float"
?prompt "Height:"
?defValue 25e-10
?use "cdfgData->modelType->value == \"thin
film\" && cdfgData->resType->value == dq\"bulk\""
?callback "myHeightCB()")

Levels of CDF Data

You can attach a CDF description to either a component or a library of components. When
you attach a description to a library, all components in the library inherit the description. When
you attach a description to a component, you actually attach the description to the cell.

Librarians usually create and maintain the CDF descriptions for components. Library users
see the effects of the descriptions as they use components from the library. However, you
might want to override parts of the default description. For example, you might want to specify
your own default values for CDF parameters. So that you can override the description,
Cadence defines the following description levels:
■ The base-level CDF is the description that is permanently attached to a cell or library.
When the system reads in a cell or library, any base-level CDF attached to it is also read.
The creator of a base-level CDF must have write permission on the object to which the
description is being attached.

December 2005 17 Product Version 5.1.41

 Component Description Format User Guide
What Is CDF?

■ The user-level CDF is a user-specific overlay description on top of the base-level CDF
for a cell or a library (although the base level does not have to exist). The system never
saves a user-level CDF, so you must save it to a file and reload this level on startup. You
do not need write permission to attach user-level CDF information to an object.
■ The effective-level CDF is an overlay of a user-level CDF description on top of a
corresponding base-level CDF description.
Note: Because user-level CDF is specific to a user and not a library or cell, simulation results
can be different for different users. The use of user-level CDF is therefore discouraged.

“Overlay” means “effectively replaces.” A user cell value of 10,000 ohms for the resistor
parameter r takes the place of any base- or library-level value for r. Each level of CDF is
optional. You can create cell-level CDFs without having library-level CDFs, and you can
create user-level CDFs without having base-level CDFs.

Both libraries and cells can have base- and user-level CDF descriptions. The effective-library
CDF is the overlay of the user-library CDF on top of the base-library CDF. The effective cell-
level CDF is the overlay of the user cell CDF on top of the base cell CDF, overlaid on top of
the effective library-level CDF. Finally, the effective instance CDF is the effective cell-level

December 2005 18 Product Version 5.1.41

 Component Description Format User Guide
What Is CDF?

CDF with instance-specific information added to the CDF description. The following figure
shows the relationship between CDF levels and the Cadence database.

Base Lib CDF

ddLibId User Lib CDF

CDF Effective Lib CDF CDF

Database IDs Description
Levels

Base Cell CDF

ddCellId User Cell CDF

Effective Cell CDF

dbInstId Effective Inst CDF

CDF Worksheet
Before you start to build a component, you need to decide how you intend to use it, in which
applications, and with what requirements. Before you create the cell and CDF, you need to
determine the following:
1. What applications will you use?
❑ Design Entry
❑ Simulation
❑ Layout
❑ Design Checking

December 2005 19 Product Version 5.1.41

 Component Description Format User Guide
What Is CDF?

❑ Place and Route

2. What views are required by the applications you will use?
❑ Design Entry — symbol
❑ Layout — layout, flat, or pcell
❑ Simulation — spectre, spectreS, etc.
❑ Design Checking — auLvs, lvs
3. What parameters are required by the applications?
❑ Layout — pcell parameters, view names
❑ Simulation — model parameters
4. What restrictions do the applications impose on the parameters?
❑ Simulation — parameter representation
❑ Layout — pcell parameters representation, correspondence between the pcell and
CDF default values
❑ Are parameter values discrete values or any value in a range?
5. What dependencies do the parameters have on each other?
6. How do you want the user interface (the Add Component and Edit Property forms) to
appear?
❑ Are there parameters that you want to hide from the user?
❑ Are there parameters that you want the user to see but not be able to edit?
❑ What parameters should appear at the top of the forms?

This user guide illustrates each task with examples based on two components. In the first
example, you examine the CDF of the functional block description of a complex pole. This
component uses a macro model and is set up for simulation under Cadence spectre and
SpectreS. This example illustrates parameter passing in hierarchical designs.

In the second example, you create the CDF for a BICMOS NFET from the CDF information
for another component. The NFET is specified as a parameterized cell (pcell) that can be
simulated with Analog LVS. This example illustrates some of the consequences of specifying
particular values for a component CDF description.

For example, the answers to the set of questions for the BICMOS NFET cell would be as
follows:

December 2005 20 Product Version 5.1.41

 Component Description Format User Guide
What Is CDF?

1. What applications will you use?

❑ Design Entry — Yes
❑ Simulation — Yes
❑ Layout — Yes
❑ Design Checking — Yes
❑ Place and Route — No
2. What views are required by the applications?
❑ Schematics — symbol with G, S, and D pins for the gate, source, and drain. You do
not need a bulk node connection in this case.
❑ Simulation — spectre and spectreS interfaces to the Spectre simulator.
❑ Layout — pcell for the usual case, but also a large power transistor for I/O in a flat
layout. Both layouts have G, S, and D connections.
❑ Design Checking — auLvs
3. What parameters are required by the applications?
❑ Schematics — none
❑ Virtuoso XL Layout Editor (Virtuoso XL) — width, length, bend, contact information,
and maskLayoutViewName (because there are two different layouts)
❑ Spectre — model parameters, including w and l and a bulk node
❑ spectreS — model parameters, including w and l and a bulk node
❑ LVS — w and l
4. What restrictions do the applications impose on the parameters?
❑ Simulation — simulator parameters must be strings.
❑ Layout — pcell parameters need to be floating-point numbers and the defaults
between the CDF and the pcell must match.
❑ The length parameter has a minimum and a maximum. The width parameter has a
minimum and a maximum. The length and width must be on grid. There are no other
specific discrete values.
5. Do the parameters have dependencies on each other?

December 2005 21 Product Version 5.1.41

 Component Description Format User Guide
What Is CDF?

❑ The number of gates is a function of width. If the width exceeds 120, the layout
changes to have two gates (a bend with connections).
❑ If a designer selects the alternate large layout, the width and length are updated to
reflect the value.
❑ Length and width of each two parameters: one a string for simulation and one a
floating-point number for the pcell. You should not have to enter both and they
should be synchronized.
❑ The model parameters are dependent on the physical parameters. For example, you
might calculate a value for the drain diffusion area and a value for the source
diffusion area based on the physical parameters.
6. How do you want the user interface to appear? Are there parameters that you want to be
hidden from designers? Are there parameters that you want the designers to see but not
edit? What parameters should appear at the top of the list?
❑ If a designer selects the alternate large size layout, the width and length cannot be
edited because there is only one large size device.
❑ The bends parameter is not visible to a designer because it is not under the
designer’s control.
❑ The pcell length and width are not visible to a designer. You want the designer to
specify the width and length only once so that the forms show only the string version.
❑ Because drain and source diffusion areas are calculated, they should not be
editable.
❑ Show width and length and the layout choice at the top of the forms because those
are the parameters that designers must frequently edit.

December 2005 22 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

2
CDF Commands

■ CIW CDF Commands on page 24

■ SKILL CDF Functions on page 43
■ Saving User-Level CDF Information on page 44
■ Complex Pole Example on page 45
■ NFET Example on page 46

December 2005 23 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

CIW CDF Commands

This section describes commands for managing Component Description Format (CDF)
information. These CDF commands are on the Design Manager menu of the Command
Interpreter Window (CIW). The following menu is just one of many different Tools menus you
might see, depending on which Cadence® software you own. However, all CIW Tools menus
have the same CDF commands.

icms - Log: /usr/design/CDS.log

Edit lets you modify CDF information for a cell or library.

Copy lets you copy CDF information from one cell or library to another.

Delete lets you remove CDF information from a cell or library.

Scale Factors lets you set scaling factors for displaying CDF parameters.

December 2005 24 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

Edit

Tools ➾ CDF ➾ Edit ...

Creates or edits CDF data for a library or cell. With this command, you can
■ Add component parameters
■ Delete component parameters
■ Modify attributes of component parameters
■ Rearrange the display order of component parameters
■ Modify simulation information for all available simulators
■ Modify interpreted label information for schematic annotation
■ Modify special CDF properties that define CDF-generated forms, such as the Add
Component form

You can select CDF data for a cell or library on the Edit Component CDF form. Once you
make a CDF selection, this form modifies itself and expands to include any existing CDF
information. After you enter or modify information, you can redisplay the form by pressing the
Tab key or by moving the cursor out of the form.

December 2005 25 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

Edit Component CDF Form (Initial)

CDF Selection
■ Cell lets you edit CDF data for a cell. (The default value.)
■ Library lets you edit CDF data for a library.
■ CDF Type lets you choose a CDF level of Base, User, or Effective.
❑ Base mode writes information to the library or cell when you click Apply, so you must
have write privileges to the library and cell.
❑ User mode writes data to the user-level CDF when you click Apply. The editor does
not store data permanently in the library or cell, so you do not need write privileges.
Note: Because user-level CDF is specific to a user and not a library or cell, simulation
results can be different for different users. The use of user-level CDF is therefore
discouraged.
❑ Effective mode, in contrast to user mode, writes information to the user-level CDF
only when the new information differs from base-level CDF information. The editor
does not store data permanently in the library or cell, so you do not need write
privileges.
■ Library Name is the name of the library whose CDF data you want to edit.
■ Cell Name is the name of the cell whose CDF data you want to edit.

December 2005 26 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

■ Browse displays the Component Browser.

Note: If you press Return or click OK, the values you have entered take effect, but the Edit
Component CDF form also closes. Move the cursor or press the Tab key when you finish an
entry in a field and want to move to another field. Click Apply if you want the information you
entered to take effect and you want the form to stay open.

December 2005 27 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

CDF Editor
You can use the CDF editor to create, view, or edit a CDF description. The Edit Component
CDF form gives you access to all portions of cell and library CDF descriptions. It consists of
a header for identifying the library or cell CDF description and four subsections for details
about parameters, simulation information, interpreted labels, and other information.

Edit Component CDF Form

Header

Component Parameters

Simulation Information

Interpreted Labels Information

Other Information

The following sections in this chapter describe the fields and buttons in the header.
Subsequent chapters describe each of the four subsections.

December 2005 28 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

Edit Component CDF Form (Redisplayed)

After you enter a CDF cell or library description name, the Edit Component CDF form
expands to look like the following figure.

Edit Component CDF

Most of the form is cut off by the size of the window. Stretch the CDF Editor window to expose
the entire form, or scroll down to examine each section of the form. The more parameters a
component has, the longer the form. Often, you cannot display the entire form at once, even
on a large monitor. The following chapter sections scroll down the form and show each form
section by itself.

Form Header

File Name is the name of a file that you can create where you can store all the current editor
field values. You can store the CDF description of a cell, including User or Effective levels,
for future use by the CDF editor. You can type in the name of the file or click File Name Select
to choose a name.

December 2005 29 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

Load loads the CDF information in the file in the File Name field. The Edit Component CDF
form redisplays to show the new information. The editor does not save CDF data to the cell
or library until you click OK or Apply.

Save saves the contents of the CDF editor to the file name in the File Name field.

File Name Select displays a list of file names if you enter a directory name in the File Name
field. You can select a filename from the list by clicking on the name, then clicking OK. The
name appears in the File Name field.

Choose File or Directory

Change Directory redisplays the filename list to show the contents of the directory you
chose with File Name Select.

December 2005 30 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

Component Parameters

You use the Component Parameters section of the Edit Component CDF form to modify
component parameters. The editor displays parameter names and the default attribute values
for any existing component parameters.

Edit Component CDF

You can change the default value of parameters directly on the form by typing a value into the
appropriate parameter field. You need to press the Tab key at the end of a value. You can
also add, delete, and rearrange parameters and edit their attributes.

Name lets you choose Name or Prompt to view either the name or the prompt of the
parameter for the fields in this section. The form redisplays to show the value you selected.

Add displays the Add CDF Parameter form for adding a new parameter to the component.

Edit displays Edit CDF Parameter form for editing parameter attributes.

Move displays the Move CDF Parameter form, which lets you move a parameter relative to
other parameters in the Component Parameters section of the Edit Component CDF form.

Select To Delete lets you delete the parameter from the CDF. Click on the button to the right
of the parameter you want to delete. There is no undo.

Chapter 3 describes these three forms in detail.

December 2005 31 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

Simulation Information

The Simulation Information section of the form lets you edit simulation information for
simulators registered in the software.

Edit Component CDF

The value at the right of each simulator name is a list of the names of the terminals
(connection points) of the component.

Edit displays the Edit Simulation Information form.

Chapter 4 describes how to modify the options for simulation information.

December 2005 32 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

Interpreted Labels Information

The Interpreted Labels Information section of the Edit Component CDF form lets you edit
interpreted label information in the CDF description. Chapter 5 describes the options for
interpreted labels.

Edit Component CDF

December 2005 33 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

Other Information

The Other Information section of the Edit Component CDF form lets you edit CDF property
fields that are not in any of the previous categories. Chapter 6 describes the options for other
information.

December 2005 34 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

Copy

Tools ➾ CDF ➾ Copy ...

Copies CDF data from one library or cell to a new library or cell.

Copy Component CDF Form

CDF Source Information

CDF Type is the type of CDF to copy. You can specify either a base-level or a user-level CDF
description as the source CDF type. You cannot specify the effective-level CDF description
as the source or the destination.

Library Name is the name of the library whose CDF data you want to copy.

Cell Name is the name of the cell whose CDF data you want to copy.

Browse displays the Library Browser.

December 2005 35 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

CDF Destination Information

CDF Type is the type of destination CDF. You can specify either a base-level or a user-level
CDF description as the destination CDF type. You cannot use effective-level CDF as a
destination CDF type.

Library Name is the name of a library or a library that contains a cell into which you want to
copy a CDF description.

Cell Name is the name of the cell into which you want to copy a CDF description.

Browse displays the Library Browser.

Caution
You cannot copy cell CDF data back to the same cell, and you cannot copy
a level of CDF data if it has not been created yet. In both cases, the
system issues an error message.

If you copy a cell (component) or library to another cell or library, the CDF data might not be
copied. The transfer of CDF data occurs only if the target cell or library never existed and is
created in the copy operation. If, however, the target cell or library already exists, or you copy
only a cellview (a view of the component), the CDF data of the target cell (or library) remains
unchanged. Use the Copy CDF command to avoid this problem.

December 2005 36 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

Delete

Tools ➾ CDF ➾ Delete ...

Removes CDF data from a library or cell.

Delete Component CDF Form

CDF Type to Delete lets you choose the type of CDF to delete.

Library Name is the name of either a library whose CDF description you want to delete or
a library that contains a cell whose CDF description you want to delete.

Cell Name is the name of the cell whose CDF description you want to delete.

Browse displays the Library Browser.

December 2005 37 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

Scale Factors

Tools ➾ CDF ➾ Scale Factors ...

Displays the Units Scaling Factors form, which you can use to set scaling factors for
displaying CDF parameters. These abbreviations are used on the form:
f femto (one quadrillionth)
p pico (one trillionth)
n nano (one billionth)
u micro (one millionth)
m milli (one thousandth)
c centi (one hundredth)
k kilo (one thousand)
M mega (one million)
G giga (one billion)

Rounding with Scaling Factors

When a scaling factor is set to a value other than the input value, the output string is rounded
based on the following standards:
■ If the input value is an integer, no rounding will occur.
Input: 1497Output: 1.497K

■ If the number of decimal digits of the output value is greater than12, the output value will
be rounded to maintain no more than12 decimal digits.
Input: 1234.12345678901234 Output: 1.234123456789K
Input: 123456.7890123456 Output: 123.456789012346K
Input: 123.4567890123456 Output: 123.456789012346

■ If the number of decimal digits of the input value is12 or less, the number of decimal digits
in the output will be maintained unless it exceeds12.
Input: 1497.5 Output: 1.4975K
Input: 1497.56789 Output: 1.49756789K
Input: 123.12345678987 Output: 1.231234567899K

■ Trailing decimal zeros will be truncated to the number of decimal zeros in the input.
Input: 120000.0 Output: 120.0K
Input: 11000000.00 Output: 11.00M

December 2005 38 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

Input: 11000000.0 Output: 11.0M

Input: 11 Output: 11000000.0u

Units Scaling Factors Form

Units Scaling Factors

The following two entries are common to every field on the Units Scaling Factors form:
auto automatically scales to the most efficient
representation
none uses value as entered, without modifying it

Resistance has a cyclic field for choosing how to show resistance.

t_unitName t_scaleFactor Description
resistance Ohms ohms
kOhms one thousand ohms
MOhms one million ohms

Conductance has a cyclic field for choosing how to show conductance.

t_unitName t_scaleFactor Description
conductance uMhos microsiemens
mMhos millisiemens
Mhos siemens

Capacitance has a cyclic field for choosing how to show capacitance.

t_unitName t_scaleFactor Description

December 2005 39 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

capacitance fF femtofarad
pF picofarad
nF nanofarad
uF microfarad
mF millifarad
F farad

Inductance has a cyclic field for choosing how to show inductance.

t_unitName t_scaleFactor Description
inductance pH picohenry
nH nanohenry
uH microhenry
mH millihenry
H henry

Length (Metric) has a cyclic field for choosing how to show length.
t_unitName t_scaleFactor Description
lengthMetric uM micrometer
mM millimeter
cM centimeter
M meter

Time has a cyclic field for choosing how to show time.

t_unitName t_scaleFactor Description
time pS picosecond
nS nanosecond
uS microsecond
mS millisecond
S second

December 2005 40 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

Frequency has a cyclic field for choosing how to show frequency.

t_unitName t_scaleFactor Description
frequency Hz hertz
kHz kilohertz
MHz megahertz
GHz gigahertz

Voltage has a cyclic field for choosing how to show voltage.

t_unitName t_scaleFactor Description
voltage uV microvolts
mV millivolts
V volts
kV kilovolts

Current has a cyclic field for choosing how to show current.

t_unitName t_scaleFactor Description
current pA picoamperes
nA nanoamperes
uA microamperes
mA milliamperes
A amperes

Power has a cyclic field for choosing how to show power.

t_unitName t_scaleFactor Description
power uW microwatts
mW milliwatts
W watts

December 2005 41 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

SKILL Function

You can use the following Cadence SKILL language commands to set scale factors.
cdfGetUnitScaleFactor( t_unitName) => t_scaleFactor
cdfSetUnitScaleFactor( t_unitName t_scaleFactor) => t / nil

For example, to set lengthMetric to mM, use the following command:

cdfSetUnitScaleFactor("lengthMetric" "mM")

You can use abbreviations listed earlier instead of specifying the complete scale factor. For
example, to change the scale factor for resistance to MOhms, enter:
cdfSetUnitScaleFactor("resistance" "M")

To display the current scale factor for power, enter:

cdfGetUnitScaleFactor("power")

The system returns the present value, such as auto.

You can also set the t_scaleFactor to auto or none using the
cdfGetUnitScaleFactor SKILL command. In SKILL expressions only, you can use the
following ISO1000 standard units in scale factors.

Symbol Name Factor

Y yotta 1024
Z zetta 1021
E exa 1018
P peta 1015
T tera 1012
h hecto 102
da deca 10
d deci 10-1
centi c 10-2
a atto 10-18
z zepto 10-21
y yocto 10-24

December 2005 42 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

To display the Units Scaling Factors form and edit the scale factors, use the following SKILL
function.

cdfEditScaleFactors
cdfEditScaleFactors()
=> t / nil

Displays the Units Scaling Factors form which can be used to set scaling factors for displaying
CDF parameters.

SKILL CDF Functions

You can use SKILL functions to create, view, or modify CDF information. Chapter 8 describes
the SKILL functions you can use to do the same data entry operations you perform with the
Edit Component CDF form.

To view the SKILL for a CDF description, use the cdfDump function.
cdfDump(
"libName"
filename
?cellName "cellName"
?level {’base | ’user}
?edit b_edit
)

A specific example is:

cdfDump("analogLib" "edtmp" ?cellName "nfet" ?level ’base ?edit t)

If you specify the libName only, cdfDump writes only the library CDF data to the file
specified in filename. The other arguments are optional. You must always specify a
libName and a filename.

The ?cellName argument dumps the CDF description from that cell.

With the ?level argument, you can choose base- or user-level CDF. The default is base.

The ?edit option lets you automatically load the dump file into a temporary editor window.
The default is nil.

You can specify a text editor by adding a line to your .cdsinit file. The following example
uses the vi editor.
editor = "xterm -g 80x35+30+30 -e vi"

December 2005 43 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

If you have write permission for a cell, you can reload the CDF description with the following
command in the Command Interpreter Window (CIW):
load("filename")

If you have the skillDev software, you can get a list of all the available CDF functions by typing
the following command in the CIW:
listFunctions("cdf")

If you want to change the CDF description of a component but do not have the appropriate
permissions, you can copy the cell to one of your own libraries. Copying a cell to a new cell
also copies its CDF description, which you can then edit. Alternatively, you can change the
user-level CDF description.

To create a CDF description for a new cell, you can first dump the CDF description from a
similar cell, change the cell or library name, then load the file as described in this section. If
the library and cell you specified exist, a new CDF description is created.

Saving User-Level CDF Information

User-level CDF information is not written to your system disk. When you exit the Cadence
software, all user-level CDF information that you have not specifically saved is lost. If you want
to save your user-level data, you must write it to a file.

If you are using the Edit Component CDF form in the Cadence Analog Design Environment,
you can save any CDF description that you are working on with the File Name field and its
associated buttons.

Use to save
user-level CDF
information

1. Use File Name Select and Change Directory to enter the name of the file where you
want to save the CDF data, and click Save.
2. To retrieve this CDF information, open the Edit Component CDF form again.

December 2005 44 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

3. Type the name of the library, cell, and CDF Type (User).
4. Type the filename in the File Name field, and click Load.

Using SKILL
If you are using SKILL, use adcDump to save the user-level CDF data and load to restore it.
1. Type the following SKILL command:
cdfDump("lib_name" "filename" ?cellName "cell_name"
?level ’user ?edit t)

An editor window appears, showing the file specified in filename.

2. Save the contents of this new file.
3. To restore the user-level CDF data, type
load("filename")

The load command reads the information in the file to determine which library, cell, and
level of CDF description to access. It then executes the SKILL functions contained in the file,
replacing the saved user-level data in the correct library and cell.

Complex Pole Example

To look at the CDF information for this component, you need the functional library, which is in
the following location:
your_install_dir/etc/cdslib/artist/functional

Make sure that your cds.lib file includes the definition of this library.

Using the CDF Editor

Follow these steps to edit the CDF of the complexPole1 cell in the functional block library:
1. From the CIW menu banner, choose Tools->CDF->Edit.
The Edit Component CDF form appears.
2. Click Browse to display the Component Browser.
3. Select the library functional and the cell complexPole1.

December 2005 45 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

The library and cell names are automatically entered in the form fields. After a moment,
the form expands and the CDF description for the complexPole1 component appears
in the form.

Form header

Component
Parameters
section

4. Using the scroll bar at the right, move down the form.
You should see a header and four subsections, each with information specific to the
complex pole cell.
Using the header, you can save the current CDF descriptions to a named file or load
another CDF description onto the named cell from a file. This is useful if you want to see
the results of changing certain CDF properties. You can store the alternatives in different
files and load them as required. You can also save the CDF data from one cell and then
load (transfer) it onto another cell.
At the end of the next chapter you examine the Component Parameters section in detail.

NFET Example
In the complex pole example, you examine an existing CDF description. In the NFET
example, you create a new cell and a new CDF description.

December 2005 46 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

Using the CDF Editor

When you copy a cell to another cell, all views and the CDF descriptions are copied if the
destination cell did not previously exist. Copy the nbsim component from the analog library
to create the basis for the new nfet component. You can then edit settings for parameters,
simulation, labels, and other information in the CDF description of the nfet component.
1. Create a library called bicmos.
Make sure that you include this library in your cds.lib file.
2. In the CIW, open the Library Manager by choosing Tools->Library Manager.
3. Select analogLib to display its components (cells).
4. Scroll down the cell list and select nbsim.
5. Choose Edit->Copy and copy the nbsim cell to the nfet cell in the bicmos library.

Library Manager

A copy of nbsim appears in the bicmos library under the cell name of nfet.
If you have already made a new NFET part by creating a symbol or a schematic, you still
need to copy over the CDF description.
6. In the CIW, choose Tools–>CDF–>Copy.
The Copy Component CDF form opens.

