OTA-C VOLTAGE-MODE PROPORTIONAL- INTEGRAL- DERIVATIVE

(PID) CONTROLLER FOR DC MOTOR SPEED CONTROL

1
MARY ANN GEORGE, Research Scholar and 2‡DATTAGURU V. KAMATH, Professor
1,2Department of Electronics and Communication Engineering, Manipal Institute of Technology, Manipal Academy of
Higher Education (MAHE), Manipal, India
E-mail: 1maryann.george@learner.manipal.edu, 2dv.kamath@manipal.edu
‡Corresponding Author

Abstract- The paper discusses voltage mode analog circuit implementation of PID controller for speed control of DC motor
using Operational Transconductance Amplifiers (OTAs). OTA-C approach is used to realize the controller due to its
features like electronic tunability of its transconductance and greater bandwidth. The controller parameters are computed
in MATLAB using auto-tuning method. The controller with the system is simulated in Cadence environment with 180 nm
GPDK technology. Time domain analysis and the performance of compensated system are presented.

Keywords- PID controller, Speed control of DC Motor, Operational Transconductance Amplifier, Tuning Methods

controller. The Section VI presents in detail the circuit

I. INTRODUCTION1 realization and simulation results of the plant without and
DC motors have been extensively used in both industrial and with controller. The concluding comments are given in
commercial applications like robotics, mechatronic systems, Section VII.
electric vehicle, disk drive, etc. [1], hence its speed control is II. DC MOTOR SPEED CONTROL: SYSTEM
very important. The proportional-integral-derivative (PID) MODELING
controllers are popular and widely used in industrial control
problems due to its simplicity. The PID controllers are DC motor is one of the common actuators in the control
popularly used in DC motor speed control applications in system that converts the electrical energy to rotational
industries due to its simplicity and flexibility in tuning and mechanical energy and is widely used in industrial
implementation. The PID controllers are widely used in applications [11]. The electrical equivalent circuit of DC
applications like electric vehicles [2], attitude stabilization of motor showing an armature circuit and free-body diagram of
an unmanned aerial vehicle quad-copter [3], water level the rotor [12] is given in Fig. 1. The armature voltage V and
controlling [4], mobile robot path tracking [5], speed control rotational speed of the shaft 𝜃̇ are the input and output of DC
of DC motor [6]-[7] etc. The tuning of PID controller can be motor speed controller respectively. The DC motor
based on classical tuning methods like dominant pole design, parameters and specifications are given in Table I.
self-tuning and auto-tuning [8]. Auto-tuning method is
widely used in industrial applications in view of its
advantages like no risk of loop instability and loop output
can be maintained close to the set-point all the time with
proper choice of parameters.
The earlier PID controllers were implemented using
analog blocks like Opamps [9]. Recently, Operational
Transconductance Amplifiers (OTAs) have replaced
Opamps in realizing analog controllers due to their merits
like electronic tunability of transconductance, wider
frequency range and resistor-less realization [10]. OTA-C
based PID controllers are realized using only OTAs and Fig. 1 Electrical equivalent circuit of DC motor
capacitors.
The paper deals with the feedback control of DC motor The motor torque T is proportional to the armature current i
speed using a PID controller and circuit realization of the DC and back emf e is proportional to the angular velocity 𝜃̇ .
motor control system using OTA-C approach. This paper has 𝑇 = 𝐾𝑡 𝑖 (1)
the following sections. Section II explains the DC motor
system modeling, Section III discusses the PID controller 𝑒 = 𝐾𝑒 𝜃̇ (2)
auto-tuning method, Section IV describes the CMOS circuit Also
of balanced OTA, Section V illustrates OTA-C based PID 𝐾𝑒 = 𝐾𝑡 = 𝐾 (3)

1
By using Newton’s second law and Kirchhoff’s voltage law
in Fig. 1, the following equations are obtained

𝐽𝜃̈ + 𝑏𝜃̇ = 𝐾𝑖 (4)

