Electrical Engineering

ECG Transmitter

PROJECT REPORT ON DESIGN OF ECG

TRANSMITTER USING AGILENT ADS

BY
MANU MITRA 0795410

UNIVERSITY OF BRIDGEPORT
2008-2009

i University of Bridgeport
Electrical Engineering

ECG Transmitter

Name : - Manu Mitra

Student Id : - 0795410

Major : - Electrical Engineering

Class : - Microware Engineering

Subject : - Project Report on Design of ECG Transmitter

using Agilent ADS

Submission : - Dec 23rd 2009

Comments:-

Marks Obtained Professor’s Signature.

ii University of Bridgeport
Electrical Engineering

ECG Transmitter

Contents
1. Project overview

1.1 Introduction
1.2 Importance of work and aims of project
1.3 Project Outline
2. Background
2.1 ECGs
2.2 Holter monitors
2.3 Electrical Safety
2.4 Mains Interference
2.5 Defibrillator Protection
2.6 Low Pass Filter
2.7 Telemetry
3. Isolated ECG Hardware
3.1 ECG Amplifier
3.2 Driven-Right-Leg Systems
3.3 Defibrillator Protection

3.4 Filtering and gain Stage

3.5 A/D Conversion
4. Antennas
4.1 Antenna Analysis and Design
4.2 Wireless link hardware implementation

i University of Bridgeport
Electrical Engineering

ECG Transmitter

5. Simulation
6. References

ii University of Bridgeport
 1 Introduction

1.1 Introduction

In 1996, Intellidesign Pty Ltd (then IntelliMed) was approached by a

cardiologist to design an ECG Holter monitor. This original device was a two or
three lead, single channel ECG device, which could continuously record for a
maximum of one hour. Additionally, the device had Polar Chest Strap capabilities
for the added functionality as a Heart Rate monitor. The device could operate in
four different modes: 1-hour ECG Recording Mode, in which the device would
record one continuous hour of near diagnostic quality ECG trace during exercise;
Event Recording Mode, in which the device would record up to 60, one minute
segments around a recorded event, over a period of up to 24 hours; Heart rate
Recording Mode, in which the unit would have the capacity to record up to 24
hours of heart rate information; and ECG Telemetry Mode, in which the unit
would transmit, via a Radio Frequency (RF) link, a real-time ECG signal to a
receiver unit.

1.2 Importance of Work and Aims of Project

The American Heart Association claims that cardiovascular disease is the

cause of an average of one death every 33 seconds. In Australia, one person dies
every 10 minutes as a result of cardiovascular disease. With these statistics, there is
an obvious need for diagnostic devices in the prevention and detection of heart
disease.
The aim of the project is to design two ECG devices to aid in the detection
of heart disease.

1.3 Project Outline

The purpose of this project is to design ECG transmitter using the software
Agilent ADS.

1.3.1 Tasks

The following tasks need to be undertaken:

 Design the analogue circuitry for the ECG device
 Instrumental Amplifier
 Right-Leg-Drive

1 University of Bridgeport
  Defibrillator Protection
 Low pass filtering
 High pass filtering
 Gain stage (for maximum signal)
 Analogue to Digital Conversion

Design of Wireless Link

 Transmitter design and implementation

 Receiver design and implementation

2 University of Bridgeport
 2 Background of ECG

2.1 ECGs

In order to be able to design an ECG recording device, one must have some
background into the physiology of the heart (with respect to its bio potentials) and
an understanding of exactly what it is an ECG is, and its function.

“The electrical activity of the heart can be recorded to monitor cardiac

changes or diagnose potential cardiac problems. The principle involved is simple:
body fluids are good electrical conductors. Electrical impulses generated in the
heart are conducted through body fluids to the skin, where they can be detected and
printed out by a sensitive machine called an electrocardiograph. This printout is
called an electrocardiogram, or ECG.”