December 2005 47 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

7. Type the analogLib library and the nbsim cell in the CDF Source Information section.
8. Type the name of your destination library (bicmos) and the cell name (nfet) in the CDF
Destination Information section.
Press the Tab key at the end of each entry.

Copy Component CDF

The source cell is entered

automatically when
browsing

Enter the name and

library of the
destination cell.

9. Click Apply.

December 2005 48 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

You can look at the CDF description of nfet using the Edit Component CDF form, as you did
in the complex pole example in this chapter.

Edit Component CDF

In the next chapter, you will change some of the parameter attributes to make nfet a different
part from nbsim.

December 2005 49 Product Version 5.1.41

 Component Description Format User Guide
CDF Commands

Using SKILL
The following SKILL function copies the base-level CDF description of nbsim in analogLib
to nfet in bicmos.
cdfCopyCDF( dest_cell "baseCellData" srcCDF )

This is possible only after you have performed these assignments, in order:
source_cell = ddGetObj( "analogLib" "nbsim" )
srcCDF = cdfGetBaseCellCDF( source_cell )
dest_cell = ddGetObj( "bicmos" "nfet" )
if( destCDF = cdfGetBaseCellCDF(dest_cell) then
cdsDeleteCDF( destCDF )
)

If the destCDF already has a CDF description of the specified type, a warning message will
be displayed in the CIW indicating that the Base/User CDF already exists. In this case, the
CDF copy function will not be performed.

Refer to Appendix C, “NBSIM Transistor CDF SKILL Description,” for an example of the SKILL
CDF description of the nbsim cell.

December 2005 50 Product Version 5.1.41

 Component Description Format User Guide
Defining Parameters

3
Defining Parameters

■ Overview on page 52
■ Component Parameters on page 53
■ Editing Parameters on page 60
■ Passing Parameters in a Design on page 63
■ Parameterized Cells on page 65
■ Complex Pole Example on page 66
■ NFET Example on page 68

December 2005 51 Product Version 5.1.41

 Component Description Format User Guide
Defining Parameters

Overview
This chapter describes how to create cell parameters and set parameter attributes.

Each cell has a set of one or more parameters that describe the function of the cell. At any
one time, a parameter has one value and a set of attributes. Unlike the value, attributes
impose features and limitations on the parameters themselves, independent of their values.
For example, the Boolean attribute means that a parameter can have only one of two values.
A float attribute requires that the value of the parameter be a floating-point number.

Another attribute can determine whether the parameter value is displayed on a form or in the
schematic.

Cell
parameter
parameter
parameter
parameter value
attribute
attribute
attribute

The parameters section of a Component Description Format (CDF) describes what

parameters the component has, and the attributes of a component’s parameters. You can
describe the value of a parameter using long expressions based on the Analog Expression
Language (AEL). These expressions can include mathematical operations and functions that
call the value of other parameters. Attribute values are limited to very specific values.

When a cell is instantiated into a design, the parameters of that cell become properties.

The difference between a parameter and a property is the following:

■ A parameter is a characteristic of a component that has special meaning to the
component. Different instances of the same component do not always use the same
parameters and almost never use the same values for their parameters.
■ A property is how a component parameter is viewed by the Cadence® database when it
deals with each instance.

You can set up a hierarchy of CDF descriptions, one overriding the other, because you might
want to override specific attributes of some parameters. For example, if you define the xyz

December 2005 52 Product Version 5.1.41

 Component Description Format User Guide
Defining Parameters

parameter on a base-level CDF, you can define the same xyz parameter on the corresponding
user-level CDF. Because the user-level description overlays the base-level description, the
user-level attributes override the base-level attributes. However, certain parameter attributes,
such as the parameter name, cannot be overridden.
Note: Because user-level CDF is specific to a user and not a library or cell, simulation results
can be different for different users. The use of user-level CDF is therefore discouraged.

Component Parameters
The Component Parameters section of the Edit Component CDF form lists all of the
parameters defined for the selected CDF description. (Refer to Chapter 2 to learn how to
select libraries and cells.) Use the scroll bar on the right side of the form to view more
parameters. If a parameter has a default value, it is displayed in the form field next to the
parameter name.

Default
values

Parameter names

Name lets you choose whether to display the name of the parameter or the prompt for the
parameter on this form. Switching to Prompt redisplays the parameter information by prompt,
rather than by parameter name. This is useful if you become familiar with parameters from
their details in other forms that use the prompt rather than the name.

Select To Delete lets you delete the parameter from the CDF. Click on the button to the right
of the parameter you want to delete. The only way you can undo a deletion is to cancel all of
your edits.

Add displays the Add CDF Parameter form, which lets you add parameters to a CDF
description. You can place the new parameters in whatever order you like relative to the

December 2005 53 Product Version 5.1.41

 Component Description Format User Guide
Defining Parameters

existing parameters. The Add After Parameter field lets you specify where you want the new
parameter inserted. This cyclic field contains the names of all existing component
parameters.

Add CDF Parameter

The Add CDF Parameter form has fields for setting all attributes available to a CDF
parameter. When creating a new parameter, you can leave some fields blank. This does not
affect the overall CDF.

After you set the attributes you want, click on Apply to insert the new parameter as the first
parameter or after an existing parameter.

December 2005 54 Product Version 5.1.41

 Component Description Format User Guide
Defining Parameters

Edit displays the Edit CDF Parameter form, which lets you change the value and format of a
parameter.

Edit CDF Parameter

The form has fields for modifying all attributes available to a CDF parameter. You can select
a parameter to modify with Choose Parameter. When you select a parameter from the cyclic
field, the form redisplays to show the attributes for that parameter. When you change the
name of a parameter, you replace the first parameter with a second parameter.

After you change the attributes of one parameter, click Apply before you change the attributes
of another parameter.

Move lets you change the sequence of existing parameters. You can move a parameter
relative to other parameters in the Component Parameters section of the Edit Component
CDF form:

Move CDF Parameter

December 2005 55 Product Version 5.1.41

 Component Description Format User Guide
Defining Parameters

Commands that use CDF data display CDF parameters in the same order as the CDF editor.
To avoid unnecessary scrolling, place frequently used parameters at the top.
Note: The CDF description does not change until you click Apply or OK on the Edit
Component CDF form. Clicking Apply or OK on a subform updates the Edit Component CDF
form, not the CDF.

Edit CDF Parameter Form

Each parameter has attributes that distinguish it from other parameters. The Edit CDF
Parameter form is automatically modified depending on the parameter attributes you select.

The following figure shows the attribute list for the string parameter bn.

Edit CDF Parameter

paramType is the data type of the parameter. You need to specify paramType for every
parameter. Valid data types are string, int (integer), float (floating-point number), radio
(radio button), cyclic (cyclic field), boolean, button, and netSet.

parseAsNumber specifies whether this parameter can evaluate to a floating-point number.

Use this attribute only for string type parameters that contain numeric data.

December 2005 56 Product Version 5.1.41

 Component Description Format User Guide
Defining Parameters

If parseAsNumber is yes, the system converts the string to a floating-point number, converts
the floating-point number to the most efficient notation for execution, and reconverts the
floating-point number into a string.

If parseAsNumber is no (the default) or don’t use, the system does no conversion. Use this
setting if the parameter to be defined is a literal that contains numeric characters, such as the
name of a file or a model.

parseAsNumber must be used when parseAsCEL is yes.

units determines the unit suffix and scale factor to use when displaying the parameter value.
Use this attribute only with string type parameters where the response to parseAsNumber
is yes. (This specification has no effect if you set it for other parameter types.)

The units attribute must be one of the following: resistance, capacitance, inductance,
conductance, time, frequency, power, powerDB, lengthMetric, lengthEnglish, angle, voltage,
current, or temperature.

parseAsCEL specifies whether the parameter is processed as a CDF Expression Language

(CEL) expression. Use this attribute only for string type parameters.

If parseAsCEL is yes, the parameter is processed as a CEL expression. (Expressions such

as iPar(x) or pPar(x), indicating inheritance from an instance parameter or a parent
parameter, are processed as CEL expressions.)

If parseAsCEL is no (the default), the parameter is not processed as a CEL expression. It

is taken literally.

storeDefault specifies whether the parameter default value is saved as a property of the
instance. All tools based on the Cadence Analog Design software use the CDF description
to find default values if there is no property on an instance.

If set to no (the default) or don’t use, a property is not saved on the instance when the
parameter value is the default. Also, if the default value of the parameter changes, all
instances that use the default automatically get the new default value. (To see the change in
an open window you must choose Window – Redraw from the Cadence menu.)

If set to yes and the parameter value is set to the same value as the default, a property is
saved on the instance. If you are a user of the Open Simulation Software (OSS), and you used
to set this attribute to yes because you could only netlist using instance properties, you can
now set it to no because OSS users now have the option of using CDF. One disadvantage of
this attribute is that if the default value of a parameter changes, the instances that use the
default do not automatically change to the new default.

December 2005 57 Product Version 5.1.41

 Component Description Format User Guide
Defining Parameters

name is the parameter name. All parameters must have a name defined, and parameter
names cannot begin with a number. Sometimes this name is dictated by the application that
uses the parameter.

prompt is a string used as the prompt for the value of this parameter on the Add Component
form. If you do not specify a prompt, the parameter name is used. Because parameter names
are not always clear (r, ad, m, and so forth), you can use this field to create a more meaningful
prompt (such as resistance, diffusion area, multiplier, and so forth).

choices is a list of selections for a cyclic or radio data type. This attribute does not apply to
other types of parameters.

If each selection is a single word, you can separate the selections with spaces. If any
selection is a group of words, you must separate each selection with a comma (,). (When
using commas, do not leave extra spaces between selections because these spaces become
part of the selection value.)

A typical entry in the form field might be

choice 1,choice 2,choice 3

Notice that there is no blank between the number 1 and the comma, or between the comma
and the c in choice.

defValue is the default for the parameter. Most parameters must have a default defined. The
default value depends on the data type for this parameter. Parameter values can be long
mathematical expressions with variables and functions of other parameters.

use determines if the parameter is to be used. Often, you can enter Cadence SKILL
language expression that evaluates to t or nil in this field to determine if this parameter is
applicable. When the field evaluates to nil, the system never displays the parameter. When it
evaluates to non-nil (the default), then based on the value of display, the system will display
the parameter.

display determines if this parameter is displayed in forms that display CDF parameters, such
as the Edit Properties form or the Add Component form. You must enter t, nil, or a SKILL
expression that evaluates to t or nil in this field to determine if this parameter is to be
displayed. If the field evaluates to non-nil (the default), the parameter is displayed. If the field
evaluates to nil, the parameter is not displayed.

dontSave determines if the parameter value is to be saved on the instance. This attribute is
for programming use only. Typically, you should set this field to nil and not use it. If the field
evaluates to nil (the default), the parameter value is saved as a property on the instance. If
the field evaluates to non-nil, the parameter value is not saved as a property on the instance.

December 2005 58 Product Version 5.1.41

 Component Description Format User Guide
Defining Parameters

Caution
The dontSave attribute overrides the storeDefault setting. If storeDefault
is yes and dontSave is yes, not even the default property is saved.

editable determines if this parameter can be edited in forms that display CDF parameters,
such as the Edit Properties form or the Add Component form. Often, you can enter a SKILL
expression that evaluates to t or nil in this field to determine if the parameter is editable. If the
field evaluates to non-nil (the default), the parameter is editable. If the field evaluates to nil,
the parameter is not editable. (If not editable, the parameter is dimmed so that you can see
the value but you cannot edit it.)

callback specifies a SKILL routine to be executed whenever the value of the parameter
changes. The value of this optional field must be a string. Your entry can be the entire
function to be executed or a call to a function that is defined elsewhere. If you do not enter
anything in this field, the system assumes that there is no callback.
Note: The CDF parameter callback is primarily a GUI based callback. GUI based callbacks
occur when you modify the values in the parameter form fields. A GUI based callback is active
when the CDF parameters are displayed in the Add Instance form or the Edit Object
Properties form when you use the Create Instance or Edit Properties commands. For more
information on callbacks, refer to Chapter 7, “Overview on page 116,”.

Attributes and Parameter Types

Certain attributes are meaningful for some parameter types but not for others. For example,
the editable attribute is meaningful for string, integer, and float types, but not for radio.

The following table shows the attributes that are required (req), optional (opt), and unused for
each data type. The use attribute is for programming use only (prog). You should set the
dontSave attribute to the default value nil for all parameter types.

Attribute String Integer Float Radio Cyclic Boolean Button

paramType req req req req req req req
parseAsNumber opt unused unused unused unused unused unused
units opt unused unused unused unused unused unused
parseAsCEL opt unused unused opt opt unused unused
storeDefault opt opt opt opt opt opt unused
name req req req req req req req

December 2005 59 Product Version 5.1.41

 Component Description Format User Guide
Defining Parameters

Attribute String Integer Float Radio Cyclic Boolean Button

prompt req req req req req req req
choices unused unused unused req req unused unused
defValue req req req req req req unused
use prog prog prog prog prog prog prog
display opt opt opt opt opt opt opt
dontSave nil nil nil nil nil nil nil
editable opt opt opt unused unused unused unused
callback opt opt opt opt opt opt opt

Editing Parameters
All the parameter attributes descriptions in the previous section apply whether you are using
SKILL or the CDF editor. The following sections describe how you edit parameters using
either method.

Using the CDF Editor

Typically, you use the Component Parameters section of the Edit Component CDF form to
edit parameter attributes to test variations in your design. You can also create a new
parameter by editing an existing parameter and changing its name. This way, you can use
some or all of the parameter attributes from the original parameter.
1. Scroll to the Component Parameters section of the Edit Component CDF form.
2. Click Edit to display the Edit CDF Parameter form.
3. In the Choose Parameter cyclic field, select the parameter you want to edit.

December 2005 60 Product Version 5.1.41

 Component Description Format User Guide
Defining Parameters

The settings for the selected parameter are displayed as shown in the following figure.

Edit CDF Parameter

4. Change the parameter attributes.

5. Click Apply to enter the changes.

December 2005 61 Product Version 5.1.41

 Component Description Format User Guide
Defining Parameters

Using SKILL
You can use SKILL functions to create and modify parameters and their attributes. The
following SKILL functions create the parameters for the sample CDF description represented
here by the variable cdfDataId.

The first function defines the sample parameter maskLayoutViewName:

cdfParamId = cdfCreateParam(cdfDataId
?type "cyclic"
?name "maskLayoutViewName"
?prompt "Type of layout"
?choices ("layout" "layout.pwr")
?defValue "layout"
?callback "mosLayout( )")

The following functions define the sample parameter width with a callback from the
parameter w:
cdfParamId = cdfCreateParam(cdfDataId
?type "string"
?parseAsNumber "yes"
?units "lengthMetric"
?parseAsCEL "yes"
?storeDefault "no"
?name "w"
?prompt "width"
?defValue 7u
?display "artParameterInToolDisplay(’w)"
?editable "cdfgData->maskLayoutViewName->..."
?callback "pcMOSw( )")

The following SKILL functions define the parameters bend and bendt:
cdfParamId = cdfCreateParam(cdfDataId
?type boolean
?name "bend"
?prompt "PCell add bend?"
?defValue nil
?display nil)
cdfParamId = cdfCreateParam(cdfDataId
?type float
?name "bendt"
?prompt "PCell add bend?"
?defValue 0.0
?display nil)

The following SKILL function defines the sample parameter diffD:

cdfParamId = cdfCreateParam(cdfDataId
?type boolean
?name "diffD"
?prompt "Drain pin diffusion only?"
?defValue t
?display "(cdfgData->bend->value) && ...")

December 2005 62 Product Version 5.1.41

 Component Description Format User Guide
Defining Parameters

Passing Parameters in a Design

In algebraic expressions that define the value of parameters, the pPar, iPar, atPar, and
dotPar functions inherit the value of a parameter from a particular instance of the component.
■ pPar means parent parameter, NLP (netlist processor) syntax [+name]
■ iPar stands for instance parameter, NLP syntax [~name]
■ atPar is named after the NLP syntax [@name]
■ dotPar is named after the NLP syntax [.name]

The syntax of all four functions is

_Par("parameter")

You can use these inherited parameter value functions to pass parameters in a design. If you
change the value of one parameter, you also change every other parameter whose value is
defined by an expression that includes a Par function of the parameter that you changed.
Different Par functions can be used in the same expression.

You can define variables by expressions that use any combination of pPar, iPar, atPar, and
dotPar. Those functions can reference other variables that are defined by expressions that
include one of the Par functions. This chain can continue indefinitely. Circular references
generate errors.

pPar
pPar("x") lets you reference the value of parameter x from the parent instance of the
component.

An example of pPar used in a parameter definition is

rate = pPar("slewRate") + 100.0

When Analog Expression Language (AEL) evaluates this expression, it searches the parent
instance for the value of slewRate and substitutes it into the expression. If it cannot find
slewRate on the parent instance, AEL searches the parent cell’s effective CDF for a value
for slewRate.

Top-level schematic instances with parameters defined by expressions using pPar follow a
special procedure. The netlister first searches the property list of the top-level schematic for
a value for the referenced parameter. If it cannot find it, it searches the top-level cell’s effective
cell CDF for a default value for the parameter. This value can also be defined with an
expression, but that expression can only include the iPar and dotPar functions.

December 2005 63 Product Version 5.1.41

 Component Description Format User Guide
Defining Parameters

iPar
iPar("x") lets you reference the value of a parameter on the current instance.

Suppose the parameter i12 of a component is defined as a function of its input noise current,
n:
i12 = V1/r12 + iPar("n")

The AEL evaluates this expression by searching the current instance for the value of n, which
it substitutes into the expression. If the current instance does not have the parameter n, AEL
uses the cell’s effective CDF value as a default value for n. Changing the value of n does not
have an effect unless you change the value of n on this specific instance.

Caution
Use the iPar function when the value of a parameter depends on the value
of another parameter on the current instance but ensure that these
parameters are not dependant on each other. For example, if you use the
cap symbol and specify the length as iPar("w") and width as iPar(“l”) you
will get a segmentation fault error as this results in infinite recursion.

atPar
The use of atPar is discouraged. Use pPar instead. atPar searches the entire design
hierarchy, looking at the parent, grandparent, and great-grandparent of the instance, and so
on, until it finds the original value. Using atPar with hierarchical netlisting causes an error for
socket interfaces such as spectreS.

If atPar cannot find the parameter value anywhere in the design hierarchy, it then searches
the current instance master, which is typically the simulation primitive for the device in a
library. If that fails, atPar checks the global block for your simulator (nlpglobals).

dotPar
The use of dotPar is discouraged. It is not supported in several netlist situations.

When trying to evaluate a parameter on a primitive at the bottom level of a design hierarchy,
dotPar behaves like iPar. If a property on a component or functional block higher up in the
design hierarchy uses dotPar("x") in its definition, dotPar goes down to the bottom of the
design hierarchy, searching for x on the leaf instance or primitive currently being output to the
netlist.

December 2005 64 Product Version 5.1.41

 Component Description Format User Guide
Defining Parameters

At the top level of a design, dotPar searches the top-level cellview rather than the instance.

Inherited Parameters in Callbacks

Do not use callbacks based on parameters whose values are defined by expressions that
include inherited parameter functions. Changes to the inherited parameters do not trigger the
callbacks. Use callbacks only to establish dependencies between parameters whose values
are numeric constants.

For example, in a transistor with the length parameter L, L is defined by a callback. Normally,
changes in the value of the width parameter W trigger the L callback. The expression for this
is
cdfgData -> L -> value = cdfgData -> W -> value

In this example, W is defined by the following expression:

W = pPar("test_width")*10u

If you change the value of test_width, the callback that defines L is not triggered, and the
value of L becomes out of date.

Parameterized Cells
Although parameterized cells (pcells) provide great design flexibility, they also have unique
requirements for their CDF descriptions. These requirements include the following.
■ The default value for CDF parameters that are directly mapped to pcell stretch
parameters must be the same as the minimum value of the pcell parameter.
If the CDF default value is larger than the minimum pcell stretch parameter value, when
you place a pcell with a parameter value less than the CDF default value, the pcell
shrinks to the minimum pcell stretch value.
■ In layout, pcells do not understand string variables. All variables must be floating-point
numbers if you are using stretch parameters.
■ In a schematic, the system places pcell symbols according to string variables because
it must understand expressions (using the Analog Expression Language). Variables are
most likely to be strings to allow scaling units (nanometers, micrometers, and so on).

The second and third items might appear to conflict, but you can use one of the two methods
to simultaneously support the requirements of layout and schematics.

December 2005 65 Product Version 5.1.41

 Component Description Format User Guide
Defining Parameters

■ Create two CDF parameters. One is a string variable for the schematic instantiation of
the pcell symbol. The other is a floating -point number variable for the pcell layout view.
You must maintain both variables.
■ Create a callback for the floating point parameter based on the value of the string
parameter. The user enters the string value and both values are set.

Complex Pole Example

Using the same complex pole example from the previous chapter, you should have the Edit
Component CDF form open and have scrolled down to the Component Parameters section
of the form.

The Component Parameters section lists all component parameters. For complexPole1,
these are sigma, wn, and macro.

Using the Edit CDF Parameter Form

You can examine the different features of the Edit CDF Parameter form using the
complexPole1 as an example.

December 2005 66 Product Version 5.1.41

 Component Description Format User Guide
Defining Parameters

1. Click on Edit to open the Edit CDF Parameter form.

Edit CDF Parameter

The first field, Choose Parameter, is a cyclic field whose choices are the parameters
listed in the main form. The remaining fields give the current values and types of CDF
attributes attached to the parameter specified in the Choose Parameter cyclic field. The
parameter sigma has a paramType equal to string. The parseAsNumber and
parseAsCEL attributes are set to yes (true). The parameter name is set to sigma.
The prompt used in the Edit Object Properties and Add Component forms is Sigma.
When you are done, you see the parameter name Sigma, but the system keeps track of
the parameter by its real name, sigma. The default value of sigma is -5, so every instance
of the complexPole1 in a schematic has this value initially.
The first two parameters, sigma and wn, are component specific. The third parameter,
macro, identifies the file that contains the macro (subcircuit) definition. The default value
of macro is the f_cmplxP1 file. Because macro is a CDF parameter, you can change its
value to point to another file. This lets you create your own macros and reference them
from the CDF editor. Editing at this level changes all instances of complexPole1. To
change complexPole1 in a particular design, use the Edit Properties command in the
schematic editor.
2. Cancel the Edit CDF Parameter form to return to the Edit Component CDF form. (You
have not modified anything.)

If you want to make changes to the complex pole CDF description, copy the cell to a new
library that you have write access to.

December 2005 67 Product Version 5.1.41

 Component Description Format User Guide
Defining Parameters

The complex pole example continues at the end of the next chapter, where you examine the
options available for CDF simulation information.

NFET Example
This section shows you how to use the CDF editor to set up parameters and their attributes
for the NFET cell. You should have the Edit Component CDF form open for the NFET cell
created in Chapter 2.

Edit Component CDF

December 2005 68 Product Version 5.1.41

 Component Description Format User Guide
Defining Parameters

The sample NFET component requires a maskLayoutViewName parameter and five

parameters needed for the pcell layout. The following table describes the parameters.

Parameter Description
maskLayoutViewName Cyclic field that distinguishes between two layout views
width Floating-point number that defines the width of the gate
length Floating-point number that defines the length of the gate
bend Boolean to control conditional inclusion of shapes for bend
bendt Boolean to help stretch shapes correctly when bend occurs
diffD Boolean to control conditional inclusion of shapes for
diffusion drain contact versus metal drain contact

December 2005 69 Product Version 5.1.41

 Component Description Format User Guide
Defining Parameters

Adding New Parameters

Adding a new parameter is similar to editing a parameter. Use the following procedure to add
new parameters to the NFET CDF description.
1. Scroll to the Component Parameters section of the Edit Component CDF form.
2. Click Add to display the Add CDF Parameter form to create the maskLayoutViewName
parameter.
maskLayoutViewName is a cyclic parameter that lets you choose between two layouts.
The following figure describes decisions you must make about parameter position,
prompt use, parameter type selection, and callback actions for the
maskLayoutViewName parameter for the NFET component.

Add CDF Parameter

Add the parameter after
parameter m.

This prompt is more

meaningful than the
name required by
Virtuoso XL.