𝑑𝑖 (5)
𝐿 + 𝑅𝑖 = 𝑉 − 𝐾𝜃̇
𝑑𝑡
Applying Laplace transform to (4) and (5) and further
simplification gives the below given transfer function of DC Fig. 2 Block diagram of PID controller incorporating auto-tuning
motor speed controller.
(6) PID tuner uses frequency response estimation to compute the
̇
𝜃(𝑠) (𝐾 ⁄𝐽𝐿) 𝑟𝑎𝑑/𝑠𝑒𝑐 PID gains. This method maintains the performance and
= 2 + 𝑏𝑅)
𝑉(𝑠) 𝑅 𝑏 (𝐾 𝑉 robustness for any given control bandwidth and minimum
𝑠2 + 𝑠 ( + ) +
𝐿 𝐽 𝐽𝐿 phase margin. The PID controller for DC motor speed
control is designed for a settling time smaller than 2 sec,
Table I DC motor parameters and specifications
overshoot lower than 5% and zero steady state error. The PID
Parameters Specifications gain values satisfying the above mentioned criteria is
obtained, i.e. Kp = 20.27, Ki = 49.41 and Kd = 1.805. By
J Moment of inertia of the rotor 0.01 kg. m2 substituting the gain values in (8), the transfer function of the
b Motor viscous friction constant 0.1 N. m. s PID controller for DC motor speed control is determined as
Ke Electromotive force constant 0.01 V/rad/s 49.41 (8)
𝐺𝑐 (𝑠) = 20.27 + + 1.805𝑠
𝑠
Kt Motor torque constant 0.01 N.m/A
R Electric resistance 1 ohm IV. OPERATIONAL TRANSCONDUCTANCE
AMPLIFIER (OTA)
L Electric inductance 0.5 H
An ideal OTA is a differential-input current-output analog
Substituting the parameter specifications of Table I in DC block modelled as voltage controlled current source (VCCS)
motor transfer function of (6), the system/plant transfer having high input and output impedance. The electronic
function is obtained as tunability of its transconductance by altering the DC bias
current, capability to operate at greater frequency (200 kHz
𝜃̇(𝑠) 1.96078 𝑟𝑎𝑑/𝑠𝑒𝑐 (7) – 500 MHz) and lesser chip area are the important features
= of OTA [14]. OTA based circuits employ only OTAs and
𝑉(𝑠) 𝑠 2 + 11.7843𝑠 + 19.627 𝑉
capacitors and are suited for complete integrated and
III. PID CONTROLLER: TUNING resistor-less implementation. Hence, in recent years, most
analog circuits are designed using OTAs and capacitors. The
PID controller is extensively employed in industrial process symbols of single output OTA (SO-OTA), dual output OTA
control applications as a result of its simplicity, optimal (DO-OTA) and multiple output OTA (MO-OTA) are given
performance and robustness [13]. PID controller is a in Fig. 3
dominant technique of feedback control and almost 95% of
the industrial controller are PID type. A suitable set of
proportional, integral and derivative gains can provide a
desired closed loop system response. The transfer function
of the PID controller is given in (8). PID controller can
maintain a reasonable transient response by eliminating the Fig. 3 Circuit symbol of SO-OTA, DO-OTA and MO-OTA
steady-state error and reducing the settling time.
The output current of the DO-OTA is given by
𝐾𝑖 (8)
𝐺𝑐 (𝑠) = 𝐾𝑝 + + 𝐾𝑑 𝑠 𝐼0 + = 𝐼0 − = 𝑔𝑚 (𝑉𝑖 + − 𝑉𝑖 − ) (9)
𝑠
where Kp, Ki and Kd are proportional gain, integral gain and where 𝐼0 + , 𝐼0 − are dual outputs and 𝑔𝑚 is the
derivative gain respectively. tranconductance of the DO-OTA, 𝑉𝑖 + and 𝑉𝑖 − indicates non-
The control parameters are calculated using auto-tuning inverting and inverting input voltages of the DO-OTA
method. Fig. 2 illustrates the schematic diagram of the PID respectively. CMOS technologies are suitable for realizing
controller using the auto-tuning technique. OTAs as their MOSFETS are essentially VCCS. The CMOS
circuit of balanced OTA [15] is shown in Fig. 4.
 𝐿1 = 𝐶1 /𝑔𝑚2 𝑔𝑚3 (10c)
The expression for Kp, Ki and Kd are given as
𝐾𝑝 = 𝑔𝑚5 ⁄𝑔𝑚8 (11a)
𝑔𝑚4 𝑔𝑚7 (11b)
𝐾𝑖 = ( )
𝐶2 𝑔𝑚8
𝑔𝑚1 𝐶1 𝑔𝑚6 (11c)
𝐾𝑑 = ( )
𝑔𝑚2 𝑔𝑚3 𝑔𝑚8