Typically, an ECG is comprised of a series of three distinguishable waves or

components (known as deflection waves), each representing an important aspect of
cardiac function (as can be seen in Figure 1). The first wave, known as the P wave,
represents atrial depolarization, and is a result of the depolarization wave from the
Sinoatrial node (SA node) through the atria. This action precedes and is the cause
of atrial contraction.

The QRS complex is the result of ventricular depolarization. It is caused by

the electrical activity spreading from the Atrioventricular node (AV node), through
the ventricles via the Purkinje fibers, and precedes ventricle contraction. During
this time, atrial repolarization is also occurring; however its occurrence is usually
masked by the large QRS complex being detected.

Finally, the T wave occurs when the ventricles repolarize. Repolarization, is

slower than repolarization, hence the T wave is usually wider than the P wave and
the QRS complex.

“In a healthy heart, the size, duration, and timing of the deflection waves
tend to be consistent. Thus, changes in the pattern or timing of the ECG may reveal
a diseased or damaged heart or problems with the heart’s conduction system.”

3 University of Bridgeport
(Figure 1): An electrocardiogram tracing (lead 1) illustrating the three normally recognizable deflection
waves and the important intervals

2.1.1 Lead Configuration

To perform a clinical electrocardiograph, it is important that more than one

lead (also known as a channel) be recorded in order to accurately describe the
heart’s electrical activity. There are two planes in which these leads may lie: the
frontal plane (the plane of the body that is parallel to the ground when one is lying
on one’s back); and the transverse plane (the plane of the body that is parallel to
the ground when one is standing erect). For a two or three channel ECG, only the
leads in the frontal plane are required.

4 University of Bridgeport
(Figure 2): Position and orientation of the three bipolar limb leads. (Leads I, II, and III)

The frontal plane of an ECG consists of three basic leads, as can be seen in
figure 2. These leads are the result of the various combinations of pairs of
electrodes located on the right arm (RA), the left arm (LA) and the left leg (LL) of
the patient. The resulting leads are: lead I, LA to RA; lead II, LL to RA; and lead
III, LL to LA. Very often an electrode is also placed on the right leg (RL) and
grounded or connected to a Driven-Right-Leg circuit.

2.1.2 ECG Parameters

The hardware of an ECG monitor, in its most simple form, is a biopotential

amplifier. The maximum voltage that can be acquired at the skin electrodes is 0.5
to 4 mV (Figure 3). In order to perform any analysis on this data, the signal will
need to be amplified, before any signal analysis is performed.

5 University of Bridgeport
(Figure 3): Voltage and Frequency Ranges for Some Important Parameters that are measured in the Human
Body

2.1.2 Exercise ECGs

“Electrocardiograms often are obtained during exercise. These are valuable

diagnostic tests. As exercise intensity increases, the heart must beat faster and
work harder to deliver more blood to active muscles. If the heart is diseased, an
indication may show up on the electrocardiogram as the heart increases its rate of
work. Exercise ECGs are also invaluable tools for research in exercise physiology
because they provide a convenient method for tracking cardiac changes during
acute and chronic exercise.”

2.2 Holter monitors

A Holter monitor is a small portable ECG device named after its inventor,
the American biophysicist Norman Holter. The device records the electrical
activity of the heart, typically over a 24-hour period, while the patient keeps a
diary recording their activities and any symptoms felt. The ECG recording is then
analyzed, and irregular heart activity is correlated with the patient’s record of their
activities and symptoms. The Holter monitor is useful for identifying disturbances
which are sporadic and which are not readily identified with the usual resting
electrocardiogram test. In some instances, the patient may manually activate an
“event” if they are experiencing symptoms such as palpitations, chest pain, or
irregular rhythms are noticed.

6 University of Bridgeport
2.3 Electrical Safety

For electricity to have an effect on the body, two conditions must be

satisfied: an electrical potential difference must be present; and the individual must
form part of the electrical circuit, permitting current to flow in a loop through the
body. Assuming these conditions are met, the actual effect of the electrical current
passing through the body is dependent upon the magnitude of the current and the
pathway the current takes through the body.