This cyclic field limits the

user to two choices.

Use this callback to

update width and length.

3. In the Add After Parameter field, add the new parameter to the parameter list after the
parameter m.
4. Specify the parameter type as cyclic and enter the other attributes shown in the sample
form, pressing the Tab key at the end of each field.
5. When you are done entering all the selections, click Apply so that your entries take
effect.

December 2005 70 Product Version 5.1.41

 Component Description Format User Guide
Defining Parameters

Setting Sample Callbacks

Because you copied the CDF description for the NFET component from the nbsim CDF
description, there is already a width parameter w in the CDF setup for the simulators. You can
set up a callback that updates the width parameter whenever the w parameter is updated.
Then designers don’t have to enter the width twice.
1. Click Add to display the Add CDF Parameter form again to create the width or pcell
width parameter.
2. Add the new width parameter after the geo parameter.
Because this parameter is set automatically by another parameter, put it near the end of
the parameter list. The new width parameter does not need to be displayed.
3. Set paramType to float and leave storeDefault as don’t use.

Add CDF Parameter

Place this parameter near
the bottom of the list, after
geo.

Enter a default value that

matches the w default.

Do not display this

parameter. The callback
sets the value.

4. Enter a default value that matches the default value for w.

5. Type nil for display.
Don’t display this parameter. The callback sets the value.
6. Click Apply to enter the new parameter.
7. On the Edit Component CDF form, click Edit.
8. Choose the w parameter.

December 2005 71 Product Version 5.1.41

 Component Description Format User Guide
Defining Parameters

9. Type in the callback name pcMOSw( ).

(Refer to Chapter 7 for a description of callback procedures.)

Edit CDF Parameter

Make the parameter not

editable when using
layout.pwr.

Add the callback name.

A change to this
parameter’s value starts
the callback.

10. In the editable field, make the parameter not editable by typing
cdfgData->maskLayoutViewName->value ==
cdfgData->maskLayoutViewName->defValue

The expression in the editable field of parameter w controls the designer’s ability to change
this parameter. This expression evaluates to nil if the attribute maskLayoutViewName is not
the same as its default value. In other words, if the designer selects the layout.pwr view
name, this parameter is not editable.

The expression in the display field of parameter w (copied as part of the w parameter from
the analogLib library) controls the display of parameters when the user uses the Edit –
Properties – Tool Filter command. The expression evaluates to t if this parameter is
associated with the tool selected.

You can expand the display attribute expression. For example, to turn off the display of w
whenever the maskLayoutViewName is layout.pwr (instead of making it uneditable), you
can use the following:

December 2005 72 Product Version 5.1.41

 Component Description Format User Guide
Defining Parameters

artParameterInToolDisplay(’w) &&
(cdfgData->maskLayoutViewName->value ==
cdfgData->maskLayoutViewName->defValue)

The function artParameterInToolDisplay takes a parameter name as its argument.

It evaluates that parameter in the current CDF description and decides to display it or not,
depending on whether the parameter is used by the simulator you are running.

December 2005 73 Product Version 5.1.41

 Component Description Format User Guide
Defining Parameters

Completing Parameter Definitions

Fill in the required information for the remaining parameters.
1. Repeat the previous procedure for the l and length parameters that you used for w and
width.
2. Using the Add CDF Parameter form again, enter the attributes for bend and bendt as
illustrated in the following figure.
The pcell parameters bend and bendt can also be controlled by the width parameter.
You can place both parameters at the bottom of the parameter list because they are
driven by the callback and, therefore, do not need to be displayed.

Add CDF Parameter Add CDF Parameter

3. Enter the attributes for the diffD parameter as shown below.

The last pcell parameter is the Boolean, diffD, which prompts the designer to specify
whether or not to use diffusion to make contact with the drain. The display of this
parameter depends on the value of bend and the type of layout selected. If you use

December 2005 74 Product Version 5.1.41

 Component Description Format User Guide
Defining Parameters

layout.pwr, there is no reason to ask about the drain connection. The same is true if you
are using a bend.

Place this parameter near Add CDF Parameter

the top of the parameter list
because the user needs to
select a value.

The display of this

parameter depends on the
value of bend and
maskLayoutViewName

The complete expression in the display field is as follows:

!(cdfgData->bend->value) &&
!(cdfgData->maskLayoutViewName->value=="layout.pwr")

You are ready to set the simulation options in the NFET CDF description, which are described
in the next chapter.

December 2005 75 Product Version 5.1.41

 Component Description Format User Guide
Defining Parameters

December 2005 76 Product Version 5.1.41

 Component Description Format User Guide
Modifying Simulation Information

4
Modifying Simulation Information

■ Overview on page 78
■ Edit Simulation Information Form on page 78
■ Editing Simulator Options on page 86
■ Complex Pole Example on page 88
■ NFET Example on page 90

December 2005 77 Product Version 5.1.41

 Component Description Format User Guide
Modifying Simulation Information

Overview
Before you can simulate a component, you must enter information required by each of the
simulators you intend to use. The Simulation Information section of the Edit Component CDF
form lists the simulators associated with a CDF description.

Edit Component CDF

Click here to enter

detailed information
for each simulator.

The value in the field to the right of the simulator name indicates that the component CDF has
simulation information for that simulator. The value is the default value of the terminal order
attribute, a list of the names of the terminals (connection points) of this component. You can
modify this value for each simulator on the form to accommodate the different requirements
of each simulator.

You must click on Edit to display the Edit Simulation Information form. Use this form to enter
detailed information for each simulator.

Edit Simulation Information Form

Unlike the parameter attributes that you enter with the Edit Component CDF form, the
information you enter into the Edit Simulation Information form is a set of directions,
parameters, and terminal names used by the simulator when it simulates the cell.

December 2005 78 Product Version 5.1.41

 Component Description Format User Guide
Modifying Simulation Information

Each simulator requires a unique set of information. Not every field described in this section
appears for each simulator. Only those fields that are appropriate for the simulator you select
are displayed. You can modify each of the fields in this form.

Choose Simulator is a cyclic field that lets you select the simulator to modify. When you
select a new value, the form redisplays to show attributes for that simulator. After you change
the attributes of one simulator, click on Apply before you change the attributes of another
simulator.

netlistProcedure is the name of the netlist procedure to use. Procedures that come with the
system are defined in tools/dfII/etc/context. You can also define your own procedures.

For the spectre interface, the netlister chooses a default netlist procedure if none is specified.
For socket interfaces, a netlist procedure must be specified.

The netlist procedures for spectre that are supplied by Cadence start with spectre. All
Cadence-supplied netlist procedures for socket interfaces start with ans, followed by an
abbreviation of the simulator name. The abbreviations used are: Cdl, Hpmns, Libra4, Lvs,
Spectre, SpectreS and Spice. Using SpectreS as an example, use
■ ansSpectreSCompPrim for elements (components without an associated model)
■ ansSpectreSDevPrim for devices with models
■ ansSpectreSSrcPrim for all voltage and current sources

December 2005 79 Product Version 5.1.41

 Component Description Format User Guide
Modifying Simulation Information

ansSpectreSCompPrim prints out the name of the component and its terminals.
ansSpectreSDevPrim prints out the name of the component, its terminals, and the model
name. ansSpectreSSrcPrim calls a generic routine to print out the sources.
Note: The netlist procedure name can be defined as either a symbol (without quotation
marks) or a string (with quotation marks). For example, ansSpectreSDevPrim (symbol) or
"ansSpectreSDevPrim" (string).

enumParameters are parameters of SPICE and Spectre primitives, where the values of the
parameters are enumerations. The parameter names must be names recognized by the
targeted simulator.

For details see Appendix B, "Updating Legacy SimInfo for Analog Primitives" in the Cadence
AMS Environment User Guide.

referenceParameters are parameters that have instance names as their values. The
parameter names must be names recognized by the targeted simulator.

For details see the Appendix B, "Updating Legacy SimInfo for Analog Primitives" in the
Cadence AMS Environment User Guide.

stringParameters are parameters that are treated as strings when they are written to the
netlist. The parameter names must be names recognized by the targeted simulator.

For details see the Appendix B, "Updating Legacy SimInfo for Analog Primitives" in the
Cadence AMS Environment User Guide.

arrayParameters are parameters that must be written to the netlist as arrays. The parameter
names must be names recognized by the targeted simulator.

For details see the Appendix B, "Updating Legacy SimInfo for Analog Primitives" in the
Cadence AMS Environment User Guide.

extraTerminals contains information for writing inherited connection terminals on instances.

This information is used when the simulator view of an instance contains more terminals than
are present on the symbol view.

For details see the Appendix B, "Updating Legacy SimInfo for Analog Primitives" in the
Cadence AMS Enviroment User Guide.

otherParameters for spectre are parameters not defined in instParameters. These

parameters are processed in a special way by the netlist procedure.

For socket interfaces, these are parameters not yet defined in modelArguments,
macroArguments, or instParameters. Typical values are macro, model, and bn. (The

December 2005 80 Product Version 5.1.41

 Component Description Format User Guide
Modifying Simulation Information

parameter names you list must reflect the property mapping you specify in the propMapping
field.) If you use a tool filter, parameters that are not used by the tool are not displayed. Any
field you specify in otherParameters, however, is displayed regardless of the tool filter (with
the artParameterInToolDisplay function).

For details see the Appendix B, "Updating Legacy SimInfo for Analog Primitives" in the
Cadence AMS Enviroment User Guide.

instParameters a list of simulator names of parameters that you want to include with this
device in the netlist. (The parameter names you list must reflect the property mapping you
specify in the propMapping field.)

For details see Appendix B, "Updating Legacy SimInfo for Analog Primitives" in the Cadence
AMS Environment User Guide.

Some simulators require that a property have a different name or different capitalization from
the parameter name in the parameter definition. For LVS (Layout Versus Schematic), this
depends on the parameter names used in the LVS rules in your technology file. For example,
the capacitor parameter might be c, and the LVS rules can refer to the property C. You can
set the propMapping field to map the value of c to C so that the parameters listed with each
instance are C.

modelArguments is a list of the cell CDF parameter names that are to be passed as
arguments to the cell model. This entry is not used by the spectre interface. (The parameter
names you list must reflect the property mapping you specify in the propMapping field.) The
order of the names defines the corresponding substitution pattern in the file; for example, &1
&2 &3. The first element of the list (&1) should not be considered part of the parameter list.
It is replaced by the model name when the list is read by the simulator. The
modelArguments list is also used with Spectre® SKILL-based models.

macroArguments is a list of the cell CDF parameter names that are passed as arguments
to the cell macro. This entry is not used by the spectre interface. The order of the names
defines the corresponding substitution pattern in the file; for example, &1 &2 &3. The first
element of the list (&1) should not be considered part of the parameter list. It is replaced by
the macro name when the list is read by the simulator.

Property mapping is not supported for macros.

When using a function like pPar(prop), prop has to be in the macroArguments field when
passing parameters down the hierarchy.

componentName is the type of component you are creating, such as BJT, JFET, and
MOSFET.

December 2005 81 Product Version 5.1.41

 Component Description Format User Guide
Modifying Simulation Information

For the spectre interface, if the "model" parameter ("Model Name") has a value, that value is
used as the name of the model in the netlist. Otherwise, the value of this entry is used as the
name of the model. If neither the model parameter nor the componentName entry has a
value, the name of the cell is used as the name of the model.

For details see Appendix B, "Updating Legacy SimInfo for Analog Primitives" in the Cadence
AMS Environment User Guide.

When you are using auLVS and auCDL, component names can be anything. For other
simulators, you must use lowercase component names similar to those used in analogLib.
This is critical for SPICE types of netlists.

termOrder is a list of terminals that define the net order for your component. The nets are
considered those nets connecting the terminals in the order given in this list.

For details see Appendix B, "Updating Legacy SimInfo for Analog Primitives" in the Cadence
AMS Environment User Guide.

You can also use programmable nodes to define the net order. The use of programmable
nodes is discouraged in releases 4.4.3 and later. Instead, use inherited connections. You can
use the conversion tools to convert your designs. Programmable nodes are not supported by
spectre.

To do this, you need to define a bulk node property (usually bn). The following entry is an
example of termOrder input:
PLUS MINUS progn(bn)

This indicates that the third net, progn, is defined by the bulk node property instead of by a
connection to a pin. The property bn must have a default value when using LVS.
Note: You can set a net name to a global net (such as gnd! or vdd!), to one of the instance
pins (such as PLUS or MINUS), or to a net name, as long as the net, pin, or name exists in
the cellview. You can assign a net name to bn. (This is called parameterizing the bulk node.)
You can also set a net name to any other net. If the netlisting is flat, the net name must be
fully qualified, identifying all levels of hierarchy in the design. If the netlisting is hierarchical,
the net name does not need be fully qualified if it is a local net name on the same page.

Caution
You must escape invalid characters while specifying termOrder using
symbols to make them valid symbols in SKILL. All symbols must be
preceded by the backslash (\) as used in the following code:
cdfId = cdfCreateBaseCellCDF( cellId )
.
.

December 2005 82 Product Version 5.1.41

 Component Description Format User Guide
Modifying Simulation Information

.
cdfId->simInfo->spectre = ’(nil
otherParameters (model)
instParameters (w l as ad ps pd nrd nrs ld ls m trise region)
termOrder ( a\<0\> a\<1\> )
termMapping (nil D \:d G \:g S \:s B \:b )
)

deviceTerminals lets you define the terminal names to be listed in the device definition line.
If you leave this field blank, the terminals listed use the same names as those in the
termOrder field. Used by auLVS only.

The entries in this field are usually different from those in the termOrder field when you have
a programmable node (the bulk node). For example, if you used progn(bn) in the termOrder
field for the three-terminal capacitor, you might see the following entry in deviceTerminals:
PLUS MINUS BULK

termMapping is used when the simulator has different names for the component terminals
than the cell schematic. It is set up as a disembodied property list. Therefore, the first entry
in the list is nil. This is followed by a set of name-value pairs, where the first name is the
terminal name on the symbol and the second name is its corresponding terminal name used
by the simulator for the device.

For example, in the nbsim cell in analogLib, the terminals names D, G, and S are specified
in uppercase. Although uppercase terminal names are appropriate for SPICE, the Spectre
simulator expects the terminal names d, g, and s in lowercase. Therefore, in the SKILL
version, to map between the two sets of terminal names for the Spectre simulator, you can
enter the following code in the simInfo field:
termMapping(nil D \:d G \:g S \:s B \:b)

Note: Make sure all non-alphanumeric characters are preceded by a backslash (\).

This is important when you use backannotation because the simulator saves the results
under the terminal name that it recognizes.

propMapping lets you enter your own CDF parameter names in place of the names
recognized by the simulator. You can define different name mappings for each simulator. A
single parameter can be identified a different way for each simulator.

For details see Appendix B, "Updating Legacy SimInfo for Analog Primitives" in the Cadence
AMS Environment User Guide.

You specify an entry in the propMapping field in the following format:

nil simParamName1 CDFparamName1
simParamName2 CDFparamName2 ...

This format represents a disembodied property list, so the first element is always nil.

December 2005 83 Product Version 5.1.41

 Component Description Format User Guide
Modifying Simulation Information

The LVS propMapping list is nil C c. The CDF parameter name is c, but under the LVS rules,
it must be C. In another simulator, the same CDF parameter value might be referred to as F.
In this case, the propMapping list for this simulator is nil F c. A third simulator might refer to
this value as cap, making its propMapping list nil cap c.

modelName is used only by the analog-microwave version of CDL.

permuteRule, used only in auLVS, lets you indicate the permutability of the pins, that is,
whether pins can be swapped or if they must be fixed. For example, the capacitor in the
propMapping definition shows that the PLUS and MINUS terminals are permutable.

namePrefix lets you enter a prefix for defining the instance in the netlist. This is not used by
the spectre interface. Typically, this is the first letter of your component. The netlister takes
this prefix (R, for example), and the name of your component (12, for example), and forms the
name (R12).

current lets you save the simulation and definition of the current plot set. This is not used by
the spectre interface. The only choices for this field are port or nil. If you do not specify a
choice, no currents are saved. You can choose to save or plot the current for the port of a
device or through a component. For SPICE-based simulators, you cannot save currents on
specific ports. Currents can be saved only on the component (for example, resistors). The
Spectre simulator lets you save currents on specific ports. For devices (for example, an npn
transistor), you probably want to save currents on a port.

Typically, only the positive port current of a component current is written to the simulation PSF
file, and the Cadence Analog Design software calculates the negative port by taking the
negative of the positive port. To use this feature to extract component currents, you must set
the termMapping field as follows:
termMapping =
(nil PLUS \:p MINUS "(FUNCTION minus(root(\"PLUS\")))")

modelParameters is used only by the microwave simulator HP MNS.

opParamExprList lets you define the terminal names to be listed in the device definition
line. If you leave this field blank, the terminals listed use the same names as those in the
termOrder field. Used by auLVS only.

December 2005 84 Product Version 5.1.41

 Component Description Format User Guide
Modifying Simulation Information

Simulator Options
The following table lists six simulators and the options that apply to them.

Cadence
Option Spectre SpectreS auLVS AMS
SPICE
netlistProcedure Yes Yes Yes Yes No
otherParameters Yes Yes No Yes Yes
instParameters Yes Yes Yes Yes Yes
enumParameters No No No No Yes
modelArguments No Yes No Yes No
macroArguments No Yes No Yes No
componentName Yes Yes Yes Yes Yes
termOrder Yes Yes Yes Yes Yes
deviceTerminals No No Yes No No
termMapping Yes Yes No Yes No
propMapping Yes Yes Yes Yes Yes
excludeParameters No No No No Yes
extraTerminals No No No No Yes
isPrimitive No No No No Yes
permuteRule No No Yes No No
namePrefix No Yes Yes Yes No
current No Yes No Yes No
arrayParameters No No No No Yes
stringParameters Yes No No No Yes
referenceParameters No No No No Yes
opParamExprList Yes Yes No Yes No

December 2005 85 Product Version 5.1.41

 Component Description Format User Guide
Modifying Simulation Information

Editing Simulator Options

Setting simulator options can be a lengthy process. You should copy CDF simulator
information from other components and libraries whenever possible. This section describes
editing simulator options with the CDF editor and with SKILL functions.

Using the CDF Editor

Use the following procedure to set simulator options with the CDF editor:
1. Click Edit in the Simulation Information section of the Edit Component CDF form.
The Edit Simulation Information form appears.
2. In the Choose Simulator cyclic field, choose a simulator.
3. Add or modify the simulator settings as needed.
4. Click Apply.
Set this option for the simulator before you change to another simulator.
5. Repeat steps 2 through 4 for the other simulators that you need.
6. Click OK.
You return to the Simulator Information section of the Edit Component CDF form.
7. Scroll to the Component Parameters section to check that all simulator parameters are
defined.
Note: When you edit the simulator information the CDF termOrder may not be automatically
updated. Instead, you may get the Want to Update dialog box to confirm overwriting the
existing CDF termOrder.
You can suppress the Want to Update dialog box by setting the .cdsenv variable
queryCDFtermOrder to nil. This will ensure that the CDF is automatically updated
without checking for confirmation. You can undo the automatic update by setting this variable
back to its default value, t.

December 2005 86 Product Version 5.1.41

 Component Description Format User Guide
Modifying Simulation Information

Using SKILL
You can use the SKILL variable cdfId and the right arrow (->) operator to specify simulator
options. Enter the options for each simulator as elements of a disembodied property list.

The following SKILL expression sets the options for the Spectre (spectre) simulator:
cdfId = cdfCreateBaseCellCDF( cellId )
.
.
.
cdfId->simInfo->spectre = ’(nil
otherParameters (model)
instParameters (w l as ad ps pd nrd nrs ld ls m trise region)
termOrder ( D G S B )
termMapping (nil D \:d G \:g S \:s B \:b )
)

The following SKILL expression sets the options for the Analog LVS (auLvs) simulator, given
a definition for cellId:
cdfId = cdfCreateBaseCellCDF( cellId )
.
.
.
cdfId->simInfo->auLvs = ’(nil
netlistProcedure ansLvsCompPrim
instParameters (m l v)
componentName n0bsim
termOrder (D G S progn(bn))
deviceTerminals (D G S B)
propMapping (nil L l W w)
permuteRule "(p D S)"
namePrefix "Q")

The following SKILL expression sets the options for the Spectre (spectreS) simulator:
cdfId = cdfCreateBaseCellCDF( cellId )
.
.
.
cdfId->simInfo->spectre = ’(nil
netlistProcedure ansSpectreDevPrim
otherParameters (bn model)
instParameters (w l as ad ps pd nrd nrs ld ls m trise)
componentName bsim1
namePrefix "S"
current port)

The following SKILL expression sets the options for the Cadence SPICE (cdsSpice)
simulator:
cdfId = cdfCreateBaseCellCDF( cellId )
.
.
.
cdfId->simInfo->cdsSpice = ’(nil
netlistProcedure ansSpiceDevPrim

December 2005 87 Product Version 5.1.41

 Component Description Format User Guide
Modifying Simulation Information

otherParameters (model bn)

instParameters (m w l ad as pd ps nrd
nrs ld ls off vds vgs vbs)
componentName bsim
termOrder (D G S progn(bn))
propmapping (nil vds Vds vgs Vgs vbs Vbs)
namePrefix "S"
current port)

Complex Pole Example

In this portion of the complex pole example, you use the Edit Simulation Information form to
look at the simulation information provided with the complexPole1 component.
1. Scroll down to the Simulation Information section on the Edit Component CDF form.
This section shows the simulator-specific data. You see the terminal names of the current
simulation representations for the complex pole component. Only the values for spectre,
spectreS, and cdsSpice are defined.

2. Click Edit to see the details for each simulator.

The Edit Simulation Information form appears.
3. In the Choose Simulator cyclic field, select spectre.

December 2005 88 Product Version 5.1.41

 Component Description Format User Guide
Modifying Simulation Information

The form displays all of the fields applicable to spectre. However, not all fields are
required by different netlist procedures.

For spectre, complexPole1 requires no netlist procedure.

The other important fields are termOrder and instParameters. The termOrder field is an
ordered list of the terminals as named on the symbol. The netlister uses this list. The order is
important, especially if the component is defined by a subcircuit: the order in termOrder maps
to the node numbers specified in the first line of the subcircuit.

Because complexPole1 uses a macro model (subcircuit definition) to define its operation,
the CDF parameter model must have a value in the Component Parameters section for
spectre. If you are using a schematic to describe the component, the model information is
unnecessary.

For the spectre interface, the instParameters field determines the names of the parameters
of all types of components, primitives, subcircuits, model cards, and Verilog®-A modules. The
subcircuit definition for f_cmplxP1 is
subckt f_cmplxP1 (1 2)
parameters sigma wn
E1 3 0 1 0 1E-6 (??)
R1 (4 5) resistor r=1
L1 (3 4) inductor l=(1/(-2*sigma))
C1 (5 0) capacitor c=(-2*(sigma))/((wn*wn)+(sigma*sigma))
ends

December 2005 89 Product Version 5.1.41

 Component Description Format User Guide
Modifying Simulation Information

The socket interfaces use the macro parameter instead of the model parameter. In order to
avoid duplicate information, the value of the model parameter is set to iPar("macro"), and the
model parameter is not displayed.

This model, and others used by the functional block library, are in the following file:
your_install_dir/tools/dfII/etc/cdslib/artist/spectre/functional.scs

For socket interfaces, the macroArguments field determines the names and order of
parameters that are passed down through the design hierarchy. Again, the order is
particularly important in the case of macro models. The first parameter in the list is assigned
to the variable &2 in the subcircuit description, the second to &3, and so forth. In this example,
complexPole1 is defined by a macro model and the parameters sigma and wn on the symbol
instance are passed to the model file f_cmplxP1 as &2 and &3.