Sensitivity: The sensitivity of the proposed PID controller

gains in regard to active devices and passive components can
Fig. 4 CMOS circuit of Balanced OTA [15] be calculated to check the extent to which changes in values
of these parameters affect the system behavior and hence the
V. OTA-C BASED PID CONTROLLER performance.
The analogue circuit realization of PID controller using The sensitivity of output variable F with respect to input
OTA-C approach [16] shown in Fig. 5 is used in our work. parameter P is defined as the ratio of the fractional variation
The controller gains (Kp, Ki and Kd) are adjusted in F to the fractional change in P as the fractional variation
conveniently by varying the DC bias current (Itail) of OTA of the parameter approaches zero. The sensitivity of F with
that in turn changes the transconductance of the OTA. respect to P is mathematically expressed as
(∆𝐹⁄𝐹 ) 𝑃 ∆𝐹 𝑃 𝛿𝐹
𝑆𝑃𝐹 = lim = lim =( ) (12)
∆𝑃→0 (∆𝑃⁄𝑃 ) ∆𝑃→0 𝐹 ∆𝑃 𝐹 𝛿𝑃

The sensitivity of the controller gains Kp, Ki and Kd with

respect to transconductances and capacitance values are
shown to be
𝐾
𝑝 𝑖 𝐾 𝑖 𝐾 𝑑 𝑑𝐾 𝐾 𝐾 (13a)
𝑆𝑔𝑚5 = 𝑆𝑔𝑚4 = 𝑆𝑔𝑚7 = 𝑆𝑔𝑚1 = 𝑆𝑔𝑚6 = 𝑆𝐶1𝑑 = 1
𝐾𝑝 𝑖 𝐾 𝐾 𝐾 𝐾 𝐾
𝑆𝑔𝑚8 = 𝑆𝑔𝑚8 = 𝑆𝐶2𝑖 = 𝑆𝑔𝑚2
𝑑 𝑑
= 𝑆𝑔𝑚3 𝑑
= 𝑆𝑔𝑚8 = −1
(13b)

VI. CIRCUIT IMPLEMENTATION AND

SIMULATION RESULTS
This section describes the steps involved in simulating the
Fig. 5 OTA-C based PID controller realization [9] PID controller for DC motor speed control in Cadence design
environment. Both the controller and plant transfer function
Using the circuit in Fig. 5 an OTA-C based PI controller can are realized for the block diagram shown in Fig. 6 and the
be realized by eliminating the path III and PD controller can overall response is observed.
be realized by taking out the path II within the input Vi and
output Vo.
The controller gain 𝐾𝑝 is expressed as the ratios of the
transconductance of the OTAs 𝑔𝑚5 and 𝑔𝑚8 . A voltage-
mode OTA-C integrator is realised using OTA 𝑔𝑚4 and
grounded capacitor 𝐶2 . The transfer function 𝐴𝐼 of OTA-C
integrator is given by
𝐴𝐼 = 𝑔𝑚4 ⁄𝑠𝐶2 (10a)
A voltage-mode OTA-C differentiator is realised using OTA Fig. 6 Block diagram of DC motor control system
𝑔𝑚1 and grounded inductor 𝐿1 . The transfer function 𝐴𝐷 of Step1: Verification of the correctness of DC motor speed
OTA-C differentiator is given by control transfer function first using (i) MATLAB and
𝐴𝐷 = 𝑔𝑚1 (𝑠𝐿1 ) (10b) Simulink environment and then using (ii) Cadence
environment
The DC motor transfer function of (7) is converted into state
The grounded inductance 𝐿1 is realized by OTAs 𝑔𝑚2 and
𝑔𝑚3 and grounded capacitor 𝐶1 . The value of OTA simulated space model using MATLAB. The block diagram of state
grounded inductance is given by space model of DC motor implemented in Simulink
environment is given in Fig. 7 and the step response obtained

3
is shown in Fig. 8.