From Ohm’s Law (V = IR), it is evident that the magnitude of current

flowing in a circuit depends on both the voltage between the connections and the
electrical resistance of the various pathways through the body. Most tissue in the
body contains a high percentage of water hence it is usually considered to be a
fairly good conductor of electricity. The body’s surface resistance also affects the
amount of current that flows through the body, with contacts being made directly
onto moist skin having little surface resistance and passing a considerable amount
of current directly into the body.

“The danger in one’s exposure to various electrical devices is not the voltage
but the magnitude and pathways of electrical current that flows through body tissue
and major organs. In addition, for electrically sensitive organs such as the heart,
physiological effects are also dependent on the frequency and waveform of the
current.”

As previously mentioned, the physiological effect of current depends on the

path of current through the body. Due to the distribution of current flow being
determined by the resistance of local tissue, the electrical current tends to spread
out. This leads to a number of alternate pathways for current flow through the
body. It is possible for an individual to come into contact with a relatively large
voltage source and experience only a minor electrical shock, whilst at the same
time, application of a relatively low voltage directly to the heart (e.g., during
surgery or via the invasive components of certain diagnostic and therapeutic
equipment) can produce ventricular fibrillation, a life threatening arrhythmia.

Whenever electrical current passes through a resistive element, some of the

electrical energy is dissipated in the form of heat. If the temperature is high enough
to affect the biological tissue, damage, in the form of a burn, will result. This
phenomenon is almost certain to occur when the current intensity is quite high, and
the pathway is limited to a relatively small area.

7 University of Bridgeport
 When a medical device (e.g., electrocardiogram or blood pressure monitor)
is connected directly to a patient, there must be some sort of patient input isolation.
“An isolated patient connection not only limits leakage current levels from the
patient contact points to very low levels but also reduces the amount of current that
can flow into the patient connection from another (defective) device. It is
recommended that equipment with isolated patient leads be used whenever the
device is to be in contact with the heart or with a fluid-filled catheter located within
the heart of great vessels.”

Although the two devices being designed will be powered by battery

(usually in this situation there would be no need for any isolation), there is a USB
(Universal Synchronous Bus) connection for download capabilities. Due to USB
architecture, it is difficult to use the same input receptacle for patient leads and
USB cable. Since two different plugs will be used, a situation in which the patient
is accidentally connected to the computer (via the USB cable) could arise. In the
event of an electrical fault, the ECG hardware could see a very high voltage, and
hence a high dose of current, being passed into the patient. A situation like this
must be avoided at all costs, and so it is imperative that all signal lines to the
patient are electrically isolated, and that any circuitry connected to these isolated
lines be powered by an isolated voltage source.

2.4 Mains Interference

Power lines and electrical equipment at 240V AC generate ambient

electromagnetic fields at a frequency of 50 Hz and its corresponding harmonics.
By Faraday’s law, if a closed loop is placed in a time-varying magnetic field, a
current is induced in the loop.

When acquiring an ECG, a differential signal, generated by the patient’s

heart, appears between the electrodes placed on the patient’s limbs. This signal is
that required by the cardiologist. However, a large 50 Hz (or 60 Hz [US])
common-mode signal is typically present between each electrode ant the local
power-system ground. This is due to very small capacitances between the patient’s
body and the power line, causing the patient to be connected to the voltage line.
Similar small capacitances connect the patient to ground. A differential signal of
approximately 4mV should be present at the input of the ECG amplifier; however,
it will be masked by a much larger 50 Hz common mode signal. Ideally the ECG
amplifier should only be receptive to the differential signal.

8 University of Bridgeport
 There are several different methods of blocking the common-mode voltage.
The first being to use good quality limb leads that have been twisted and insulated.
Following this, the ECG amplifier should have a common-mode rejection ratio
(CMRR) of at least 100dB, and finally low-pass filtering the signal below 50 Hz
will cause the common mode signal to be removed.

Another method is to implement a negative feedback circuit where the

common mode voltage is fed to a third electrode via a feedback amplifier to
oppose the common mode signal. This method is known as right-leg-drive.