The model file (f_cmplxP1) for the complex pole is

.SUBCKT &1 1 2
E1 3 0 1 0 1E-6
R1 4 5 1
L1 3 4 (1/(-2*(&2)))
C1 5 0 (-2*((&2)))/(((&3)*(&3))+((&2)*(&2)))
.ENDS &1

This model, and others used by the functional block library, are in the following directory:
your_install_dir/tools/dfII/etc/cdslib/artist/cdsSpice/functional

There are several other fields in the Edit Simulation Information form. The simInfo parameters
vary depending upon whether the component is a model or a macro for socket interfaces.
For the spectre interface, the fields are the same for models and macros. A model is a
component that references a model statement in the netlist. A macro is a component
described by a subcircuit. For example, complexPole1 is a macro, but a diode could be a
model. You determine whether a component is treated as a model or a macro by selecting
the appropriate netlist procedure: ansSpiceDevPrim or ansSpiceSubcktCall. Once you have
selected either model or macro, you can assign the appropriate fields in the Simulation
Information section. If you set the type to macro, the macroArguments field must contain the
arguments to assign to &2, &3, and so on in the subcircuit. Similarly, if the type is model, the
modelArguments field must contain the ordered list of arguments to pass to the model
statement.

NFET Example
Before proceeding with this example, open the Edit Component CDF form and set it to edit
the nfet cell in the bicmos library. Scroll down to the Simulation Information section.

December 2005 90 Product Version 5.1.41

 Component Description Format User Guide
Modifying Simulation Information

For this NFET example, the next step in CDF development is to select simulator options for
the new component. Because you copied the CDF description for the NFET component from
analogLib, there are six simulators with their options already defined, including auLvs,
cdsSpice, and spectre.

Edit Component CDF

For the Analog-Microwave LVS (auLvs) simulator, you need to define three parameters in the
parameter section of the CDF description, m, l, and w. For the NFET example, these
parameters were copied from the analogLib nbsim cell.

Edit Simulation Information

December 2005 91 Product Version 5.1.41

 Component Description Format User Guide
Modifying Simulation Information

You also need parameters for the spectre and spectreS interfaces but these are already
defined in the original CDF that was copied.

Edit Simulation Information

The Spectre simulator options are similar to the Cadence SPICE options, but the Spectre
simulator does not require property mapping.

Edit Simulation Information

December 2005 92 Product Version 5.1.41

 Component Description Format User Guide
Specifying Label Information

5
Specifying Label Information

■ Interpreted Labels Information Section on page 94

■ The Edit Component Display Options Form on page 98
■ Creating Labels with SKILL on page 102
■ Complex Pole Example on page 103
■ NFET Example on page 104

December 2005 93 Product Version 5.1.41

 Component Description Format User Guide
Specifying Label Information

Interpreted Labels Information Section

You can use interpreted labels to display parameter values, evaluated parameter values, net
connectivity information, backannotated simulation information, and more.

You can set the default label display options in the Interpreted Label Information section of
the Edit Component CDF form. The Interpreted Labels Information section contains different
fields, depending on whether you are editing the CDF description of a library or of a
component.

Scroll down the Edit Component CDF form and look at the Interpreted Labels Information
section. Library label information differs from cell label information. The display that you see
depends on whether you are viewing a library CDF form or of a cell CDF form.

Library Label Information

The following figure shows the Interpreted Labels Information section for a library CDF
description.

Edit Component CDF

You can specify a value of not set for any field. This means that the parameter is deleted from
or not added to the base-level library CDF.

paramDisplayMode is a cyclic field that selects the kind of parameters to display.

December 2005 94 Product Version 5.1.41

 Component Description Format User Guide
Specifying Label Information

■ parameter selects instance or CDF parameter values.

■ model selects model parameter values from simulation.
■ opPoint selects DC or transient operating point results from simulation.
■ none selects no parameters.
■ not set specifies that you made no selection for this field.

paramEvaluate selects a method for displaying parameter values. (These different methods
of display are necessary because a parameter value can be a constant, a design variable, an
expression using constants or design variables, or a value or expression inherited from the
hierarchy.)
■ not set specifies that you want no setting for this field.
■ literal displays the literal parameter value as entered.
■ suffixes converts suffixes in the parameter value to numerical values. For example, 1K
becomes 1000.
■ globals evaluates global design variables in expressions. For example, if the value is
Rin*100 and Rin is 5, the label displays 5 * 100.
■ inheritance evaluates inherited parameters in expressions. For example, if the value is
pPar(Rin)*100 and the parent instance has a parameter Rin set to 10, the label displays
10 * 100.
■ full performs complete numeric evaluation of the expression and displays the result. For
example, if the value is Rin*10 and Rin is 5, it displays 50. If the value is pPar(Rin)*100
and the parent instance has a parameter Rin set to 10, it displays 1000.

paramSimType selects the simulation type for the operating point result that you want to
annotate. The choices are not set, DC, or transient. This field is for operating point
parameters only.

termDisplayMode selects what is displayed next to the component terminals.

■ netName displays the terminal net names.
■ pinName displays the terminal pin names.
■ voltage displays the terminal voltage.
■ current displays the terminal current.
■ not set specifies that you made no selection for this field.

December 2005 95 Product Version 5.1.41

 Component Description Format User Guide
Specifying Label Information

termSimType selects the simulation type for the result that you want to annotate. The valid
choices are not set, DC, or transient. This field is used only when termDisplayMode is
either voltage or current.

netNameType selects the type of net name to display if termDisplayMode is set to

netName. The valid choices are not set, schematic, or simulator.

instDisplayMode selects what is displayed next to the component symbol.

■ instName displays the instance name of the component.
■ cellName displays the cell name of the component.
■ not set specifies that you made no selection for this field.

instNameType selects the type of instance name. The valid choices are
not set, schematic, or simulation. This field is used only when the instDisplayMode is
instName.

December 2005 96 Product Version 5.1.41

 Component Description Format User Guide
Specifying Label Information

Cell Label Information

The Interpreted Labels Information section of the Edit Component CDF form for components
shown in the following figure includes three fields not present in the library version of this
section. These fields are paramLabelSet, opPointLabelSet, and modelLabelSet.

Edit Component CDF

If a parameter name in these lists is preceded by a minus sign (-), the parameter value is
displayed but the parameter name is not. Using the minus sign is the same as setting the
Display Value Only option in the Label Display form.

If there is no minus sign before the parameter name and the value is the default, the display
has the format parameter name : value. If there is no minus sign before the parameter
name and the value is not the default, the display has the format parameterName = value.

Each label set in the following list is limited to the number of cdsParam labels on the cell.

paramLabelSet lists the parameters to be displayed if paramDisplayMode is set to

parameter. This list is limited to the number of cdsParam labels on the cell.

opPointLabelSet lists the parameters to be displayed if paramDisplayMode is set to

opPoint.

December 2005 97 Product Version 5.1.41

 Component Description Format User Guide
Specifying Label Information

modelLabelSet lists the parameters to be displayed if paramDisplayMode is set to model.

Creating Labels Using the CDF Editor

You can make selections for label information for both library and cell CDFs. You can use the
following procedure to create labels for a library CDF.
1. Scroll to the Interpreted Labels Information section of the Edit Component CDF form.
2. Make edits for library or cell CDF label specifications.
3. Click Apply to apply these specifications to the current CDF.

The Edit Component Display Options Form

You can control the creation of CDF labels by using the Component Display command in
the Edit menu in the schematic editor to open the Edit Component Display Options form.

December 2005 98 Product Version 5.1.41

 Component Description Format User Guide
Specifying Label Information

Each section shown below is displayed when you click on the corresponding Select Label
button.

As you click each of

these buttons,

these sections
appear.

When you change the display of information in the Edit Component Display Options form, the
display of ilLabels (labels generated automatically during automatic symbol generation) and
interpreted labels is affected. ilLabels are usually defined on symbols and are displayed

December 2005 99 Product Version 5.1.41

 Component Description Format User Guide
Specifying Label Information

when you place the symbol in a schematic. For example, the following figure shows the
symbol for a capacitor from analogLib:

The cdsTerm( ) labels (usually placed on the layer annotate drawing8) display information
about the pin or about the nets attached to the pin. These labels must be of the form
cdsTerm(pinName). In the Edit Component Display Options form, you use the Terminal
Labels settings to control the cdsTerm labels.

If pinName contains special characters, you must quote or escape them properly.

The cdsName( ) labels (usually placed on the layer annotate drawing7) display information
about the cell name or the instance name. In the Edit Component Display Options form, you
use the Instance Labels settings to control the cdsName label.

The cdsParam( ) labels (usually placed on the layer annotate drawing) display information
about parameter values or backannotated parameter values. Three labels are usually
generated during automatic symbol generation, but you can define additional labels. The only
requirement for the parameter labels is that they are sequential, starting with 1 if they are
numbers. In the Edit Component Display Options form, you use the Parameter Labels
settings to control the cdsParam labels.
Note: cdsParam( ) can also take parameter names. Therefore, as cdsParam(n), you can
use cdsParam(s_paramName), for example, cdsParam(’pm), where pm is a parameter
name. You must declare the parameter names in the paramLabelSet field of the Edit
Component CDF form. The cdsParam points to those parameters in an orderly manner. If you
change the order of those parameters, the displayed values will be changed.

In the Edit Component Display Options form, there are three cyclic fields for parameter name
labels. If you add a fourth parameter label to your cell, a fourth cyclic field appears on the Edit
Component Display Options form. The syntax for the fourth parameter label is cdsParam(4).

If you choose the library option in the Apply To field in the Edit Component Display Options
form, the display information is written to the library user-level CDF. If you choose the cell
option, the display information is written to the cell user-level CDF. If you choose the instance
option, the display information is saved as properties on that instance. Only the instance label

December 2005 100 Product Version 5.1.41

 Component Description Format User Guide
Specifying Label Information

display is saved when you close the library because user-level CDF data applies only while
the library is open. (It is not written to the database or saved when you exit the Cadence®
software.) To keep this library and cell CDF information, copy it into a library or base CDF
description that is permanently saved.

Creating Labels in the Schematic Editor

You can set up label display with the Edit Component Display Options form and the Edit
Component CDF form as follows:
1. Set up your component and place it in a schematic.
2. Choose Edit–>Component Display to set up the labels you want placed with the
component.
The Edit Component Display Options form appears.
3. On the Edit Component CDF form, look at the effective-level CDF description for the
labeled component.
The effective-level cell CDF is the combination of the base-level CDF (with all parameter
and simulation information defined) and the user-level CDF that you created with the Edit
Component Display form.
4. Type in a filename and click on Save to save the contents of the form in a file.
5. Change the CDF type to base.
6. Enter the same filename and click Load.
The effective-level CDF is copied to a base-level CDF with all of the interpreted label
information set up as you saw it.
7. Click OK to save this new base CDF description.
8. Repeat this procedure to create a library CDF description.

When you place a component in the schematic editor, the Add Component form shows the
parameters associated with this component. No CDF properties (labels) appear unless you
assign parameter values other than the default values. If you want to see the value of any
parameter displayed in the design window, you can toggle the display setting for that
parameter to make it visible.

December 2005 101 Product Version 5.1.41

 Component Description Format User Guide
Specifying Label Information

Creating Labels with SKILL

You can use the following Cadence SKILL language functions with the right arrow operator
(->) to select label specifications. You enter the specifications as properties of either the
library or cell CDF. Not all properties need to be set. Enter only those properties that you want
set.

The following SKILL expressions set label CDF specifications at the library level:
■ paramDisplayMode setting
libId = ddGetObj("analogLib")
cdfId = cdfCreateBaseLibCDF(libId)
cdfId->paramDisplayMode = “parameter”

■ paramEvaluate setting
cdfId->paramEvaluate = "t nil nil nil nil"

For the paramEvaluate option list, the first entry applies to literal, the second to suffixes,
the third to globals, the fourth to inheritance, and the fifth to full.
■ paramSimType setting
cdfId->paramSimType = "DC"

■ termDisplayMode setting
cdfId->termDisplayMode = "netName"

■ termSimType setting
cdfId->termSimType = "DC"

■ netNameType setting
cdfId->netNameType = "schematic"

■ instDisplayMode setting
cdfId->instDisplayMode = "instName"

■ instNameType setting
cdfId->instNameType = "schematic"

The following expressions set label CDF specifications at the cell level:
■ paramLabelSet setting
cell = ddGetObj(libId "NFET")
cdfId = cdfCreateBaseCellCDF(cell)
cdfId->paramLabelSet = "-model l w"

■ opPointLabelSet setting
cdfId->opPointLabelSet = "id vgs vds"

December 2005 102 Product Version 5.1.41

 Component Description Format User Guide
Specifying Label Information

■ modelLabelSet setting
cdfId->modelLabelSet = "vfb phi eta"

Complex Pole Example

In previous chapters, you used the CDF editor on the complexPole1 component to look at
the Component Parameters and Simulation Information sections of the CDF description. Now
you look at the Interpreted Labels Information in the complex pole CDF description and see
how it affects the display of the component’s symbol.

The following figure shows the Interpreted Labels Information section of the Edit Component
CDF form for complexPole1.

Edit Component CDF

1. Look at the complexPole1 symbol in the Symbol Editor. (Open the complexPole1
design, specifying the symbol cellview.)

December 2005 103 Product Version 5.1.41

 Component Description Format User Guide
Specifying Label Information

You see a generic version of the component, with no values displayed.

cdsName()

COMPLEX
cdsTerm(in) POLE 1 cdsTerm(out)
in out

param(1)
param(2)
param(3)

2. Open an empty schematic editor window by creating a new design.

3. Use the Add Component command to add a complexPole1 component to the
schematic.
Its symbol looks like this:

X3
COMPLEX
POLE 1
in out

sigma:-5
wn:6283.1853

The two default parameter values from the Component Parameters section and the instance
name are set to be displayed (with paramDisplayMode and instDisplayMode). However,
termName is set to netName, and the component is by itself, not connected to any nets.
There are no net names to display on the terminals (in and out).

NFET Example
The NFET symbol in the following figure shows the effect of setting paramDisplayMode to
parameter and paramLabelSet to -model l w. Three parameters, model, l, and w, are
displayed. There is no parameter name displayed with bmemne (the model name) because
a minus (-) symbol precedes the parameter name model in paramLabelSet.

The default values from the CDF description are displayed for the l and w parameters, as
shown by the format of parameter name : value (l:3u and w:7u). In addition,

December 2005 104 Product Version 5.1.41

 Component Description Format User Guide
Specifying Label Information

termDisplayMode is set to netName and instDisplayMode is set to instName. Unlike the

complex pole example, this transistor is connected to nets in a schematic, so the net and
instance names are all displayed.

Try changing the fields to different settings and then view the results.

You can use labels in any view. For example, you can set up a layout with labels so that you
can see instance and parameter information when you place an instance:

Layout Master Instance

c = 128f
w = 20
cdsParam(1) l = 20
cdsParam(2)
cdsParam(3)

I1
cdsName()

This can be especially useful when you build the ivpcell used in layout extraction. In this way,
when you look at the extracted layout, you see the instance found and the parameters
measured.
Note: These labels do not pick up hierarchical information, so cdsName will only show the
current name, not the hierarchical name.

December 2005 105 Product Version 5.1.41

 Component Description Format User Guide
Specifying Label Information

December 2005 106 Product Version 5.1.41

 Component Description Format User Guide
Other CDF Information

6
Other CDF Information

■ Other Information Section on page 108

■ Setting Up Fields and Prompts on page 109
■ Initialization Procedure on page 112
■ Postprocessing Procedure on page 113

December 2005 107 Product Version 5.1.41

 Component Description Format User Guide
Other CDF Information

Other Information Section

You use the last section of the Edit Component CDF form to specify the sizes of input fields
and the initialization and closing procedures to be executed when a new component is
instantiated.

The first four fields let you control the size of parameters on an instantiation form (such as the
Add Component or Edit Properties forms). These fields do not affect the display in the Edit
Component CDF form.

The last two fields let you apply procedures to the CDF. These procedures are performed
when you instantiate the component or when you change the component’s parameters.

buttonFieldWidth lets you specify the width of a button on the instantiation form. The
default width is 350 pixels.

fieldHeight lets you specify the height of a field on the instantiation form. The default height
is 35 pixels.

fieldWidth lets you specify the width of a field on the instantiation form. The default width is
350 pixels.

promptWidth lets you specify the width of the prompt on the instantiation form. The default
width is 175 pixels.

formInitProc lets you specify an optional procedure (a Cadence® SKILL language routine
that you provide) that executes automatically when the component is placed on an
instantiation form.

doneProc lets you specify an optional procedure (a SKILL routine that you provide) that
executes after you change any parameter on the instantiation form.

December 2005 108 Product Version 5.1.41

 Component Description Format User Guide
Other CDF Information

Setting Up Fields and Prompts

The input fields and prompts for a component are displayed on an instantiation form, such as
the Add Component or Edit Properties forms. The following examples show the Add
Component form as it appears when a component is instantiated. The sample CDF
description in these examples defines four parameters: a button, a string, a cyclic field, and
a radio button field.

Adjusting Button Fields

You can adjust the width of a field with the buttonFieldWidth parameter in the Edit
Component CDF form. The following examples show button fields on two Add Component
forms. The first form shows the button with the default width of 340 pixels, and the second
form shows the button with a width of 500 pixels.

Form with default button Form with buttonFieldWidth

FieldWidth (340 pixels) set to 500 pixels

Add Component Add Component

Default button size Wide button

December 2005 109 Product Version 5.1.41

 Component Description Format User Guide
Other CDF Information

Adjusting Field Heights

You can adjust the height of a field with the fieldHeight parameter in the Edit Component
CDF form. The following example shows fields on two Add Component forms. The first form
shows the fields at the default height of 35 pixels, and the second form shows the field height
at 70 pixels.

Form with default fieldHeight (35 pixels) Form with fieldHeight set at 70 pixels
Add Component Add Component

Default fields

Tall fields

December 2005 110 Product Version 5.1.41

 Component Description Format User Guide
Other CDF Information

Adjusting Field Widths

You can adjust the width of a field with the fieldWidth parameter in the Edit Component CDF
form. This parameter adjusts the width of string input fields only. (You adjust the width of a
button field with the buttonFieldWidth parameter.) The following examples show string input
fields on two Add Component forms. The first form shows the field at the default width of 350
pixels, and the second shows the field width of 500 pixels.

Form with default Form with fieldWidth

fieldWidth (350 pixels) set to 500 pixels

Add Component Add Component

Default field Wide field

December 2005 111 Product Version 5.1.41

 Component Description Format User Guide
Other CDF Information

Adjusting Prompt Widths

You can adjust the width of a prompt with the promptWidth parameter in the Edit Component
CDF form. The following example shows prompts on two Add Component forms. The first
form shows the prompts at the default width of 175 pixels, and the second form shows the
prompt width of 200 pixels:

Form with default Form with promptWidth

promptWidth (175 pixels) set to 200 pixels
Add Component Add Component

Default prompt width Wider prompt width

Initialization Procedure
The formInitProc field in the Other Information section of the Edit Component CDF form lets
you enter the name of an optional procedure for preprocessing CDF data. The procedure
executes when the contents of the CDF are displayed on a form, such as the Add Component
form or the Edit Properties form. The procedure takes a single argument, cdfDataId, which
is the database object that represents the CDF description for the component being loaded
into the form. You enter only the function name (with no parentheses) in the formInitProc
field.

In the NFET example, you might enter an initialization procedure to reset all NFET
parameters to their defaults each time the component is displayed on a form. (This is only an

December 2005 112 Product Version 5.1.41

 Component Description Format User Guide
Other CDF Information

example. It is not practical to reset parameters to their default values every time you return to
the Edit Properties form.)

Enter the procedure name, setDefaults, in the formInitProc field. This procedure takes only
one argument, cdfDataId, which is used to point to the parameters and values of the
associated component.
You can create the SKILL procedure with a standard editor. You must load the file in order for
the initialization procedure to be called and run. (See “Loading Callbacks” on page 117 for
details about loading or attaching a SKILL procedure.)
procedure(setDefaults(cdfDataId)
;used as formInitProc to set all nfet values to the
;default values as form is entered
cdfDataId->w->value = cdfDataId->w->defValue
cdfDataId->width->value = cdfDataId->width->defValue
cdfDataId->l->value = cdfDataId->l->defValue
cdfDataId->length->value=cdfDataId->length->defValue
)

An initialization procedure runs every time the form is updated. The procedure runs when you
change a value on the component (a callback), and when you point to a new component.

You can also enter the following SKILL expression to set the formInitProc property when you
create the cell CDF:
cdfId = cdfCreateBaseCellCDF( cellId )
.
.
.
cdfId->formInitProc = "setDefaults"

Postprocessing Procedure
The doneProc field in the Other Information section of the Edit Component CDF form lets
you enter the name of an optional procedure to be executed after you change a parameter on
a component instance. You execute the procedure when you click on Apply on the Edit
Properties form or when you place an instance of the component using the Add Component
form.

For the NFET example, a doneProc procedure might recalculate the simulation parameters
every time the instance changes. Enter the procedure name in the doneProc field. In the
following example, the procedure name is calcSimParams. This procedure takes a single
argument, cellId, that identifies the cell of the instance that was modified. Throughout the
procedure, cdfgData, the global variable for CDF information, accesses the CDF
information.

December 2005 113 Product Version 5.1.41

 Component Description Format User Guide
Other CDF Information

You can create the SKILL procedure with a standard editor. You must load this file in order for
the procedure to be called and run. (See “Loading Callbacks” on page 117 for details about
loading or attaching a SKILL procedure.)
procedure(calcSimParams(cellId)
cdfgData = cdfGetInstCDF(cellId)
tmp=cdfParseFloatString(cdfgData->w->value)
cdfgData->ad->value=sprintf(s "%g" (tmp * 7.0 * 1e-6))
cdfgData->as->value=sprintf(s "%g" (tmp * 7.0 * 1e-6))
cdfgData->pd->value=sprintf(s "%g" ((tmp + 7e-6) * 2))
cdfgData->ps->value=sprintf(s "%g" ((tmp + 7e-6) * 2))
)

You can also enter the following SKILL expression to set a postprocessing procedure when
you create the cell CDF:
cdfId = cdfCreateBaseCellCDF( cellId )
...
cdfId->doneProc = “calcSimParams”

December 2005 114 Product Version 5.1.41

 Component Description Format User Guide
Writing Callbacks

7
Writing Callbacks

■ Overview on page 116

■ Loading Callbacks on page 117
■ NFET Example on page 118

December 2005 115 Product Version 5.1.41

 Component Description Format User Guide
Writing Callbacks

Overview
A callback procedure is a Cadence® SKILL language expression that is evaluated when a
parameter value changes. Using callbacks is a powerful way to control relationships between
parameters and the restrictions on the parameter itself.
Note: Cadence advises you to have sufficient knowledge of callback procedures before
using them.

Callbacks can be GUI based or database based. GUI based callbacks occur when you modify
the values in the parameter form fields. Database based callbacks occur when the parameter
value is modified via a database action. The CDF parameter callback is primarily a GUI based
callback. A GUI based callback is active when the CDF parameters are displayed in the Add
Instance form or the Edit Object Properties form when you use the Create Instance or Edit
Properties commands.
Note: The CDF parameter callback is invoked only when you use the Create Instance or the
Edit Properties commands. If you modify the parameter values through other commands then
the CDF parameter callback is not invoked.

You can enter a callback as a SKILL expression that you type directly in the callback field of
the parameter section of the Edit Component CDF form. You can also enter a callback by
using the ?callback keyword as an argument to the cdfCreateParam function:
cdfParamId = cdfCreateParam(cdfDataId
?type "string"
?parseAsNumber "yes"
?units "lengthMetric"
?parseAsCEL "yes"
?storeDefault "no"
?name "pName2"
?prompt "width"
?defValue 7u
?display "artParameterInToolDisplay(’w)"
?editable "cdfgData->maskLayoutViewName->..."
?callback "cdfgData->pName1->value =
cdfgData->pName2->value")

In this example, the callback on parameter pName2 sets the value of parameter pName1 to
the new value when the value of pName2 changes.

A callback can also be a call to a function that you loaded previously.

cdfParamId = cdfCreateParam(cdfDataId
?type "string"
?parseAsNumber "yes"
?units "lengthMetric"
?parseAsCEL "yes"
?storeDefault "no"
?name "pName2"
?prompt "width"
?defValue 7u

December 2005 116 Product Version 5.1.41

 Component Description Format User Guide
Writing Callbacks

?display "artParameterInToolDisplay(’w)"
?editable "cdfgData->maskLayoutViewName->..."
?callback "myFunc( )")

In the second example, the callback on parameter pName2 calls the function myFunc( )
when the value of pName2 changes.

When using callbacks, always use the global variable cdfgData to access information about
parameter values and default values.