Fig. 7 Simulink state space model of DC motor (uncompensated

system) transfer function

Fig. 10 Step response of DC motor observed using Cadence Spectre

Step 2: CMOS OTA-C circuit implementation of PID
controller using Cadence environment

The CMOS schematic and symbol of Balanced OTA

implemented in Cadence Virtuoso environment are shown in
Fig. 11 (a) and (b) respectively. Table II shows the
dimensions of MOSFETS used in CMOS OTA simulation.
Table II. MOS transistor aspect ratio used to realize OTA
Fig. 8 Step response of DC motor (uncompensated system) transfer MOS Transistor Aspect ratio (W/L)
function using MATLAB and Simulink environment M1, M2 30 μm/ 3 μm
Next, we consider the circuit implementation of DC motor M3, M4 12 μm/ 3 μm
transfer function of (7) using voltage controlled voltage M5, M6 36 μm/ 3 μm
M7, M8 12 μm/ 3 μm
source (VCVS) model. The DC motor schematic and symbol
(subsystem block) implemented in Virtuoso Analog Design The CMOS balanced OTA based circuits are simulated using
Environment (ADE) of Cadence is shown in Fig. 9 (a) and Cadence Virtuoso and Spectre RF simulator tool using
(b) respectively. GPDK 180 nm technology parameters with supply voltages
Vdd = +5 V, Vss = -5 V.

(a)

(a)
(b)
Fig. 9 (a) Schematic circuit and (b) symbol of DC motor
implemented in Cadence using VCVS model

The DC motor schematic in Fig. 9(a) is simulated using

Spectre and the obtained step response is displayed in Fig.
10. (b)
Fig. 11 (a) CMOS schematic and (b) symbol of balanced OTA
implemented in Cadence environment
Next, the schematic and symbol of OTA-C PID controller
circuit in Fig. 5 is constructed in Cadence Virtuoso. The
transconductance values of OTAs are adjusted by changing
DC bias current to get the required proportional, integral and
derivative gain, Kp = 20.27, Ki = 49.41 and Kd = 1.805. The
transconductance and capacitance values used in the design
of OTA-C based PID controller in Fig. 5 are listed in Table
III.
Table III. Transconductance and capacitance values for OTA-C
based PID controller

g m1,7,8 40.18 μS (Ibias = 3.6 μA)

g m2 22.29 μS (Ibias = 1.88 μA)
g m3 19.91 μS (Ibias = 1.66 μA)
g m4 49.47 μS (Ibias = 4.53 μA)
g m5 814.73 μS (Ibias = 134 μA)
g m6 800.02 μS (Ibias = 130.25 μA) (b)
C1 1 μF Fig. 13 Step response of the compensated system (Plant with PID
1 μF controller) using (a) MATLAB, (b) Cadence
C2
The closed loop response of the compensated system for
Step 3: Finally, PID controller block is integrated with DC
different values of Kp, Ki and Kd are shown in Fig. 14 (a)-(c)
motor block of Fig. 9(b) as shown in Fig. 12 and simulated
respectively. Tables IV, V and VI show the closed loop
in Cadence environment with a step input.
performance of the system by independently varying the
parameters Kp, Ki and Kd respectively. Time domain
performance parameters like peak amplitude (PA),
Overshoot (% OS), settling time (Ts) and rise time (Tr) are
measured and tabulated for each case.

Table IV. Closed loop performance by varying Kp

Fig. 12 Schematic of PID controller with DC motor plant in Cadence Kp PA OS % Ts (s) Tr (s)
environment 15 1.09 9.3 1.83 0.538
17.5 1.07 6.76 1.76 0.516
Fig.13 shows the step response of the compensated system 20 1.05 4.64 1.63 0.489
(plant with PID controller). From the response of the 22.5 1.03 2.87 1.41 0.465
compensated system it is observed that the settling time is 25 1.01 1.43 0.66 0.445
1.61sec, overshoot is 4.44% and steady state error is
negligibly small (zero). Table V. Closed loop performance by varying Ki

Ki PA OS% Ts(s) Tr(s)

30 1 0 1.56 0.743
37.5 1.01 0.63 0.952 0.601
45 1.03 2.93 1.55 0.523
52.5 1.05 5.48 1.61 0.467
60 1.08 7.97 1.56 0.434

Table VI. Closed loop performance by varying Kd

Kd PA OS% Ts(s) Tr(s)

0.5 1.06 5.61 1.32 0.399
1.125 1.05 4.72 1.47 0.442
1.75 1.04 4.44 1.60 0.485
2.375 1.04 4.33 1.73 0.527
(a)
3 1.04 4.30 1.84 0.563

5
 realization; hence it requires less chip area. The controller is
integrated with DC motor plant in Virtuoso Analog Design
Environment (ADE) of Cadence and its response is observed
by varying Kp, Ki and Kd independently. The response of the
compensated system is found to have an overshoot of 4.44%
(less than 5%), settling time of 1.61 sec (less than 2 sec) and
zero steady state error and thus is in good agreement with
specification.
REFERENCES
[1] Li, Zhiqiang, and Changliang Xia, “Speed control of brushless DC
motor based on CMAC and PID controller,” IEEE Sixth World
Congress on Intelligent Control and Automation (WCICA 2006), vol.
2, pp. 6318-6322, Jun 2006.
[2] Amanda Danielle O. da S. Dantas, André Felipe O. de A. Dantas, João
Tiago L. S. Campos, Domingos L. de Almeida Neto, and Carlos
Eduardo T. Dórea, “PID Control for Electric Vehicles Subject to
(a) Control and Speed Signal Constraints,” Journal of Control Science
and Engineering, vol. 2018, Article ID 6259049, 11 pages, 2018.
[3] Qasim, M., Susanto, E. and Wibowo, A.S., “PID control for attitude
stabilization of an unmanned aerial vehicle quad-copter,” 5th IEEE
International Conference on Instrumentation, Control, and
Automation (ICA), pp. 109-114, Aug 2017.
[4] Getu B N, “Water Level Controlling System Using PID
Controller,” International Journal of Applied Engineering Research,
vol. 11, no. 23, pp. 11223-11227, 2016.
[5] Normey-Rico JE, Alcalá I, Gómez-Ortega J, Camacho E F, “Mobile
robot path tracking using a robust PID controller,” Control
Engineering Practice, vol. 9, no. 11, pp. 1209-1214, Nov 2001.
[6] Safwan A. Hamoodi, Rasha A. Mohammed, Bashar M. Salih, “DC
motor speed control using PID controller implementation by Simulink
and Practical,” International Journal of Electrical Engineering, vol.
11, no. 1, pp. 39-49, 2018
[7] Khan M.R., Khan A. A. Umer G. “Speed Control of DC Motor under
Varying Load Using PID Controller,” International Journal of
Engineering (IJE), vol. 9, no. 3, pp. 38-48, 2015.
[8] Petráš, I., “Tuning and implementation methods for fractional-order
controllers,” Fractional Calculus and Applied Analysis, vol. 15, no.
2, pp. 282-303, Jun 2012.
(b) [9] Tehrani, K. A. and Mpanda, A., “PID control theory”, In Introduction
to PID Controllers-Theory, Tuning and Application to Frontier Areas,
In Tech Publishers, pp. 213-228, 2012.
[10] Kamat, D. V., Mohan, P. A. and Prabhu, K. G., “Novel first-order and
second-order current-mode filters using multiple-output operational
transconductance amplifiers,” Circuits, Systems and Signal
Processing, vol. 29, no. 3, pp. 553-576, Jun 2010.
[11] Dimeas, I., Petras, I. and Psychalinos, C., “New analog
implementation technique for fractional-order controller: a DC motor
control,” AEU-International Journal of Electronics and
Communications, vol. 78, pp. 192-200, Aug 2017.
[12] Chaudhary, H., Khatoon, S. and Singh, R., “ANFIS based speed
control of DC motor,” IEEE Second International Innovative
Applications In Computational Intelligence on Power, Energy and
Controls with their Impact on Humanity (CIPECH), pp. 63-67, Nov
2016.
[13] Lennartson, B. and Kristiansson, B., “Evaluation and tuning of robust
PID controllers,” IET control theory & applications, vol. 3, no.3, pp.
294-302, Mar 2009.
[14] Kamath, D. V., “Overview of OPAMP and OTA based
Integrators,” International Journal of Innovative Research in
(c)
Electrical, Electronics, Instrumentation and Control Engineering,
vol. 3, no. 9, pp. 74-81, Sept 2015.
Fig. 14 (a)-(c) Closed loop response of the compensated system by varying
[15] Al-Hashimi, B. M., Dudek, F., Moniri, M. and Living, J., “Integrated
Kp, Ki and Kd respectively
universal biquad based on triple-output OTAs and using digitally
programmable zeros,” IEE Proceedings-Circuits, Devices and
VII. CONCLUSION Systems, vol. 145, no. 3, pp.192-196, Jun 1998.
[16] Erdal, C., Toker, A. and Acar, C, “OTA-C based Proportional-
The voltage-mode OTA-C PID controller is designed for DC Integral-derivative (PID) controller and calculating optimum
parameter tolerances,” Turkish Journal of Electrical Engineering &
motor speed control. The realized OTA-C PID controller Computer Sciences, vol. 9, no. 2, pp. 189-198, Sep 2001.
uses only OTAs and grounded capacitors. It is a resistor-less