To get the default value for a parameter, type the following expression:
cdfgData->paramName->defValue

To get the current value for a parameter, type the following expression:
cdfgData->paramName->value

To set the value of a parameter, type either of the following expressions:

cdfgData->paramName->value = 10
cdfgData->paramName->value =
cdfgData->paramName->defValue

Loading Callbacks
Once you define the SKILL callback procedures, you need to load the files. From the
Command Interpreter Window in your .cdsinit file, type
load("path/callbacks.il")

You must make sure that the callbacks are loaded before you use the library and that the
callback files are archived whenever you archive the library.

As an alternative, you can attach a callback file to a library. You attach callback files to a library
by including a libInit.il file in that library. libInit.il is a text file that contains calls
to load other SKILL files or contexts in the library or to initialize autoload properties.

You must always use the file name libInit.il or libInit.ils. You can have one
such file in each library. The following example is the contents of a libInit.il file in a
microwave library:
load( ddGetObj( "microwave" nil nil "libraProcs" )~>path )
load( ddGetObj( "microwave" nil nil "hpmnsProcs" )~>path )
load( ddGetObj( "microwave" nil nil "mharmProcs" )~>path )
load( ddGetObj( "microwave" nil nil "callbacks" )~>path )

where ddGetObj uses the following syntax:

ddGetObj( libName cellName viewName fileName)

December 2005 117 Product Version 5.1.41

 Component Description Format User Guide
Writing Callbacks

If you attach the file to the library, the file is automatically loaded whenever the library is
opened. This guarantees that the functions are defined. In addition, when you archive your
library, the CDF callback functions are automatically archived.

NFET Example
To understand how to use callbacks, look at the NFET CDF description. The following
callback functions are specified for the NFET component:
■ mosLayout( ) on parameter maskLayoutViewName
■ pcMOSw( ) on parameter w
■ pcMOSl( ) on parameter l

December 2005 118 Product Version 5.1.41

 Component Description Format User Guide
Writing Callbacks

mosLayout Callback
In the NFET CDF description, when you select the layout.pwr option of the
maskLayoutViewName parameter, the values of the other parameters are constant. You
can use the mosLayout callback procedure to set these values automatically whenever the
maskLayoutViewName is layout.pwr. As the following example shows, whenever the value
of maskLayoutViewName changes, the callback checks the value of
maskLayoutViewName and, if the value is layout.pwr, sets the parameter values:
procedure( mosLayout( )
if((cdfgData->maskLayoutViewName->value ==
cdfgData->maskLayoutViewName->defValue)
then ; It is layout. Use defaults.
cdfgData->w->value=cdfgData->w->defValue
cdfgData->width->value=cdfgData->width->defValue
cdfgData->l->value=cdfgData->l->defValue
cdfgData->length->value=cdfgData->length->defValue
cdfgData->bend->value=cdfgData->bend->defValue
cdfgData->bendt->value=cdfgData->bendt->defValue
cdfgData->ad->value=cdfgData->ad->defValue
cdfgData->as->value=cdfgData->as->defValue
cdfgData->pd->value=cdfgData->pd->defValue
cdfgData->ps->value=cdfgData->ps->defValue
cdfgData->diffD->value=cdfgData->diffD->defValue
else ; It is layout.pwr. Use known values.
cdfgData->w->value="880u"
cdfgData->width->value=880
cdfgData->l->value="12u"
cdfgData->length->value=12
cdfgData->bend->value=cdfgData->bend->defValue
cdfgData->bendt->value=cdfgData->bendt->defValue
cdfgData->ad->value="6.16e-09"
cdfgData->as->value="6.16e-09"
cdfgData->pd->value="0.001774"
cdfgData->ps->value="0.001774"
cdfgData->diffD->value=nil
)
)

December 2005 119 Product Version 5.1.41

 Component Description Format User Guide
Writing Callbacks

pcMOSw Callback
The pcMOSw callback is similar to the pcMOSl callback. pcMOSw checks the input, sets
the width parameter, determines if there should be a bend, and calculates some of the
simulation parameters:
procedure( pcMOSw( )
prog(()
if( (cdfgData->maskLayoutViewName->value ==
cdfgData->maskLayoutViewName->defValue)
; Do this only if maskLayoutViewName is
; layout. Otherwise, mosLayout sets it.
then
tmp = evalstring(cdfgData->w->value)
if( (tmp && (typep(tmp)!=’flonum))
then
artError("Width value must be a floating
point number. Set to default." )
cdfgData->w->value = cdfgData->w->defValue
cdfgData->width->value =
cdfgData->width->defValue
cdfgData->bend->value =
cdfgData->bend->defValue
cdfgData->bendt->value =
cdfgData->bendt->defValue
cdfgData->ad->value = cdfgData->ad->defValue
cdfgData->as->value = cdfgData->as->defValue
cdfgData->pd->value = cdfgData->pd->defValue
cdfgData->ps->value = cdfgData->ps->defValue
return(nil)
)
tmp = ( tmp / 1e-6 )
if( (tmp < cdfgData->width->defValue)
then
artError("Width value is less than
minimum (7u). Set to default." )
cdfgData->w->value = cdfgData->w->defValue
cdfgData->width->value =
cdfgData->width->defValue
cdfgData->bend->value =
cdfgData->bend->defValue
cdfgData->bendt->value =
cdfgData->bendt->defValue
cdfgData->ad->value = cdfgData->ad->defValue
cdfgData->as->value = cdfgData->as->defValue
cdfgData->pd->value = cdfgData->pd->defValue
cdfgData->ps->value = cdfgData->ps->defValue
return(nil)
)
if( (tmp > 240 )
then
artError("Width value is greater than
maximum (240u). Set to default." )
cdfgData->w->value = cdfgData->w->defValue
cdfgData->width->value =
cdfgData->width->defValue
cdfgData->bend->value =
cdfgData->bend->defValue
cdfgData->bendt->value =

December 2005 120 Product Version 5.1.41

 Component Description Format User Guide
Writing Callbacks

cdfgData->bendt->defValue
cdfgData->ad->value = cdfgData->ad->defValue
cdfgData->as->value = cdfgData->as->defValue
cdfgData->pd->value = cdfgData->pd->defValue
cdfgData->ps->value = cdfgData->ps->defValue
return(nil)
)
if( (tmp <= 240 )
then
cdfgData->maskLayoutViewName->value =
cdfgData->maskLayoutViewName->defValue
grid = round(tmp/0.5)
newwidth = grid*0.50
if( (newwidth != tmp)
then
artError("Width is set to nearest value
on 0.50 micron pitch.")
cdfgData->w->value =
sprintf(s "%g" (newwidth * 1e-6))
)
if( (tmp > 120) ;criteria for setting bend here
then
cdfgData->width->value = tmp/2
cdfgData->bend->value = t
cdfgData->bendt->value = 1
cdfgData->diffD->value = nil
else
cdfgData->width->value = tmp
cdfgData->bend->value =
cdfgData->bend->defValue
cdfgData->bendt->value =
cdfgData->bendt->defValue
)
)
cdfgData->ad->value =
sprintf(s "%g" (tmp * 7.0 * 1e-12))
cdfgData->as->value =
sprintf(s "%g" (tmp * 7.0 * 1e-12))
cdfgData->pd->value =
sprintf(s "%g" ((tmp + 7.0) * 2e-6))
cdfgData->ps->value =
sprintf(s "%g" ((tmp + 7.0) * 2e-6))
)
)
)

December 2005 121 Product Version 5.1.41

 Component Description Format User Guide
Writing Callbacks

pcMOSl Callback
When the length and width parameters for the NFET CDF description were created, the l
and w parameters already existed for one of the simulators. Instead of prompting the user
twice for the same information, callbacks attached to the l and w parameters can update the
length and width parameters automatically.

The pcMOSl callback runs whenever the value of l changes. The callback verifies that the
maskLayoutViewName parameter value is layout. (For the other option, layout.pwr, the
parameter values are controlled by the mosLayout callback, shown in the previous section.)
The callback then checks the new value against the minimum value, the maximum value, and
the grid. If the value is not valid, the callback resets the lengths to the default value. If the value
is valid, the callback sets the length parameter value. The following example shows this
process:
procedure( pcMOSl( )
prog(()
if((cdfgData->maskLayoutViewName->value ==
cdfgData->maskLayoutViewName->defValue)
; Do this only if maskLayoutViewName is
; layout. Otherwise, mosLayout sets it.
then
; Convert input string to float.
tmp = evalstring(cdfgData->l->value)
if( (tmp && (typep(tmp)!=’flonum))
then
artError("Length value must be a floating point
number. Set to default." )
; artError will send message to CIW
cdfgData->l->value = cdfgData->l->defValue
cdfgData->length->value =
cdfgData->length->defValue
return(nil)
)
tmp = ( tmp / 1e-6 ) ; Convert to user units.
if( (tmp < 3.0)
then
artError("Length is less than minimum (3u).
Set it to default." )
cdfgData->l->value = cdfgData->l->defValue
cdfgData->length->value =
cdfgData->length->defValue
return(nil)
)
if( (tmp > 15.0 )
then
artError("Length value is greater than
maximum (15u). Set it to default." )
cdfgData->l->value=cdfgData->l->defValue
cdfgData->length->value =
cdfgData->length->defValue
return(nil)
)
; The following checks the value against
; the grid and fixes it if necessary.

December 2005 122 Product Version 5.1.41

 Component Description Format User Guide
Writing Callbacks

grid = round(tmp/0.5)
newlength = grid*0.50
if( (newlength != tmp)
then
or("Length is set to nearest value
on 0.50 micron pitch.")
cdfgData->l->value = sprintf(s "%g"
(newlength * 1e-6))
)
; Set the pcell param value.
cdfgData->length->value = tmp
)
)
)

December 2005 123 Product Version 5.1.41

 Component Description Format User Guide
Writing Callbacks

December 2005 124 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

8
CDF SKILL Summary

■ Overview on page 126

■ CDF SKILL Function Elements on page 127
■ CDF SKILL Functions on page 138

December 2005 125 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

Overview
Short examples in the previous chapters describe how to use the Cadence® SKILL language
for specific operations. This chapter:
■ Covers all CDF SKILL functions in general
■ Describes each type of element in a CDF SKILL function
■ Summarizes all the CDF SKILL functions

When you use these SKILL functions, you can perform all the same operations that you do
when you fill in form fields and apply them. However, unlike filling out forms, with SKILL you
can create routines to perform the same operation on many components at once, automating
your handling of CDF descriptions.

December 2005 126 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

CDF SKILL Function Elements

CDF SKILL functions use or operate on data IDs, parameters, parameter attributes,
expressions, and global variables. This section describes each of these elements in the
context of SKILL.

Cell and Library Data IDs

Before working on a CDF description, you must specify the data ID of the cell or library. You
can get the data ID for the cell or library you are using with the ddGetObj() function. With
SKILL versions earlier than 4.4, you used the dmFindLib() and dmFindCell() functions.
You must assign the value returned by these functions to a variable that you create. For
example, to operate on the analogLib library, you must first create the variable mylib with
this assignment:
mylib = ddGetObj("analogLib")

When creating the variable for the object ID of a cell, you must specify both the library name
and the cell name:
test_cell = ddGetObj("analogLib" "schottky")

or
test_cell = ddGetObj(mylib "schottky")

if you have already defined mylib.

You must then use mylib when a SKILL function requires a d_id or d_libId, and test_cell
when a SKILL function requires a d_id or d_cellId.

You can use this data ID to access information about the description by using the right arrow
( -> or ~>) operator. For example
test_cell -> name

returns the ID
"schottky"

Data Objects
CDF descriptions are represented by cdfDataId objects, which are SKILL objects that you can
manipulate like other SKILL database objects. Just like data IDs, you assign the cdfDataId
of a particular CDF description to a variable that you create. The most common variable name
is cdfDataId. To create a new CDF data object for a new CDF description, use this type of
ID assignment:

December 2005 127 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

newCellId = cdfCreateUserCellCDF(test_cell)
cdf:25092160
newCellId->type
"userCellData"

To access an existing CDF data object for an existing CDF description, use this type of ID
assignment:
baseCell = cdfGetBaseCellCDF(test_cell)
baseCell->dataFile->value
"bjt"

newCellId and baseCell are arbitrary names that you create.

Although you can add any information to a cdfDataId, the object has specific fields to hold the
information that it maintains. These fields are described in the following section.

id

The database object (cell ID or library ID) to which the cdfDataId is attached. This field is not
editable.

type

There are seven types of cdfDataId:

baseLibCDF base-level library CDF
userLibCDF user-level library CDF
effLibCDF effective-level library CDF
baseCellCDF base-level cell CDF
userCellCDF user-level cell CDF
effCellCDF effective-level cell CDF
effInstCDF effective-level instance CDF

This field is not editable.

parameters

List of parameters attached to this cdfDataId. This field is not editable.

December 2005 128 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

doneProc

An optional procedure name (a string) that is evaluated after any change to a parameter on
a component instance. You can use doneProc for post-processing. The procedure must take
a single argument: the instance that has been modified.

The default value is nil. This field is editable.

formInitProc

An optional procedure name. If specified, the procedure is executed when the contents of the
CDF are displayed on a form. The default value is nil. This field is editable.

This procedure runs when you use the Add Component and Edit Properties commands. You
can use the procedure for preprocessing CDF data. This procedure must take a single
argument, the cdfDataId being added to the form.
Note: When you modify a parameter field value using the formInitProc, the Edit Properties
Object command for the schematic application is not aware of the modification and does not
update the changes made with formInitProc. You can avoid this problem by setting the
variable cdfgForm->cdfModified to the value t.

Displaying Parameters

The following fields control the display of parameters on a form.

fieldWidth Width of a field Default = 350 pixels

fieldHeight Height of a field Default = 35 pixels
buttonFieldWidth Width of a button field Default = 340 pixels
promptWidth Width of a prompt Default = 175 pixels

Parameters
CDF parameters are represented by cdfParamId objects, which are SKILL objects that you
can manipulate like other SKILL database objects. You can use a cdfParamId to access
information about a parameter by using the right arrow ( ->) operator. Although you can add
any information to a cdfParamId, the object has fields to hold information that it maintains.

In addition to getting the cdfParamIds through the parameters field of a cdfDataId, you can
also access a parameter by specifying:

December 2005 129 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

cdfDataId->paramName

paramName is the name of the parameter. However, you must not create any user-defined
properties on a cdfDataId that conflict with the name of a parameter on the cdfDataId.

use

Specifies whether to use this parameter. The use attribute is context-specific and is evaluated
when necessary. In this field, you can specify parameters that you are not using because of
the value of other parameters, or because of the function you are using. The use field must
be a string.
■ If the field evaluates to non-nil, the system uses the parameter.
■ If a value for the field exists, the system ignores the value.
■ If the field evaluates to nil, the system does not use the parameter.
■ If you do not specify this field, t is assumed, so the system always uses the field.

The “Global Variables” section describes several global variables for constructing the use
expression.

This field is editable.

paramType

Type of the parameter. Valid values for this field are

"string"
"int"
"float"
"radio"
"cyclic"
"boolean"
"button"
“netSet”

The quotation marks are required. CDF checks that the parameter value matches the type
specified. You must specify this field.

This field is not editable.

December 2005 130 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

Note: Use the netSet type to store inherited connections in CDF. Inherited connections allow
you to selectively override global signals in designs created in Virtuoso Schematic Editor. The
override information is communicated through net expressions and netSet properties. For
more information on netSet properties refer to Inherited Connections Flow Guide and
Virtuoso Schematic Composer User Guide.

defValue

Default value of the parameter. You must specify this field.

This field is editable.

value

Current value of the parameter.

This field is editable.

prompt

Prompt that appears on a form field when an application asks for the value of this parameter.
The value for this field must be a string. You must specify this field.

This field is editable.

choices

Possible values (a list of strings) of a cyclic or radio type parameter. Do not use this field for
other types of parameters.

This field is editable.

units

Unit type of the parameter. You often modify parameters that represent resistance,
capacitance, length, time, power, and so forth, and expect to see the parameter value scaled
to appropriate units (for example, pF, uM, dBm). Valid values for this field include the
following strings:
"resistance"
"capacitance"

December 2005 131 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

"inductance"
"conductance"
"time"
"frequency"
"power"
"powerDB"
"lengthMetric"
"lengthEnglish"
"angle"
"voltage"
"current"
"temperature"

The quotation marks are required. This field determines the unit’s suffix and scale factor to
use when displaying the parameter value. For example, by setting a parameter’s units field to
capacitance, the parameter value is displayed as 5 pF instead of 5e-12.

editable

Specifies whether a parameter is for display only. If set to nil, the field specifies a displayed
parameter that you cannot change. (Other parameters’ callbacks can change the parameter.)
Use this field only on int, float, and string type fields.

The editable field must be a string. This field is evaluated when necessary.
■ If the field evaluates to non-nil, the parameter is editable.
■ If the field evaluates to nil, the parameter is grayed out when displayed as a form field.
■ If you do not specify the field, t is assumed, meaning that the field is editable.

callback

SKILL code to be executed whenever the value of a parameter changes. Using callback, you
can cause any parameter’s value to affect any other parameter’s value. The value of this field
must be a string.

This field is optional. If you do not use callback, no callback is executed.

December 2005 132 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

The “Global Variables” section describes several global variables you can use in the callback
field.

parseAsNumber

String type parameter whose value can evaluate to a floating-point number. These types of
parameters often occur for circuit level simulators that allow component parameters to be
either numbers or expressions containing variables.

If you specify this field, the system parses the value of the parameter to check if the value is
a number.
■ If the value is a number, the system converts the string to a floating point number,
converts the string to the most efficient notation (taking into account any units field
specified), then reconverts the number into a string.
■ If the parameter value is not a number, the system does no conversion.

For example, if the parameter value is the string 5.4e-12 and the parameter value has a units
field of capacitance (that is, the suffix is F), the parameter value is converted to the string 5.4p
before being displayed. If, however, the parameter value is c1, no conversion takes place.

dontSave

Specifies that the parameter cannot be stored in any instance that corresponds to the
component.

The dontSave field must be a string. The system evaluates the field as necessary. If the field
evaluates to non-nil, the parameter is not stored. If you do not specify the field, nil is assumed
and the parameter value is stored.

parseAsCEL

If set to yes, the associated CDF parameter is processed as a CDF Expression Language
(CEL) expression. The parameter must be a string. If the expression resolves to a numeric
value (the usual case), the parseAsNumber flag should also be set to yes.

storeDefault

Specifies whether the parameter default value is saved as a property of the instance. All tools
based on the Cadence Analog Design Environment software use the CDF to find default
values if no property exists on an instance.

December 2005 133 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

If set to no (the default) or don’t use, a property is not saved on the instance when the
parameter value is the default. Also, if the default value of the parameter changes, all
instances that use the default automatically get the new default value. (To see the change in
an open window, you must select Window – Redraw from the Cadence menu.)

If set to yes and the parameter value is set to the same value as the default, a property is
saved on the instance. One disadvantage of this attribute is that if the default value of a
parameter changes, the instances that use the default do not automatically change to the new
default. If you are a user of the Open Simulation Software (OSS) system, and you used to set
this attribute to yes because you could only netlist using instance properties, you can now set
it to no because OSS users now have the option of using CDF.

display

Determines if this parameter is displayed in forms that display CDF parameters, such as the
Edit Properties form or the Add Component form. You must enter t, nil, or a SKILL
expression that evaluates to t or nil in this field to determine if this parameter is to be
displayed. If the field evaluates to non-nil (the default), the parameter is displayed. If the
field evaluates to nil, the parameter is not displayed.

Example

You can use the cdfDataId baseCell from the previous example to examine the parameters
in the base-level cell CDF description of the Schottky transistor in the analog library. For
example, that cell has a parameter m. You can access information about m in the following
manner:
baseCell->m->prompt
"Multiplier"
baseCell->m->editable
nil
baseCell->m->display
"artParameterInToolDisplay(’m)"
baseCell->m->paramType
"string"

Expressions
CEL (CDF Expression Language) is another name for the Analog Expression Language
(AEL) that works with CDF parameters. For information on AEL, refer to the AEL Reference
Manual. With AEL you can express a value as a mathematical expression instead of a single,

December 2005 134 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

fixed value. In such an expression, symbolic names such as sheetResistivity can refer to
values that are computed and set on one of the following:
■ CDF parameter on the same design component
■ CDF parameter on the parent design component

December 2005 135 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

Global Variables
CDF parameter values interact with each other, and one parameter’s value can affect the
existence of another. This feature is implemented primarily through the use and callback
fields of the parameters section. The following global variables, which you can access, are
set whenever parameter fields are evaluated.

cdfgData

CDF data for the component in use. Use the value field to get the current value of any
parameter in the CDF description.
■ For creating an instance, set the field to the last value you used when you created a
component of this type.
■ For editing, set the field to the value for the component you are editing.

For example, you might set the use field for the resistance parameter to
"cdfgData->resType->value == \"ideal\""

implying that the resistance parameter should be used only if the resistor type is set to ideal.
(== is an equality test, not an assignment.)

When setting the value of a parameter, set

cdfgData->paramName->value = paramValue

If you use this setting, in the future you should be able to use the property list editor to edit
the parameter values directly without going through the form.

cdfgForm

Form on which the CDF data is displayed, if there is one. You can modify data stored in
cdfgData in the CDF form initProc, or by modifying the cdfgForm fields with a callback
function. When doing this, set the Boolean variable cdfgForm -> cdfModified to t.

gLabelsNumNotation

Displays the cdsTerm and cdsParam values in different notations, such as scientific or
engineering. The syntax of gLabelsNumNotation is as follows:
gLabelsNumNotation = ‘suffix

The possible values can be ‘suffix, ‘scientific, ‘engineering, or ‘default.

December 2005 136 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

The default value displays the labels according to the existing setting. For example, the
nmos symbol shows the engineering notation, 300e-3 for w, by default.

To set the number of signif digits, use the aelPushSignifDigits function as follows:
aelPushSignifDigits(10)

December 2005 137 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

CDF SKILL Functions

This section describes the functions that operate on CDF descriptions.

Create Functions
You can create base and user CDF descriptions for libraries and cells using the following
functions.

cdfCreateBaseLibCDF
cdfCreateBaseLibCDF(
libId
[?doneProc t_doneProc]
[?formInitProc t_formInitProc]
[?fieldWidth x_fieldWidth]
[?fieldHeight x_fieldHeight]
[?buttonFieldWidth x_buttonFieldWidth]
[?promptWidth x_promptWidth]
)
=> cdfDataId / nil

Description

Creates the Base Library CDF that is applied to all the devices in the library.

You must open the library in write mode. The base-level CDF description must not already
exist for the library. The CDF description is created with no parameters or simModels.

Arguments

libId This is the library ID.

t_doneProc Lets you specify an optional SKILL routine that executes after
you change any parameter on the instantiation form.
t_formInitProc Lets you specify an optional SKILL language routine that
executes automatically when the component is placed on an
instantiation form.
x_fieldWidth Lets you specify the width of a field on the instantiation form.
The default width is 350 pixels.

December 2005 138 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

x_fieldHeight Lets you specify the height of a field on the instantiation form.
The default height is 35 pixels.
x_buttonFieldWidth Lets you specify the width of a button on the instantiation form.
The default width is 350 pixels.
x_promptWidth Lets you specify the width of the prompt on the instantiation
form. The default width is 175 pixels.

cdfCreateUserLibCDF
cdfCreateUserLibCDF(
libId
[?doneProc t_doneProc]
[?formInitProc t_formInitProc]
[?fieldWidth x_fieldWidth]
[?fieldHeight x_fieldHeight]
[?buttonFieldWidth x_buttonFieldWidth]
[?promptWidth x_promptWidth]
)
=> cdfDataId / nil

Description

Creates the user-level library CDF that is applied to all the devices in the library. The user-
level CDF can override entries in the base-level CDF. Therefore, a combination of the base-
level CDF and the user-level CDF becomes the effective CDF.

The user-level CDF description must not already exist for the library. The CDF description is
created with no parameters or simulation models.

Arguments

libId This is the library ID.

t_doneProc Lets you specify an optional SKILL routine that executes after
you change any parameter on the instantiation form.
t_formInitProc Lets you specify an optional SKILL language routine that
executes automatically when the component is placed on an
instantiation form.
x_fieldWidth Lets you specify the width of a field on the instantiation form.
The default width is 350 pixels.

December 2005 139 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

x_fieldHeight Lets you specify the height of a field on the instantiation form.
The default height is 35 pixels.
x_buttonFieldWidth Lets you specify the width of a button on the instantiation form.
The default width is 350 pixels.
x_promptWidth Lets you specify the width of the prompt on the instantiation
form. The default width is 175 pixels.

cdfCreateBaseCellCDF
cdfCreateBaseCellCDF(
cellId
[?doneProc t_doneProc]
[?formInitProc t_formInitProc]
[?fieldWidth x_fieldWidth]
[?fieldHeight x_fieldHeight]
[?buttonFieldWidth x_buttonFieldWidth]
[?promptWidth x_promptWidth]
)
=> cdfDataId / nil

Description

Creates a base-level CDF description for a cell.

You must open the cell in write mode. The base-level CDF description must not already exist
for the cell. The CDF description is created with no parameters or simulation models.

Arguments

cellId This is the cell ID.

t_doneProc Lets you specify an optional SKILL routine that executes after
you change any parameter on the instantiation form.
t_formInitProc Lets you specify an optional SKILL language routine that
executes automatically when the component is placed on an
instantiation form.
x_fieldWidth Lets you specify the width of a field on the instantiation form.
The default width is 350 pixels.
x_fieldHeight Lets you specify the height of a field on the instantiation form.
The default height is 35 pixels.

December 2005 140 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

x_buttonFieldWidth Lets you specify the width of a button on the instantiation form.
The default width is 350 pixels.
x_promptWidth Lets you specify the width of the prompt on the instantiation
form. The default width is 175 pixels.

cdfCreateUserCellCDF
cdfCreateUserCellCDF(
cellId
[?doneProc t_doneProc]
[?formInitProc t_formInitProc]
[?fieldWidth x_fieldWidth]
[?fieldHeight x_fieldHeight]
[?buttonFieldWidth x_buttonFieldWidth]
[?promptWidth x_promptWidth]
)
=> cdfDataId / nil

Description

Creates a user-level CDF description for a cell.

The user-level CDF description must not already exist for the cell. The CDF description is
created with no parameters or simulation models.

Arguments

cellId This is the cell ID.

t_doneProc Lets you specify an optional SKILL routine that executes after
you change any parameter on the instantiation form.
t_formInitProc Lets you specify an optional SKILL language routine that
executes automatically when the component is placed on an
instantiation form.
x_fieldWidth Lets you specify the width of a field on the instantiation form.
The default width is 350 pixels.
x_fieldHeight Lets you specify the height of a field on the instantiation form.
The default height is 35 pixels.
x_buttonFieldWidth Lets you specify the width of a button on the instantiation form.
The default width is 350 pixels.

December 2005 141 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

x_promptWidth Lets you specify the width of the prompt on the instantiation
form. The default width is 175 pixels.

cdfCreateParam
cdfCreateParam(g_cdfDataId ?name t_name ?type t_type [?defValue g_defValue]
[?units t_units]
[?parseAsNumber t_parseAsNumber]
[?choices l_choices]
[?prompt t_prompt]
[?use t_use]
[?display t_display]
[?editable t_editable]
[?dontSave t_dontSave]
[?callback t_callback]
[?storeDefault t_storeDefault])
[?parseAsCEL t_parseAsCEL]
)
=> cdfDataId / nil

Description

Creates a parameter on the specified cdfDataId with the specified attributes. The only
attributes that are always required are the parameter’s name and type. If this parameter
description is not overriding an existing base-level parameter definition, you must also specify
the defValue. The specified parameter must not already exist on the specified CDF
description.
Note: You cannot override a parameter’s type. Also, CDF checks that the effective parameter
due to any overrides is consistent and valid.

Example
cdfParamId = cdfCreateParam(cdfDataId
?name "resistorType"
?type "radio"
?prompt "Type:"
?defValue "ideal"
?choices list("ideal" "thin film")
?callback "myModelTypeCB()")

For more examples of the cdfCreateParam function, refer to Appendix C, “NBSIM

Transistor CDF SKILL Description”.

December 2005 142 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

Queries
You can query existing CDF descriptions using the following functions. To use the returned
value, you must assign it to a variable that you create.

cdfGetBaseLibCDF
cdfGetBaseLibCDF(d_libId
)
=> cdfDataId / nil

Returns the base-level CDF description attached to a library. If one is not defined, it returns
nil.

cdfGetUserLibCDF
cdfGetUserLibCDF(d_libId
)
=> cdfDataId / nil

Returns the user-level CDF description attached to a library. If one is not defined, it returns
nil.

cdfGetLibCDF
cdfGetLibCDF(d_libId
)
=> cdfDataId / nil

Returns the effective CDF description attached to a library. If neither a base- nor user-level
CDF description is defined, it returns nil. The resulting CDF description represents the
overlay of the user-level CDF on the base-level CDF.

cdfGetBaseCellCDF
cdfGetBaseCellCDF(d_cellId
)
=> cdfDataId / nil

Returns the base-level CDF description attached to a cell. If one is not defined, it returns nil.

December 2005 143 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

cdfGetUserCellCDF
cdfGetUserCellCDF(d_cellId
)
=> cdfDataId / nil

Returns the user-level CDF description attached to a cell. If one is not defined, it returns nil.

cdfGetCellCDF
cdfGetCellCDF(d_cellId
)
=> cdfDataId / nil

Returns the effective CDF description attached to a cell. If neither a base- nor user-level CDF
description is defined for the cell or its library, it returns nil. The resulting CDF description
represents the overlay of the user-level cell CDF on the base-level cell CDF on the user-level
library CDF on the base-level library CDF.

cdfGetInstCDF
cdfGetInstCDF(d_instId
)
=> cdfDataId / nil

Returns the effective CDF description associated with an instance.

The difference between the instance’s effective CDF description and the cell’s effective CDF
description is that the values of any CDF parameter takes into account the values of the
parameters stored on the instance.

December 2005 144 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

Saving Descriptions
You can save CDF descriptions using the cdfSaveCDF function. If this CDF description
already exists, the new description is written over the old one.

cdfSaveCDF
cdfSaveCDF(g_cdfDataId
)
=> t / nil

Saves a CDF description to disk.

The CDF description is then read in every time you open the cell of the library to which the
description is attached. You can save only base-level CDF descriptions. You must have write
permission on the object to which the CDF description is attached to execute this function.

December 2005 145 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

Dumping and Editing Descriptions

You can save a CDF description to a file and edit it using the following functions. These
functions replace adcDump and cdfDumpCellCDF.

Caution
While editing dump files remember to escape invalid characters while
specifying termOrder using symbols. All symbols must be preceded by
the backslash (\) to make them valid symbols in SKILL.

cdfDump
cdfDump(t_libName t_fileName
[?cellName t_cellName]
[?level s_level]
[?edit b_edit]
)
=> t / nil

Dumps the CDF description for t_libName and t_cellName into t_fileName. If
t_cellName is not specified, then only the library CDF description is dumped.
t_fileName is created in the current working directory or the directory specified with the
filename. s_level is either ‘base or ‘user, with ‘base as the default value. If b_edit
is t, a text editor window is automatically opened on t_fileName. The default is no editor.

Example
cdfDump( "nmos" "tr.mod" ?cellName "pnp" ?level ‘base
?edit t)

cdfDumpAll
cdfDumpAll(t_libName t_fileName
[?level s_level]
[?edit b_edit]
)
=> t / nil

Dumps the CDF description for t_libName and all its cells into t_fileName. s_level
is either ‘base or ‘user, with ‘base as the default value. t_fileName is created in the
current working directory or the directory specified with the filename. If b_edit is t, a text
editor window is automatically opened on t_fileName. The default is no editor.

December 2005 146 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

Example
cdfDumpAll("asic" "lib.mod" ?level ’base ?edit t)

December 2005 147 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

Deleting Descriptions
You can delete a CDF description or its parameters using the following functions.

cdfDeleteCDF
cdfDeleteCDF(g_cdfDataId
)
=> t / nil

Deletes a CDF description, including all attached parameters.

If the CDF description has been saved, the saved versions are also deleted. If this is a base-
level CDF description, you must have write permission on the object to which the CDF
description is attached.

cdfDeleteParam
cdfDeleteParam(g_cdfParamId
)
=> t / nil

Deletes a CDF parameter.

If the CDF parameter is attached to a base-level CDF description, you must have write
permission on the object to which the CDF description is attached.

CDF checks that no invalid parameter descriptions would result from deleting the specified
parameter before it actually deletes it. This would occur if you tried to delete a base-level
parameter and a user-level parameter is defined that only partially overrides the base-level
description.

December 2005 148 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

Copying, Finding, and Updating Data and Parameters

You can copy, find, and update CDF data and parameters using the following functions.

cdfCopyCDF
cdfCopyCDF(d_id t_dataType g_dataId
)
=> g_cdfDataId / nil

Copies CDF data onto the specified ID, creating a new dataId of the specified type.

The ID must be either a library or a cell, and it must not already have a CDF description of
the specified type. The data type must be one of the following:
"baseLibData"
"userLibData"
"baseCellData"
"userCellData"

Neither the old nor the new dataId can be effective objects.

Arguments

d_id Specifies the ID of the library or cell

t_dataType Specifies the type of cdf data to be copied
g_cdfDataId Specifies the source of the cdf data to be copied from

Value Returned

cdfDataId New cdf data returned if successful

nil nil returned in case of an error

cdfCopyParam
cdfCopyParam(g_dataId g_paramId
)
=> g_paramId / nil

Copies a parameter, adding it to dataId.

December 2005 149 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

dataId must not already have a parameter with the same name. The parameter and the
dataId must not be effective objects.

cdfFindParamByName
cdfFindParamByName(g_cdfDataId t_name
)
=> g_paramId / nil

Returns the parameter ID for the specified parameter name on the specified CDF description,
if it exists. If not, it returns nil.

Use this function to search for parameters by name.

cdfUpdateInstParam
cdfUpdateInstParam(
d_instId
)
=> t / nil

Stores the CDF parameters specified in the effective cell CDF of the instance master onto the
specified instance. When the Id given is not for an instance or the instance master does not
have CDF definition, it returns nil.

aedCopyCDF
aedCopyCDF(
)
=> t / nil

Opens the Copy Component CDF form.

aedDeleteCDF
aedDeleteCDF(
)
=> t / nil

Opens the Delete Component CDF form.

December 2005 150 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

Setting Scale Factors

Use the following Cadence SKILL language commands to determine the current scale factors
or to set scale factors.

cdfGetUnitScaleFactor
cdfGetUnitScaleFactor(
t_unitName
)
=> t_scaleFactor

Description

Displays the current scale factor for the specified unit.

Arguments

t_unitName The unit name for which you want to display the scale factor.

Value Returned

t_scaleFactor The current scale factor for the specified unit name.

Example

Following command returns the scale factors for power:

cdfGetUnitScaleFactor("power")

December 2005 151 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

cdfSetUnitScaleFactor
cdfSetUnitScaleFactor(
t_unitName
t_scaleFactor
)
=> t / nil

Description

Sets the scale factor for the specified unit.

Arguments

t_unitName The unit name for which you want to set the scale factor.
t_scaleFactor The scale factor for the specified unit name.

Example

Following command sets the lengthMetric to mM:

cdfSetUnitScaleFactor("lengthMetric" "mM")

Other SKILL Functions

cdfParseFloatString
cdfParseFloatString(
t_string
)
=> nil / d_value / t_string

Description

This function uses the standard strtod (string to double) function to parse the input string.
When the input string contains trailing non-numerical characters, the fragment of the string is
compared against a supported set of scale factor designators.

For information on scale factors, refer to Chapter 2, “CDF Commands”.

December 2005 152 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

Value Returned

nil when the input string cannot be parsed as a float value or

without a valid scale factor, as shown below:
cdfParseFloatString("1g") => nil

d_value (a float value) when the input string can be parsed as a float value with or
without a valid scale factor, as shown below:
cdfParseFloatString("1.0") => 1.0
cdfParseFloatString("1.0u") => 1e-06

the given string when the input string does not contain a valid numerical
representation for a float value. For example, the input string
starting with a non-digit character as shown below:
cdfParseFloatString("abcd") => "abcd"

December 2005 153 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

cdfFormatFloatString
cdfFormatFloatString(
t_string
t_scaleFactor
)
=> nil / t_val

Description

This function formats the input string into a value representation, if possible. It formats the
input string using the input scale factor, re-converts the value to a string, and then returns the
formatted string value. If the input string cannot be converted, the input string is returned with
no change to it.

Arguments

t_string a string representing a float value

t_scaleFactor a string representing a scale factor

Value Returned

nil if the t_scaleFactor given is invalid

t_val when the input string can be formatted using the input scale
factor. Else, the input string is returned without any change to
it.

Example
cdfFormatFloatString("123.4" "m") => "123400.0m"
cdfFormatFloatString("10000" "M") => "0.01M"

December 2005 154 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

cdfSyncInstParamValue
cdfSyncInstParamValue(
d_instId1
d_instId2
)
=> t / nil

Description

This function generates all the CDF parameters for the first instance (instId1 ) and updates the
second instance (instId2 ) with the same values. Both the instances must share the same cell.

cdfUpdateInstSingleParam
cdfUpdateInstSingleParam(
d_instId
t_paramName
)
=> t / nil

Description

This function copies the specified parameter’s (t_paramName ) effective value to the specified
instance (d_instId ).

December 2005 155 Product Version 5.1.41

 Component Description Format User Guide
CDF SKILL Summary

Invoking the CDF Editor

You have the option of modifying how the CDF Editor opens. You can use aedEditCDF to
specify initial library, cell, and type values. For example, you can modify the editor so that it
starts with the base level CDF instead of the effective level CDF.

aedEditCDF
aedEditCDF(
[?libName t_libraryName
[?cellName t_cellName]
[?cdfType t_cdfType]
)
=> g_dataId / nil

Description

Opens the CDF Editor to the library, cell, and CDF type specified by libraryName,
cellName, and cdfType.
libraryName and cellName must be strings referring to an existing library or cell, and
cdfType must be ‘effective’, ‘base’, or ‘user’.

December 2005 156 Product Version 5.1.41

 Component Description Format User Guide
Verifying CDF Operation

9
Verifying CDF Operation

■ Testing the Component on page 158

■ Finishing the NFET Example on page 160

December 2005 157 Product Version 5.1.41

 Component Description Format User Guide
Verifying CDF Operation

Testing the Component

Once you have completed the design of your component (all views and the Component
Description Format are done), test the operation of the component’s CDF description before
using the component in your design. The following is a procedure you might use.

Test Procedure
1. Open an empty schematic and place instances of your symbol.
2. As you place the instances, use different parameter values.
Check the Add Component form for confirmation that your parameter definitions and
callbacks are working. Look for the following indications:
❑ Parameters have the units you expected.
❑ Parameters have the defaults you expected.
❑ The correct parameters are at the top of the list of parameters. You should not have
to keep scrolling down the form to edit important parameters.
❑ Parameters that are supposed to become noneditable do so when expected.
❑ Parameters that are not supposed to be displayed are not displayed as expected.
❑ Do your callbacks get called? First check. Do they stacktrace because you forgot to
load them?
❑ Parameters whose values depend on other parameters update as you expect.
❑ Are your callbacks doing the range checking you set up?
❑ Can you enter bad parameter values?
3. After you place the instances, use Edit Properties (in the schematic editor) to look at the
actual properties stored on the instances.
Look for the following:
❑ Are you saving the default values?
❑ Are nondisplayed parameters being set as you expected? If they are not displayed,
it is hard to know that they worked.
4. Draw a minimal schematic and go to the simulator you will be using. Generate a netlist
to make sure that your component netlists as you expect.

December 2005 158 Product Version 5.1.41

 Component Description Format User Guide
Verifying CDF Operation

5. Once you are satisfied with the schematic and simulation results, go to an empty layout
and repeat the CDF tests to see the effect on your layout.

Virtuoso XL Layout Editor Procedure

If you are going to use the Virtuoso® XL Layout Editor (Virtuoso XL),
1. Make a simple schematic with some examples of your component randomly wired to
each other.
2. Go through the Virtuoso XL design cycle.
3. Check that you have set all the parameters that Virtuoso XL requires.
4. Does Virtuoso XL place the layouts you expect with appropriate connectivity?

LVS Procedure
If you are using Layout Versus Schematic (LVS),
1. Create a simple schematic and layout with very few instances.
2. Run extraction and LVS.
3. Make sure the LVS netlist is correct.
This also gives you a chance to check out the LVS rules in a controlled experiment using
just one device.

December 2005 159 Product Version 5.1.41

 Component Description Format User Guide
Verifying CDF Operation

Making Changes
If you find a problem or change the design, follow these guidelines:
■ If problems occur during testing, fix the views or CDF and retest.
■ If the problem or change occurs after you have placed multiple instances of the
component, consider the following:
❑ You might need to change just the user interface, such as the order of the
parameters or which parameters are displayed in the form or label display. When you
make a change, it is inherited by all existing instances.
❑ If you change the default value of a parameter with the attribute storeDefault = no,
it is inherited by all existing instances. (Use the Window – Redraw command to see
the changes reflected in the current window.)
❑ If you change the default value of a parameter with storeDefault = yes, changing
the default value does not update the existing instance, because the value was
saved as a property on each instance.
❑ If you change a callback that sets the value of one parameter based on an equation
involving another parameter, changing the default value does not update the existing
instances. You must also trigger the callback again, for each instance.
In the last two cases, you can update existing instances using the Cadence Analog
Design Environment – Design Variables – Update command. This command
updates the default values and triggers all of the callbacks of all of the instances in the
current window. The Update command has the following limitations:
❑ The update works only on the open cell.
❑ The update does not traverse the design hierarchy.
❑ There is no guaranteed order of running the callbacks, so you need to make sure
that your callbacks work well together.

Finishing the NFET Example

In the last part of the NFET example you can see visible changes in the cell layout and form
displays as a result of changing the CDF data.

December 2005 160 Product Version 5.1.41

 Component Description Format User Guide
Verifying CDF Operation

Varying the Layout

Once you enter all the CDF information for the NFET into the system, you can display the
transistor layout. The pcell layout (stored in view layout) might look like the following:

This is a pcell with the following parameters:

width float, defining the width of the gates

length float, defining the length of the gates
bend Boolean, to control conditional inclusion of shapes for bend
bendt float, to help in stretching shapes correctly when bend occurs
diffD Boolean, to control conditional inclusion of shapes for diffusion
drain contact versus metal drain contact

December 2005 161 Product Version 5.1.41

 Component Description Format User Guide
Verifying CDF Operation

The NFET looks like the following with different parameter combinations:

December 2005 162 Product Version 5.1.41

 Component Description Format User Guide
Verifying CDF Operation

In addition, you need the fixed layout for the power applications. This might be stored under
view layout.pwr, have a gate width of 880 micrometers and length of 12 micrometers, and
look like this:

December 2005 163 Product Version 5.1.41

 Component Description Format User Guide
Verifying CDF Operation

Changing Parameters
Now try to use the NFET in a design.

Add Component

No default
values

When you first use the Add Component form, some fields are dimmed.

The first time you open the Add Component form for your component, it shows all default
values. Some default values are not defined. At this point, you might want to go back to the
CDF editor and add the appropriate default values.

December 2005 164 Product Version 5.1.41

 Component Description Format User Guide
Verifying CDF Operation

Notice, however, that you cannot enter values in the fields with dim names. If you use the CDF
description to set a parameter to be displayed, but not editable, the parameter field is not
editable on the Add Component form.

The type of layout is cyclic, as you specified. The width and length are prompted for only
once, as a string with units of length.

As you place instances of the NFET, you never change the value for the bulk node connection.
Therefore, you might use the CDF editor to move that parameter down in the list of
parameters until it is below other parameters that you do change.
1. Change the Type of Layout cyclic button. The Add Component form updates as follows:

Add Component

Alternate layout

Width and length

are set and are
not editable

Set to appropriate value

You can see the form modify itself to show the new fields, indicating that the callbacks
are working.

December 2005 165 Product Version 5.1.41

 Component Description Format User Guide
Verifying CDF Operation

At this point you still cannot see if the pcell parameters are being set properly.
2. Use the Edit – Properties – Object command on a few examples to verify the operation
of the pcell parameters in the CDF description.
3. Using the Edit Object Properties form, create an instance with width set to 30u, length
set to 10u, the default layout, and a drain diffusion pin.
The CDF Parameter fields in the Edit Object Properties form should look like this:

Edit Object Properties

These values indicate that the callbacks used to generate pcell parameter values
worked correctly. If you have a different result, check the callbacks in the CDF
description.
4. Create an instance of the NFET with width set to 130u, length set to 3u (default), the
default layout, and no drain diffusion pin (required).

December 2005 166 Product Version 5.1.41

 Component Description Format User Guide
Verifying CDF Operation

The Edit Object Properties form should look like the following:

Edit Object Properties

width is > 120,

so bend is turned on

width is > 120,

so no diffD

Both of the length parameters are default values, so no property is saved.

5. Create an instance with maskLayoutViewName as layout.pwr.

December 2005 167 Product Version 5.1.41

 Component Description Format User Guide
Verifying CDF Operation

You see the following:

Edit Object Properties

All the callbacks are triggered, as expected.

December 2005 168 Product Version 5.1.41

 Component Description Format User Guide
Advice and Warnings

A
Advice and Warnings

This section answers frequently asked questions not already covered, and warns you about
common mistakes.
■ Library CDF versus Cell CDF
A cell inherits all definitions from the library CDF. If there are changes in the cell CDF, the
changes override the library CDF at the cell level. It is important to be aware of what is
in the library CDF of any cell you work on.
■ CDF level differences
Remember that only the base-level CDF is saved to disk. The user-level CDF is used
only for one session and cannot be saved. If you have a base and user CDF, the effective
CDF is everything from the base CDF updated with anything new from the user CDF. In
the CDF editor, you can “edit” the effective CDF. In reality, the effective CDF is just there
to give you a way to look at the total CDF for a component. There is nothing to edit in an
effective CDF, so changing the effective CDF in the CDF editor does nothing.
■ Update
The update function under Cadence Analog Design Environment– Design
Variables – Update in the schematic editor updates only what is in the current window
and for only one level of hierarchy.
■ Parameter values and callbacks
If you use iPar, pPar, atPar, dotPar, or the old notation ({}) to inherit the value of a
parameter, as the real value changes, the callbacks associated with that parameter are
NOT triggered. So, be aware of that when determining which parameters inherit values.
Similarly, if you assign a design variable as a parameter value, the setting of the design
variable does not trigger callbacks.
■ Units
The units you can use in the CDF are limited to those offered in the CDF editor. There
are no user-defined units.
■ Range checking

December 2005 169 Product Version 5.1.41

 Component Description Format User Guide
Advice and Warnings

CDF provides no automatic range checking. You must use callbacks to do that.
■ Pcell default values
There is no system mechanism to ensure that your pcell defaults and CDF defaults are
synchronized, so you must manage this.
■ Callback files
If you define your callbacks in a separate file (not self-contained in the CDF or attached
to the library), you must remember to archive that file whenever you archive your library.
■ Callback limitation
You can control parameter values with callbacks, but you cannot affect anything in the
simulation information section.
■ Boolean properties
Be careful of using the CDF “Boolean.” It is defined as t or nil. If you add a property to
an instance using the Add button in the Edit Properties form you will see that the system
models that kind of Boolean as TRUE or FALSE. There are some applications that
depend on that. For example, the property lvsIgnore used in DLE and LVS must be
TRUE or FALSE. So, if you want to add it to your CDF, instead of making it a CDF
Boolean type, make it a cyclic type and set its choices to TRUE and FALSE.
■ Triggering callbacks
Callbacks are triggered whenever a change is made to the parameter value, even if the
change is caused by loading a default value. You need to ensure that default values do
not violate any relationships that you are trying to enforce.
Use callbacks to establish dependencies between parameters whose values are
numeric constants.
Avoid using callbacks on parameters whose values are expressions. Such values require
update evaluations and can quickly become out of date. If you do choose to use
callbacks of this type, you will need to run a hierarchical evaluator that will walk the entire
design to trigger all CDF callbacks whenever the value of the parameter is changed.
Note, too, that a hierarchical evaluator will not be reliable with pPar() parameters.
■ Layout extraction, properties, and callbacks
When you run layout extraction to find devices in the layout, you can also measure the
sizes of the devices. When you save that information, it is stored as a property on the
symbol (auLvs or an ivpcell) that is placed in the extracted layout. For example, when
extraction finds an NFET, it places the cell nfet auLvs at the location of the recognition
shape. That instance has the connectivity that was found and has the properties for the

December 2005 170 Product Version 5.1.41

 Component Description Format User Guide
Advice and Warnings

measured values saved, probably the length and width. Setting those measured
parameters by extraction does not trigger any callbacks on those parameters. This is
especially important to remember if you are doing parasitic resimulation, because you
need all parameters necessary for simulation to be set in the extracted layout. You cannot
rely on the callbacks; you must redo the calculations in the extraction rules.
■ LVS case sensitivity
Remember the property mapping example for LVS. Even though CDF might use
lowercase parameter names, LVS expects capital letters. If you do not enter this mapping
properly, the parameters are not netlisted correctly and in LVS this most likely means that
size comparisons are not made at all.
■ The Virtuoso® layout editor now accesses and uses CDF parameters.
■ Property value replacement and callbacks
Using Edit – Search – Replace to find a property and change its value across a set of
instances does not trigger callbacks associated with the parameter. The Cadence
Analog Design Environment – Design Variables – Update command triggers the
callbacks associated with each variable, but you cannot control the order in which the
callbacks are executed.
In addition, if you use the default value for a parameter, and set storeDefault to no, there
is no property for Edit – Search – Replace to find. However, changing the default value
does change it.
■ Numerical value format
Be careful about entering blanks when entering values in a form or defining values or
default values. In the CDF, “ 12” does not equal 12; nor does
“12 ”. “12 u” does not equal “12u”. This can cause errors.
■ The dontSave parameter attribute
Do not use the dontSave parameter attribute unless you are absolutely sure you need
to. It can lead to very strange results, especially when callbacks are used. The only thing
that dontSave does for you is save the space used to store the property value on a
parameter that you think is not very useful. One example is when you use the parameter
resType on a resistor to trigger callbacks to set up the rest of the parameters based on
the resistor type. Once that is done, you do not think the value of resType needs to be
saved. The problems occur when you edit the parameters on existing instances and the
values do not match.
■ Reserved words
There are words that are reserved and cannot be used as net, component, property, or
parameter names. Do not use any of the following:

December 2005 171 Product Version 5.1.41

 Component Description Format User Guide
Advice and Warnings

❑ Command or function names that your simulator uses

❑ A parameter or property name or prompt that is already in use
❑ The names of fields
❑ “others,” “loadSave,” “simInfo,” “othersFG,” “area,” “resistor,” or “multiplier”
In addition, if you want to make sure a property you define does not go into the netlist,
do not use the name of a CDF parameter.
■ Initial and final procedures
Avoid using too many initial and final procedures (initProc and doneProc). Running one
or two procedures every time you change a parameter can slow down system
performance noticeably.
■ Layout name labels
Labels in layout do not know the hierarchy, so they cannot display hierarchical names,
only current names.
■ Do not use parameter names that begin with a number.
■ Changes in the Edit Component CDF form and its subforms do not take effect until you
click Apply on the Edit Component CDF form.
■ Avoid using the dontUse attribute.

December 2005 172 Product Version 5.1.41

 Component Description Format User Guide
Accessing Subcircuit Simulation Data

B
Accessing Subcircuit Simulation Data

■ Overview on page 174

■ Accessing Subcircuit Port Currents on page 175
■ Accessing Subcircuit Model Parameter and Operating Point Information on page 176
■ Accessing Schematic Primitive Model and Operating Point Information on page 179
■ Editing a CDF Description on page 180
■ Subcircuit Simulation Data for Socket Simulation on page 181

December 2005 173 Product Version 5.1.41

 Component Description Format User Guide
Accessing Subcircuit Simulation Data

Overview
This appendix describes how to access simulation data from primitive subcircuits and
schematics when you use the spectre interface. The same information for socket simulator
interfaces is described in “Subcircuit Simulation Data for Socket Simulation” on page 181. By
modifying the Component Description Format (CDF) description of components in a design,
you can annotate and plot subcircuit port currents, primitive model parameters, and operating
point parameters. You can apply the techniques described here to any simulation primitive,
not just subcircuits. These techniques allow you to annotate derived parameters on the
schematic.

Text-based subcircuits are frequently used to

■ Provide a better model of device behavior
■ Allow quick and easy modification of the netlist representing the subcircuit

The first section describes how to modify a CDF description to obtain subcircuit port currents.
The second section describes how to obtain the derived model parameters and operating
point information. The third section describes how to obtain model parameters and operating
point information from schematics below the level of the component in question. The last
section reviews one method of editing a CDF description.
Note: Unless otherwise specified, the information in this appendix is for the spectre
simulator. Other simulators may require a variation on the syntaxes detailed here.

The subcircuit examples use this simplified operational amplifier model. You can find this in
the file
your_install_dir/tools/dfII/samples/artist/models/spectre/vcvsOpampModels.scs
subckt opamp (plus minus output)
parameters inputCap=500e-15
C1 (plus minus) capacitor c=inputCap
E1 (aout 0 plus minus) vcvs gain=1e6
R3 (aout output) r100m
R4 (plus minus) r10M
ends
model r100m resistor r=100e-3
model r10M resistor r=10e6

The subcircuit is represented at the next higher level in the design hierarchy by a symbol, with
pins INplus, INminus, and OUTPUT. The instance I16 in the cellview vcvsOpampTest
schematic in the library aExamples references the cellview.

December 2005 174 Product Version 5.1.41

 Component Description Format User Guide
Accessing Subcircuit Simulation Data

If you update the CDF, you must generate a new netlist in order for the enhancements
described here to work.

INplus +
OUTPUT
INminus -

symbol

Accessing Subcircuit Port Currents

You can obtain the port current of subcircuits of a component with simple simInfo on the CDF
of that component. For the Spectre® circuit simulator, the use of expressions is not necessary
because this simulator handles the calculation of currents.
Note: Only the port currents may be obtained. No access is provided to internal subcircuit
currents.

The required simInfo is best illustrated by example. To provide access to the currents entering
through the plus, minus, and output ports of the operational amplifier model for spectre, add
the following Cadence® SKILL language lines to the text of the spectre Simulation Information
section of the operational amplifier CDF description, using the symbol pin names INplus,
INminus, and OUTPUT:
termMapping ( nil plus \:1 minus \:2 output \:3)

The termMapping property value is a disembodied property list (DPL), which provides
information to the data access code on how to obtain data for the relevant terminal. This
information is also used by the simulator interface to generate appropriate save statements.
The first term of the termMapping entry should always be nil. Note that spectre requires the
order number of the terminal for subcircuits, not the name. So instead of ":plus", ":1" is
required.

The termMapping property describes that the schematic terminal INplus is mapped to plus,
and INminus is mapped to minus, and OUTPUT is mapped to the simulator name output. The
colon character (:) is the Spectre simulator’s instance-terminal delimiter.

To save currents, use the Save Options form on the simulation window, or use the Outputs
– To Be Saved command. You can select the individual terminals on the I16 instance in the
vcvsOpampTest schematic cellview. During simulation, only the currents you specify are
saved.

December 2005 175 Product Version 5.1.41

 Component Description Format User Guide
Accessing Subcircuit Simulation Data

Accessing Subcircuit Model Parameter and

Operating Point Information
By modifying a device CDF description, you can save additional model parameters or
operating point information using the following functions:
■ opParamExprList for specifying DC operating point information
■ optParamExprList for specifying transient operating point information
■ modelParamExprList for specifying model parameters

After simulation, you can access this information from the calculator or by using the label
display mechanism. This method differs from the one used in the previous example in that
you must fully specify the required functions. This is demonstrated in the following example.

In this example, three additional operating point parameters are resistors: rin (input
resistance), rout (output resistance), and gain. Add the following lines to the spectre
Simulation Information section of the operational amplifier CDF description:
opParamExprList (
("rin" "OP(mappedRoot(\".R4\") \"res\")")
("rout" "OP(mappedRoot(\".R3\") \"res\")")
("gain" "1e6")
)

The opParamExprList property is a list of pairs. The first element of each pair is the name of
the new property, and the second element is the expression that must be evaluated to
compute the new property. OP is the operating parameter function, which takes an instance
name and an associated parameter as arguments. The second argument res is the
resistance parameter for that component. The first argument, mappedRoot(\".R4\"), is
the expression for the name of the instance.

The mappedRoot() function takes the string ( \".R4\" ) and prepends the simulator
instance name of the current instance. The dot character (.) is the spectre subcircuit
delimiter. Since the property value is a string, substrings (such as ".R4") must be escaped
(\".R4\").

The root() function prepends the schematic name (rather than simulator name) of the device
instance in question. The root() function also adds a slash (the schematic delimiter) before
its argument.

The other two functions provided, inst() and mappedInst(), do not take arguments and
return the schematic and simulator instance names, respectively.

December 2005 176 Product Version 5.1.41

 Component Description Format User Guide
Accessing Subcircuit Simulation Data

The following table shows the difference between the four functions.

Function Syntax Value Returned

root(string) strcat(instSchematicName "/" string)
mappedRoot(string) strcat(instSimulatorName string)
inst() instSchematicName
mappedInst() instSimulatorName

The SKILL function strcat is the standard SKILL string concatenate function.

Given an instance of the subcircuit with the name I16 and the simulator instance name of
I16, the expressions in the example would evaluate as follows:
mappedRoot(\".R4\") -> "I16.R4"
root(\"inplus\") -> "I16/inplus"

The other functions would act as follows:

inst() -> "I16"
mappedInst() -> "I16"

Similarly, you can add transient operating point information and model parameters if you
replace opParamExprList with optParamExprList and modelParamExprList respectively.
Also replace OP with OPT and MP as follows:
OP(instance_name parameter)
OPT(instance_name parameter)
MP(instance_name parameter)

Following are examples of optParamExprList and modelParamExprList:

optParamExprList (("rin" "OPT(mappedRoot('.R4') 'res')"))
modelParamExprList (("afin" "MP(mappedRoot(\".R4\") \"af\")")
("afout" "MP(mappedRoot(\".R3\") \"af\")"))

In the same way, you can add parameters to cells that are similar to the ones you find in
analogLib. The following addition to a resistor CDF description adds a conductance
property:
opParamExprList (
("cond" "1.0/OP(inst() \"res\")")
)

You can define your own functions to perform computations. Consider the following example
which adds the xyz parameter:

December 2005 177 Product Version 5.1.41

 Component Description Format User Guide
Accessing Subcircuit Simulation Data

opParamExprList (
("xyz" "userFun(inst())")
)

When the xyz operating point property is accessed, the function userFun() is called with
the schematic name of the instance in question. The userFunc() should return an integer
if the operation is successful. If userFunc() does not return an integer, then the parameter
is not displayed. Some more examples of user defined functions are as follows.
opParamExprList (
("avgcurrent" "averageCurrent(root('PLUS'))")
("pwr" "instPwr(inst())")
)
defun( instPwr ( instName )
(let (i v)
i = OP(instName "i")
v = OP(instName "v")
(when (and i v) i * v )
)
defun( averageCurrent (instTerm)
(let (i)
i = IT( instTerm)
(when i average(i))
)

While using a user a user defined function with opParamExprList ensure that the function
is defined, otherwise it causes an error. This process can aid debugging.

The expression you use to define a given property can be arbitrary, but if the expression
evaluates incorrectly or the current is not saved, the system generates an error. This can be
slow down the simulation if there are many instances of the device on a schematic attempting
to display this property. For example, you might add the following avecurrent property to
a resistor to compute the average transient current through the resistor:
opParamExprList (
("avecurrent" "average(IT(root(\"PLUS\")))")
)

If you did not save the current, then the IT() function returns nil, and the average()
function complains about a nil argument. To prevent this, add logic to the expression as in
this example:
opParamExprList (
("avecurrent" "let(((i IT(root(\"PLUS\"))))
(and i average(i)))")
)

If you did not save the current, this expression prevents the system from flagging errors.

You can use standard SKILL functions like let in these expressions. IT and average are
private, internal SKILL functions that are not documented for any other use.

December 2005 178 Product Version 5.1.41

 Component Description Format User Guide
Accessing Subcircuit Simulation Data

Accessing Schematic Primitive Model and

Operating Point Information
You can apply the methods for accessing model and operating point information in subcircuits
to access that same information in schematics. By modifying the CDF description of a
component, you can access model or operating point information in a schematic one level
below the component and display it at the component’s level.

For example, the instance I19 in the cellview vcvsOpampTest is a schematic of the same
subcircuit. On the next higher level is another schematic, where the sample schematic is
represented by a symbol.

outres=3K
inplus
OP52 output

outplus Descend Edit

inres=14K

Level 1

plus
+ R3
R4 E1
- aout output
minus C1

Level 2

By putting a particular statement in the model or operating point expression list of the CDF
simInfo section of OP52, you can define new parameters such as inres and outres, base
them on primitive values one level down, such as R4 and R3, and display their values on
OP52.

To create the two parameters in the example, include these statements in the simInfo section
of OP52 for each simulator that you are using:
opParamExprList(
("inres" "OP(root(\"R4\") \"res\")")
("outres" "OP(root(\"R3\") \"res\")")
)

In this example, inres and outres are parameter names that you define. They derive their
values from the res parameter on primitives R4 and R3, respectively. Note that for
schematics you use the root function.

December 2005 179 Product Version 5.1.41

 Component Description Format User Guide
Accessing Subcircuit Simulation Data

To display the information produced by these CDF description entries, make sure that you
have created enough labels for your component and that those labels are set to display model
or operating point values. For model data, use modelParamExprList and the MP function.
Note: CDF data is not view specific, so do not reference parameters in views that are not
netlisted because of their position in the view switch list.

To refer to primitive data several design levels below the component where you want the data
displayed, you need to include all the instance names from each level.

c_in =1p 1
I10
2
E157
c = 1p 3
C13

For this example, use the following statement in the CDF description of I10 to display the
capacitance on C13 as the input capacitance of I10.
opParamExprList (
("cin" "OP(root(\"E157/C13\") \"c\"))
)

Editing a CDF Description

Writable Cells
To modify a cell CDF description for which you have write access
1. In the CIW, type
cdfDump( "lib" "full_path/filename" ?cellName "cell" ?edit ’t ?level ’base)

and wait for a window to appear.

2. Edit the file, making the necessary changes.
3. Load the file in the CIW with
load("full_path/filename")

December 2005 180 Product Version 5.1.41

 Component Description Format User Guide
Accessing Subcircuit Simulation Data

Sample CDF
The aExamples library demonstrates several different methods of parameter display.

The following is a sample CDF for several of the schematics in the aExamples library.
Spectre examples are given:
■ opamp (macro model)
cdfId->simInfo->spectre = '( nil
propMapping nil
instParameters nil
otherParameters (model)
netlistProcedure nil
termMapping ( nil IN\plus ":1" IN\minus ":2" OUTPUT ":3")
termOrder ("INplus" "INminus" "OUTPUT")
componentName opamp
opParamExprList (("rin" "OP(mappedRoot(".R4") "res")") ("rout"
"OP(mappedRoot(".R3") "res")") ("gain" "1e6"))
)

■ opamp_sch (1 level display):

cdfId->simInfo->spectre = ’( nil
termOrder ("INplus" "INminus" "OUTPUT")
componentName nil
opParamExprList (("inres" "OP(root("R2") "res")") ("outres"
"OP(root("R3") "res")"))
)

■ opamp_2Level (2 level display):

cdfId->simInfo->spectre = ’( nil
termOrder ("in_plus" "in_minus" "out_again")
componentName "subcircuit"
opParamExprList (("cin" "OP(root("I0/C0") "cap")") ("rin"
"OP(root("I0/R2") "res")"))
)

Subcircuit Simulation Data for Socket Simulation

In this section, a number of important differences for spectreS are explained.
Note: Unless otherwise specified, the information in this section is for the spectreS simulator.
Other simulators may require a variation on the syntaxes detailed here.

December 2005 181 Product Version 5.1.41

 Component Description Format User Guide
Accessing Subcircuit Simulation Data

The subcircuit examples use this simplified operational amplifier model. You can find this
model in the file
your_install_dir/tools/dfII/samples/artist/models/spectreS vcvsOpampMacro.s.
.SUBCKT &1 plus minus output
C1 plus minus 500e-15
E1 aout 0 plus minus 1e6
R3 aout output 100e-3
R4 plus minus 10e6
.ENDS &1

Accessing Subcircuit Port Currents for Socket Simulation

The spectreS simInfo required for port currents is different from the simInfo for spectre. In the
same vcvsOpampMacro example, the required simInfo is
noPortDelimite t
current port
termMapping ( nil
IN\plus "(FUNCTION
mappedRoot(\".c1\")plusmappedRoot(\".r4\"))"
IN\minus "(FUNCTION -root(\"INplus\"))"
OUTPUT "(FUNCTION mappedRoot(\".r3\"))"
)

The noPortDelimiter property indicates that the terminal (port) map should not have the port
delimiter (a netlister property) prepended.

The current property indicates to the view selection code that port currents can be obtained
for this device. The value does not have to be port, but it must not be nil or component.

Sample CDF for spectreS

■ vcvsOpampMacro
cdfId->simInfo->spectreS = '( nil
propMapping nil
modelArguments nil
instParameters nil
otherParameters (macro)
namePrefix "X"
termOrder ("INplus" "INminus" "OUTPUT")
componentName subcircuit
macroArguments nil
netlistProcedure ansSpiceSubcktCall
noPortDelimiter t
current port

December 2005 182 Product Version 5.1.41

 Component Description Format User Guide
Accessing Subcircuit Simulation Data

termMapping (nil OUTPUT "(FUNCTION mappedRoot(".r3"))" IN\plus

"(FUNCTION mappedRoot(".c1")plusmappedRoot(".r4"))" IN "(FUNCTION -
root("INplus"))")
opParamExprList (("rin" "OP(mappedRoot(\"^r4\") \"res\")") ("rout"
"OP(mappedRoot("^r3") "res")") ("gain" "1e6"))
)

■ vcvsOpamp
cdfId->simInfo->spectreS = ’( nil
namePrefix "X"
termOrder ("INplus" "INminus" "OUTPUT")
componentName "subcircuit"
macroArguments nil
netlistProcedure ansSpiceSubcktCall
opParamExprList (("inres" "OP(root("R2") "res")") ("outres"
"OP(root("R3") "res")"))
)

■ vcvs2LevelOpamp
cdfId->simInfo->spectreS = ’( nil
namePrefix "X"
termOrder ("in_minus" "in_plus" "out_again")
componentName "subcircuit"
macroArguments nil
netlistProcedure ansSpiceSubcktCall
opParamExprList (("cin" "OP(root("I0/C0") "cap")") ("rin"
"OP(root("I0/R2") "res")"))
)

December 2005 183 Product Version 5.1.41

 Component Description Format User Guide
Accessing Subcircuit Simulation Data

December 2005 184 Product Version 5.1.41

 Component Description Format User Guide
NBSIM Transistor CDF SKILL Description

C
NBSIM Transistor CDF SKILL Description

Dumping the CDF description of the nbsim cell in the analogLib library produces the
following:
/****************************************************/
LIBRARY = "analogLib"
CELL = "nbsim"
/****************************************************/
let( ( libId cellId cdfId )
unless( cellId = ddGetObj( LIBRARY CELL )
error( "Could not get cell %s." CELL )
)
when( cdfId = cdfGetBaseCellCDF( cellId )
cdfDeleteCDF( cdfId )
)
cdfId = cdfCreateBaseCellCDF( cellId )
;;; Parameters
cdfCreateParam( cdfId
?name "model"
?prompt "Model name"
?defValue ""
?type "string"
?display "artParameterInToolDisplay('model)"
?parseAsCEL "yes"
)
cdfCreateParam( cdfId
?name "bn"
?prompt "Bulk node connection"
?defValue "S"
?type "string"
?display "artParameterInToolDisplay('bn)"
?parseAsCEL "yes"
)
cdfCreateParam( cdfId
?name "m"
?prompt "Multiplier"
?defValue "1"
?type "string"
?display "artParameterInToolDisplay('m)"
?parseAsNumber "yes"
?parseAsCEL "yes"
)
cdfCreateParam( cdfId
?name "w"
?prompt "Width"

December 2005 185 Product Version 5.1.41

 Component Description Format User Guide
NBSIM Transistor CDF SKILL Description

?units "lengthMetric"
?defValue ""
?type "string"
?display "artParameterInToolDisplay('w)"
?parseAsNumber "yes"
?parseAsCEL "yes"
)
cdfCreateParam( cdfId
?name "l"
?prompt "Length"
?units "lengthMetric"
?defValue ""
?type "string"
?display "artParameterInToolDisplay('l)"
?parseAsNumber "yes"
?parseAsCEL "yes"
)
cdfCreateParam( cdfId
?name "ad"
?prompt "Drain diffusion area"
?defValue ""
?type "string"
?display "artParameterInToolDisplay('ad)"
?parseAsNumber "yes"
?parseAsCEL "yes"
)
cdfCreateParam( cdfId
?name "as"
?prompt "Source diffusion area"
?defValue ""
?type "string"
?display "artParameterInToolDisplay('as)"
?parseAsNumber "yes"
?parseAsCEL "yes"
)
cdfCreateParam( cdfId
?name "pd"
?prompt "Drain diffusion periphery"
?units "lengthMetric"
?defValue ""
?type "string"
?display "artParameterInToolDisplay('pd)"
?parseAsNumber "yes"
?parseAsCEL "yes"
)
cdfCreateParam( cdfId
?name "ps"
?prompt "Source diffusion periphery"
?units "lengthMetric"
?defValue ""
?type "string"
?display "artParameterInToolDisplay('ps)"
?parseAsNumber "yes"
?parseAsCEL "yes"
)
cdfCreateParam( cdfId
?name "nrd"
?prompt "Drain diffusion res squares"

December 2005 186 Product Version 5.1.41

 Component Description Format User Guide
NBSIM Transistor CDF SKILL Description

?defValue ""
?type "string"
?display "artParameterInToolDisplay('nrd)"
?parseAsNumber "yes"
?parseAsCEL "yes"
)
cdfCreateParam( cdfId
?name "nrs"
?prompt "Source diffusion res squares"
?defValue ""
?type "string"
?display "artParameterInToolDisplay('nrs)"
?parseAsNumber "yes"
?parseAsCEL "yes"
)
cdfCreateParam( cdfId
?name "ld"
?prompt "Drain diffusion length"
?units "lengthMetric"
?defValue ""
?type "string"
?display "artParameterInToolDisplay('ld)"
?parseAsNumber "yes"
?parseAsCEL "yes"
)
cdfCreateParam( cdfId
?name "ls"
?prompt "Source diffusion length"
?units "lengthMetric"
?defValue ""
?type "string"
?display "artParameterInToolDisplay('ls)"
?parseAsNumber "yes"
?parseAsCEL "yes"
)
cdfCreateParam( cdfId
?name "off"
?prompt "Device initially off"
?type "boolean"
?display "artParameterInToolDisplay('off)"
)
cdfCreateParam( cdfId
?name "Vds"
?prompt "Drain source initial voltage"
?units "voltage"
?defValue ""
?type "string"
?display "artParameterInToolDisplay('Vds)"
?parseAsNumber "yes"
?parseAsCEL "yes"
)
cdfCreateParam( cdfId
?name "Vgs"
?prompt "Gate source initial voltage"
?units "voltage"
?defValue ""
?type "string"
?display "artParameterInToolDisplay('Vgs)"

December 2005 187 Product Version 5.1.41

 Component Description Format User Guide
NBSIM Transistor CDF SKILL Description

?parseAsNumber "yes"
?parseAsCEL "yes"
)
cdfCreateParam( cdfId
?name "Vbs"
?prompt "Bulk source initial voltage"
?units "voltage"
?defValue ""
?type "string"
?display "artParameterInToolDisplay('Vbs)"
?parseAsNumber "yes"
?parseAsCEL "yes"
)
cdfCreateParam( cdfId
?name "trise"
?prompt "Temp rise from ambient"
?defValue ""
?type "string"
?display "artParameterInToolDisplay('trise)"
?parseAsNumber "yes"
?parseAsCEL "yes"
)
cdfCreateParam( cdfId
?name "region"
?prompt "Estimated operating region"
?defValue "triode"
?choices '("off" "triode" "sat" "subth")
?type "cyclic"
?display "artParameterInToolDisplay('region)"
?parseAsCEL "yes"
)
cdfCreateParam( cdfId
?name "rdc"
?prompt "Additional drain resistance"
?units "current"
?defValue ""
?type "string"
?display "artParameterInToolDisplay('rdc)"
?parseAsNumber "yes"
?parseAsCEL "yes"
)
cdfCreateParam( cdfId
?name "rsc"
?prompt "Additional source resistance"
?units "current"
?defValue ""
?type "string"
?display "artParameterInToolDisplay('rsc)"
?parseAsNumber "yes"
?parseAsCEL "yes"
)
cdfCreateParam( cdfId
?name "dtemp"
?prompt "Temperature difference"
?defValue ""
?type "string"
?display "artParameterInToolDisplay('dtemp)"
?parseAsNumber "yes"

December 2005 188 Product Version 5.1.41

 Component Description Format User Guide
NBSIM Transistor CDF SKILL Description

?parseAsCEL "yes"
)
cdfCreateParam( cdfId
?name "geo"
?prompt "Source/drain selector"
?defValue ""
?type "string"
?display "artParameterInToolDisplay('geo)"
?parseAsNumber "yes"
?parseAsCEL "yes"
)
;;; Simulator Information
cdfId->simInfo = list( nil )
cdfId->simInfo->auCdl = '( nil
netlistProcedure ansCdlCompPrim
instParameters (m L W)
componentName nbsim
termOrder (D G S progn(bn))
propMapping (nil L l W w)
namePrefix "M"
modelName "NM"
)
cdfId->simInfo->auLvs = '( nil
propMapping nil
netlistProcedure ansLvsCompPrim
instParameters (m l w)
componentName nbsim
termOrder (D G S progn(bn))
deviceTerminals "D G S B"
permuteRule "(p D S)"
namePrefix "Q"
)
cdfId->simInfo->cdsSpice = '( nil
netlistProcedure ansSpiceDevPrim
instParameters (m w l ad as pd ps nrd nrs ld ls off vds vgs vbs)
otherParameters (model bn)
componentName bsim
termOrder (D G S progn(bn))
propMapping (nil vds Vds vgs Vgs vbs Vbs)
namePrefix "S"
current port
dcSens t
acSens t
)
cdfId->simInfo->hspiceS = '( nil
netlistProcedure ansSpiceDevPrim
instParameters (l w ad as pd ps nrd nrs rdc rsc off Vds Vgs Vbs dtemp
geo m)
otherParameters (bn model)
componentName hnmos
termOrder (D G S progn(bn))
propMapping (nil vds Vds vgs Vgs vbs Vbs)
namePrefix "M"
current port
dcSens t
acSens t
)

December 2005 189 Product Version 5.1.41

 Component Description Format User Guide
NBSIM Transistor CDF SKILL Description

cdfId->simInfo->spectre = '( nil

propMapping nil
otherParameters (bn model)
instParameters (w l as ad ps pd nrd nrs ld ls m trise
region)
termOrder (D G S progn(bn))
termMapping (nil D d G g S s B b)
)
cdfId->simInfo->spectreS = '( nil
propMapping nil
netlistProcedure ansSpectreSDevPrim
otherParameters (bn model region)
instParameters (w l as ad ps pd nrd nrs ld ls m trise)
termOrder (D G S progn(bn))
termMapping (nil D d G g S s B b)
namePrefix "S"
componentName bsim1
current port
)
;;; Properties
cdfId->formInitProc = ""
cdfId->doneProc = ""
cdfId->buttonFieldWidth = 340
cdfId->fieldHeight = 35
cdfId->fieldWidth = 350
cdfId->promptWidth = 175
cdfId->modelLabelSet = "vfb phi eta"
cdfId->opPointLabelSet = "id vgs vds"
cdfId->paramLabelSet = "-model l w"
cdfSaveCDF( cdfId )
)

December 2005 190 Product Version 5.1.41

 Component Description Format User Guide
Glossary

Glossary

AEL
Analog Expression Language. A syntax for expressions that allows the rapid evaluation
of those expressions. AEL lets you express values as mathematical expressions instead
of single, fixed values. Refer to the Analog Expression Language Reference
Manual.

auCdl
Analog and Microwave CDL. The analog and microwave version of CDL, which originally
only ran on digital designs.

auLvs
Analog and Microwave LVS. The analog and microwave version of LVS, which originally
only ran on digital designs.

base level
The fundamental representation of a cell or library CDF description. Base level and user
level are combined to produce the effective level, with user-level values overriding base-
level values. Only the base level is permanently stored to the system disk.

callback procedure
A procedure triggered by a change in a parameter in a CDF description. Callbacks can
be SKILL procedures, expressions, or functions entered directly into the Edit Component
Parameter form, or they can be calls to procedures located elsewhere.

CDF
Component Description Format. A system for dynamically changing and storing
parameters and displaying information for components and sets of components for
different versions (levels) of designs.

December 2005 191 Product Version 5.1.41

 Component Description Format User Guide
Glossary

CDL
Circuit Description Language. CDL is a subset of SPICE used by Dracula®, with
extensions that allow for such things as global signals and empty cells.

CEL
CDF Expression Language. Another name for AEL.

cell
The Cadence software representation of a design element (or component).

CIW
Command Interpreter Window. The primary user interface for launching Cadence
software. The CDF editor starts from the Tools menu in the CIW. You can enter SKILL
commands in the CIW command line.

component
A fundamental unit within a system that encapsulates behavior and/or structure (also
known as an element). A cell, with cellviews and associated CDF.

cyclic field
A type of form field that displays only one value, but reveals a sliding menu when you
select the field.

disembodied property list

A software record used to store and retrieve component properties. New properties can
be dynamically added or removed. A disembodied property list starts with a SKILL data
object, usually nil, followed by alternating name and value pairs.

effective level
A masked sum of all the CDF values for a cell or library.

form field
The empty rectangle on a form where you type in values and names.

December 2005 192 Product Version 5.1.41

 Component Description Format User Guide
Glossary

hpmns
MNS™ microwave design simulator from Hewlett-Packard.

interpreted labels
Displayed component values which change with each instance or change as the
component parameters change.

instance parameters
Those parameters used in a module description that change with each instantiation of
that module. These parameters are defined in the first section of a module, the module
interface declarations, and are specified each time a module is called in a netlist instance
statement.

libra
Libra®. A microwave design simulator from EESof.

LVS
Layout Versus Schematic. A Dracula verification tool that confirms that your layout
matches your original schematic.

mharm
Microwave Harmonica®. A microwave design simulator from Compact Software.

model
A named instance with a unique group of parameters specifying the behavior of one
particular version of a component. You can use models to instantiate components with
parametric specifications different than those in the original component definition.

model parameter
Those parameters in a module description that remain the same with each instantiation
of the original module, but which are designed to be changed when the module is called
by a model. These parameters are defined in the second section of a module, the global

December 2005 193 Product Version 5.1.41

 Component Description Format User Guide
Glossary

module scope declarations, and they are specified each time a module is called in a
model instance statement.

nil
No value (an absence of value) or a false value.

parameter
A characteristic of a component.

parameter attributes
A characteristic of a parameter.

parse
The system evaluation of CDF information. Values can be parsed in absolute terms or
according to a predefined system.

pcell
A parameterized cell. You can change a set of cells by changing one parameter.

prompt
The title or words to the left of the form field or button that tell you the name or nature of
the parameter or other value that you enter at that place in the form.

property
A piece of information about an instance or object in a design.

radio button
An entry on a form that behaves as a push button does. You click on the button to turn
the function on (the symbol turns black) and off (the symbol turns white).

radio field
A set of interlinked radio buttons in a single form field associated with one parameter.
Typically, only one button can be on at any one time.

December 2005 194 Product Version 5.1.41

 Component Description Format User Guide
Glossary

SKILL
A proprietary Cadence programming language based on over 1,200 functions. Each
function follows a standard SKILL syntax and acts on a particular data type or data
object.

SKILL command
When SKILL functions or expressions are used alone or entered into the CIW command
line, they are often called SKILL commands.

SKILL expression
An equation where a data object is set equal to a SKILL function. AEL expressions can
be used to express individual values in a SKILL function.

SKILL function
The fundamental element in the SKILL language. An operation to be performed (usually
specific to Cadence software), usually followed by the items or data objects that the
operation is to be performed on.

SKILL procedure
A short program or “script” or routine made up of multiple SKILL expressions and
functions.

SKILL routine
A program made up of multiple SKILL expressions and functions.

T
A non-nil or true value. Used in Boolean values.

user level
A temporary level of cell or library CDF information that overrides base-level information,
but is lost at the end of each design session unless you take specific actions to save it.

December 2005 195 Product Version 5.1.41

 Component Description Format User Guide
Glossary

VLE
Virtuoso Layout Editor. Cadence’s integrated circuit layout editor that supports IC cell
layout design down to the polygon level.

watscad
WATSCAD™. A switched capacitor design simulator from the University of Waterloo.

December 2005 196 Product Version 5.1.41

 Component Description Format User Guide

Index
Symbols B
, . . . in syntax 10 base-level CDF information 26, 169, 191
. . . in syntax 10 Boolean parameter 74
[ ] in syntax 10 Boolean type 170
brackets in syntax 10
Browse button 37
A bulk nodes 82, 83
buttonFieldWidth field 109
adcDump function 45
Add CDF Parameter form 31, 70, 71
Add After Parameter field 54, 70 C
callback field 70, 71
paramType field 70 Cadence SPICE 175
prompt field 70 callback procedures. See callbacks
Add Component form 67, 164 callbacks 65, 72, 132
adding parameters 31, 53, 70, 177 definition files 170
adjusting inherited parameters 65
button field width 109 triggering 169, 170, 171
field height 110 use 118
field width 111 capacitance (scale factor), setting 39
prompt field width 112 CDF 14, 191
aedCopyCDF function 150 description levels 26
aedDeleteCDF function 150 descriptions
aedEditCDF function 156 attaching to components 17
AEL 9, 63, 65, 134, 191 attaching to libraries 17
algebraic expressions 63 base-level 17
Analog and Microwave CDL 191 overriding 17
Analog and Microwave LVS 191 user-level 18
archiving callback procedures 117 relationship to Cadence database 19
artParameterInToolDisplay function 73, SKILL functions 126 to 150
185 CDF Editor 25 to 34
atPar function 63, 64, 169 CDF Expression Language. See CEL
attaching callback procedures 117 CDF types 26
attaching files to libraries 118 cdfCopyCDF function 50, 149
attribute lists 56 cdfCopyParam function 149
attributes 54 cdfCreateBaseCellCDF function 87, 140
auCdl language 191 cdfCreateBaseLibCDF function 138
auLvs simulator 191 cdfCreateParam function 62, 116, 142,
automatic range checking 170 185
automatic symbol generation 99, 100 cdfCreateUserCellCDF function 128, 141
average function 178 cdfCreateUserLibCDF function 139
cdfDataId objects 127
cdfDeleteCDF function 148, 185
cdfDeleteParam function 148
cdfDump function 43, 146

December 2005 197 Product Version 5.1.41

 Component Description Format User Guide

?edit option 43 property values 171

?level argument 43 simulation information 32
cdfDumpAll function 146 simulator attributes 79
cdfEditScaleFactors function 43 simulator options 86
cdfFindParamByName function 150 Circuit Description Language. See CDL
cdfFormatFloatString function 154 CIW 24, 192
cdfgData variable 136 CDF commands 24
cdfGetBaseCellCDF function 128, 143, Tools Menu 24
185 CDF Menu
cdfGetBaseLibCDF function 143 Copy 35
cdfGetCellCDF function 144 Delete 37
cdfGetInstCDF function 144 Edit 25
cdfGetLibCDF function 143 complex pole example 20, 45, 88, 103
cdfGetUnitScaleFactor function 151 complex pole model file 90
cdfGetUserCellCDF function 144 complexPole1 cell 45
cdfGetUserLibCDF function 143 Component Browser 27
cdfgForm variable 136 Component Description Format. See CDF
cdfId variable 87 component parameters 28, 66
cdfParamId objects 129 Component Parameters section 53, 70
cdfParseFloatString function 152 components 192
cdfSaveCDF function 145, 190 conductance (scale factor), setting 39
cdfSetUnitScaleFactor function 152 conductance property 177
cdfSyncInstParamValue function 155 controlling parameter values with
cdfUpdateInstParam function 150 callbacks 170
cdfUpdateInstSingleParam function 155 conventions
CDL 192 menu commands 9
cdsName labels 100 converting to floating-point numbers 57
cdsParam labels 100 Copy CDF command 35
cdsTerm labels 100 Copy Component CDF form 35, 47
CEL 133, 192 CDF Source Information section 48
cells CDF Type field 35, 36
CDF descriptions 26 Cell Name field 36
definition 192 Library Name field 35
IDs 128 copying
label information 97 CDF descriptions 35, 47
macros 81 cell CDF data 36
models 81 cells 44
name selection 26 data and parameters 149
parameterized 65 simulator options 86
change directory 30 creating
changing a CDF description for a new cell 44
button field width 109 button fields 109
default values 31 CDF descriptions 25
field height 110 CDF descriptions with SKILL 43
field width 111 labels with SKILL functions 102
parameter attributes 55 parameters 60
parameter attributes with SKILL 62 parameters with SKILL 62
parameter names 55 the variable for the object ID of a
parameter sequences 55 cell 127
parameters 164 current (scale factor), setting 41
prompt field width 112 current parameter values 117

December 2005 198 Product Version 5.1.41

 Component Description Format User Guide

current property 182 dontSave field 58, 60

cyclic fields 58, 192 editable field 59, 60, 72
name field 58, 59, 67
paramType field 56, 59, 67
D parseAsCEL field 57, 59, 67
parseAsNumber field 56, 59, 67
data objects prompt field 58, 60
defined 127 storeDefault field 57, 59
id 128 units field 57, 59
parameters 128 use field 58, 59, 60
type 128 Edit Component CDF form 25 to 34, 44,
data types 56 46, 49
ddGetObj function 50, 127 Change Directory 44
default parameter values 58, 117 Component Parameters section 31, 53,
defValue attribute 131 70
Delete CDF command 37 Add button 53
Delete Component CDF form 37 Edit button 55
CDF Type to Delete field 37 Move button 55
Cell Name field 37 Name button 53
Library Name field 37 File Name field 44
deleting File Name Select field 44
CDF descriptions 37, 148 header 46
parameters 31, 53 Interpreted Labels Information
derived model information, accessing 176, section 33, 94, 103
179 invoking 156
destination CDF 36 Load 45
destination cell name 36 Name field 31
destination information 36 Other Information section 34, 108
destination library name 36 parseAsCEL field 57
determining design requirements 19 Prompt field 31
disembodied property list 175, 192 Select To Delete button 31, 53
display attribute 72 Simulation Information section 32, 78,
displaying parameter names 53 88
displaying parameters 58, 134 Edit button 78
displaying prompt names 53 Edit Component Display Options form 98,
dmFindCell function 127 100
dmFindLib function 127 Apply To field 100
doneProc procedure 129 Edit Object Properties form 67, 166
dontSave attribute 133, 171 Edit Properties form 170
dotPar function 63, 64 Edit Simulation Information form 32
componentName field 81
current field 84
E deviceTerminals field 83
instParameters field 81
Edit CDF command 25 macroArguments field 81
Edit CDF Parameter form 31, 56 to 67 modelArguments field 81
callback field 59, 60, 72 namePrefix field 84
choices field 58, 60 netlistProcedure field 79
Choose Parameter field 67 otherParameters field 80
defValue field 58, 60 propMapping field 83
display field 58, 60, 72 termMapping field 83

December 2005 199 Product Version 5.1.41

 Component Description Format User Guide

termOrder field 82, 89

editable attribute 132
H
editing interpreted labels information 33 header 46
editing parameter attributes 31, 55, 60 hpmns simulator 193
editing parameters 164
editing simulation information 32
effective-level CDF information 26, 192
entering blanks when entering values 171
I
evaluation ilLabels labels 99
full 95 inductance (scale factor), setting 40
global 95 inheritance evaluation 95
inheritance 95 inherited parameters 63, 65, 95
literal 95 in callbacks 65
to floating-point numbers 56 value functions 63
values 169
initialization procedures 112, 113
F inst function 176, 177
instance labels 100
field height 110 instance parameters 63, 193
field width 109, 111 instDisplayMode attribute 96, 102, 104,
file name selection 29, 30 105
form fields 192 instNameType attribute 96, 102
formInitProc procedure 129 internal subcircuit currents, accessing 175
forms interpreted labels 28, 193
Add CDF Parameter 31 Interpreted Labels Information section 94
Copy Component CDF 35 instDisplayMode field 96
Delete Component CDF 37 cellName option 96
Edit CDF Parameter 31, 56 to 67 instName option 96
Edit Component CDF 25 to 34 instNameType field 96
Edit Component Display Options 98 modelLabelSet field 97, 98
Edit Simulation Information 32 model option 98
Move CDF Parameter 31 netNameType field 96
Units Scaling Factors 38 opPointLabelSet field 97
frequency (scale factor), setting 41 opPoint option 97
full numeric evaluation 95 paramDisplayMode field 94
functional block library 45, 90 paramEvaluate field 95
paramLabelSet field 97
paramSimType field 95
G termDisplayMode field 95
current option 95
gLabelsNumNotation variable 136 netName option 95
global design values 95 pinName option 95
global evaluation 95 voltage option 95
global variables termSimType field 96
cdfgData 136 invoking the CDF editor 156
cdfgForm 136 iPar function 63, 64, 169
defined 136 IT function 178
gLabelsNumNotation 136 italics in syntax 10

December 2005 200 Product Version 5.1.41

 Component Description Format User Guide

K N
keywords 10, 171 net sequence 82
netlist procedures 90
netNameType attribute 96, 102
L netSet 131
NFET example 20, 46, 68, 90, 104, 118,
layout extraction 170 160
length (scale factor), setting 40 nil value 170, 194
let function 178 non-CDF libraries 19
libra simulator 193 noPortDelimiter property 182
library CDF descriptions 26 numeric evaluation 95
library CDF versus cell CDF 169
library IDs 128
Library Manager 47 O
library name selection 26
library-level label information 94 OP function 176
listFunctions function 44 operating point parameters 174 to 179
literal characters 10 operational amplifier
literal evaluation 95 CDF description 176
literal parameter values 95 model 174, 175, 182
load function 44, 45 opParamExprList 176
loading opParamExprList property 176
callback procedures 117 user defined function 177
CDF descriptions from a file 30 opPointLabelSet attribute 97, 102
functions 117 optional parameter attributes 59
LVS tool 193 optParamExprList 176
lvsIgnore property 170 other CDF information 28
Other Information section
buttonFieldWidth field 108
M doneProc field 108, 113
fieldHeight field 108, 110
macro models 89, 90 fieldWidth field 108, 111
macros 67, 90 formInitProc field 108, 112
mappedInst function 176, 177 promptWidth field 108, 112
mappedRoot function 176, 177 overlay 18
maskLayoutViewName parameter 69, 70,
118
menu commands, selection conventions 9 P
mharm simulator 193
model parameters 22, 176, 193 paramDisplayMode attribute 94, 102, 104
accessing 174 parameterized cells 9, 20, 65, 194
modelLabelSet attribute 97, 98, 103 parameterizing the bulk node 82
modelParamExprList 176, 177 parameters 52, 158, 194
models 90, 193 adding 31, 53, 70
mosLayout callback procedure 118, 119 attributes 31, 52, 54, 194
Move CDF Parameter form 31 Boolean 74
moving parameters 31 changing names 55
MP function 177 choices 131
component 66

December 2005 201 Product Version 5.1.41

 Component Description Format User Guide

creating 60 programmable nodes 82, 83

creating with SKILL 62 prompts
current values 117 defined 194
data types 56 width 112
default values 58, 117 properties 52, 194
deleting 31, 53 on symbols 170
determining 15 property mapping 83
displaying 58, 129 LVS 171
editable 132
in callbacks 65
inherited 63, 65 R
instance 63
labels 100 radio buttons 58, 194
lists 53 redisplayed form 29, 30
moving 31 removing CDF data 37
name display 97 reordering parameters 31, 56
name labels 100 required parameter attributes 59
name preceded by minus 97 reserved words 171
names 31, 58, 172 resistance (scale factor), setting 39
parent 63 root function 176, 177
passing 63
pcell 65
positioning 54 S
prompts 58, 131
reordering 31, 56 saving
sequence 55 CDF descriptions to a file 30
setting values 117 currents 84
types 70, 130 user-level CDF information 44
units 131 scale factors 38, 57
use 130 selecting
value 131 CDF data 25
visibility 101 cell names 35
paramEvaluate attribute 95, 102 file names 30
paramLabelSet attribute 97, 102, 104 library names 35
paramSimType attribute 95, 102 simulator options 91
parent parameter 63 size of input field 108
parse 194 source information 35
parseAsCEL attribute 133 sequence of parameters 55
parseAsNumber attribute 133 setting
passing parameters 63 label CDF specifications
pcell defaults and CDF defaults 170 at the cell level 102
pcell parameters 65, 74 at the library level 102
pcell stretch parameters 65 labels
pcells 9, 20, 65, 194 with SKILL 102
pcMOSl callback procedure 118, 122 with the schematic editor 98
pcMOSw callback procedure 118, 120 measured parameters by
positioning parameters 54, 70 extraction 171
postprocessing procedures 113 parameter values 117
power (scale factor), setting 41 sample callbacks 71
pPar function 63, 64, 169 scale factors 38
preprocessing procedures 112 simulator options 86

December 2005 202 Product Version 5.1.41

 Component Description Format User Guide

setting up dmFindLib 127

fields 109 listFunctions 44
label displays 101 load 44, 45
parameters and their attributes 68 skillDev software 44
prompts 109 specifying
simInfo parameters 90 a text editor 43
simulation data, accessing 174 closing procedures 108
simulation information 28 initialization procedures 108
simulator attributes 79 resistivity 14
simulator options 85 storing default values 57, 133
SKILL strcat function 177
CDF functions 43 subcircuit port currents, accessing 174,
commands 195 175
database objects 127, 129 subcircuits 67, 90, 174
defined 195 symbols 103
expressions 195 syntax conventions 10
objects 127, 129
procedures 195
queries 143 T
resistor example 16
routines 195 t value 170, 195
SKILL functions termDisplayMode attribute 95, 102, 105
adcDump 45 terminal labels 100
aedCopyCDF 150 terminals 32, 78
aedDeleteCDF 150 termMapping property 175
aedEditCDF 156 termName attribute 104
artParameterInToolDisplay 73, 185 netName option 104
cdfCopyCDF 50, 149 termSimType attribute 96, 102
cdfCopyParam 149 testing component CDF 158
cdfCreateBaseCellCDF 87, 140 time (scale factor), setting 40
cdfCreateBaseLibCDF 138 Tool Filter 72
cdfCreateParam 62, 116, 142, 185 triggering callbacks 169, 170, 171
cdfCreateUserCellCDF 128, 141
cdfCreateUserLibCDF 139
cdfDeleteCDF 148, 185 U
cdfDeleteParam 148
cdfDump 43, 146 unit suffix 57
cdfDumpAll 146 units 169
cdfEditScaleFactors 43 Units Scaling Factors form 38
cdfFindParamByName 150 unused parameter attributes 59
cdfGetBaseCellCDF 128, 143, 185 update function 169
cdfGetBaseLibCDF 143 use attribute 130
cdfGetCellCDF 144 user-level CDF information 26, 44, 169,
cdfGetInstCDF 144 195
cdfGetLibCDF 143
cdfGetUserCellCDF 144
cdfGetUserLibCDF 143 V
cdfSaveCDF 145, 190
cdfSetUnitScaleFactor 152 viewing CDF descriptions with SKILL 43
ddGetObj 50, 127 Virtuoso vs. Artist 171
dmFindCell 127 VLE 9, 196

December 2005 203 Product Version 5.1.41

 Component Description Format User Guide

voltage (scale factor), setting 41

W
watscad (WATSCAD) simulator 196
write privileges 26

December 2005 204 Product Version 5.1.41