2.5 Defibrillator Protection

Defibrillators are used in many emergency situations to restart the heart, by

applying a short pulse of very high voltage. When using a defibrillator, it is
imperative that an ECG be connected to the patient. Whilst an ECG device circuit
would usually only expect maximum input voltages of approximately 4 mV, ECG
preamplifier inputs must be designed to withstand a high-voltage (greater than a
kilovolt) burst from a defibrillator, which may last from 5 to 20 ms.

Many ECG devices use neon glow lamps across the input lines, which
bypass the voltage, and series resistors, which serve to limit the current flow. The
neon lamps are primarily a pair of electrodes mounted in a glass casing, which has
been evacuated and filled with low-pressure neon gas, or a mixture of inert gases.
The impedance across the electrodes is usually very high; however the impedance
will suddenly drop to a very low value if the voltage across the electrodes exceeds
the ionization potential of the gas. Most lamps used in medical monitors have a
firing potential between 45 and 70V. When a defibrillator is used, most of the
charge is bypassed to ground and the gas inside of the lamps will ionize.

2.6 Low Pass Filters

2.6.1 Basic Low Pass Filter

Two-stage RC second order low-pass filters as seen in Figure 4, are limited

due to a maximum Quality factor (Q) of ½. When R1=R2 and C1=C2, Q=1/3. The
Quality factor approaches the maximum value of ½ when the impedance of the
second RC stage is much larger than the first. Most filters require Qs larger than ½.

9 University of Bridgeport
(Figure 4): Basic Second Order Low-Pass Filter.

2.6.2 Sallen-Key Filter

To attain a larger Q value, a positive feedback amplifier may be used. If the

positive feedback is controlled (localized to the cut-off frequency of the filter)
almost any Q can be realized. However, Q values are limited by the physical
constraints of the power supply and component tolerances. In 1955, R. P. Sallen
and E. L. Key described these filter circuits, and hence they are generally known as
Sallen-Key filters.

The operation of a Sallen-Key filter can be described qualitatively:

 At low frequencies, the signal is buffered to the output. (C1 and C2 appear
as open circuits)
 At high frequencies, the signal is shunted to ground at the amplifier’s input.
This causes no signal to be seen at Vo. (C1 and C2 appear as short circuits)
 Near the cut-off frequency, there is positive feedback through C2, which
enhances the Q of the signal. (The impedance of C1 and C2 is of the same
order as R1 and R2)
The standard frequency domain equation for a second order low-pass filter is:

Where fc is the corner frequency and Q is the quality factor. When f<<fc Equation

10 University of Bridgeport
[2.6.2-1] reduces to K, and the circuit passes signals multiplied by a gain factor K.
When f=fc, Equation [2.6.2-1] reduces to –jKQ, and signals are enhanced by the
factor Q. When f>>fc, Equation [2.6.2-1] reduces to –K(fc/f)2 and signals are
attenuated by the square of the frequency ratio, which indicates a second order
low-pass filter.
Figure 5 shows the Sallen-Key circuit configured for low-pass:

(Figure 5): Low-Pass Sallen-Key Circuit

The ideal low-pass Sallen-Key transfer function is:

These equations can be simplified to make them more efficient.

11 University of Bridgeport
2.6.3 Quality factor considerations:

Q < 0.5 the poles of the transfer function are real.

Q > 0.5 the poles are complex.
Q > 0.707 the frequency response peaks above Ho just beyond the corner
frequency.
This peak can be quite large for large Q values.
Q = 0.707 this Q produces a maximally flat response, i.e. the sharpest fall-off near
the corner without any peaks larger than Ho.
[Ho = maximum gain]

2.6.4 Simplification 2: Set Filter Components Equal

This now means that fc and Q are independent of one another, thus simplified
greatly. The gain (K) of the circuit now determines Q. RC sets fc, (the capacitor is
chosen and the resistor calculated).

2.6.5 Component Selection

Theoretically, any values of R and C that satisfy the equations may be used,
but practical considerations call for component selection guidelines to be followed.
In the case of the low-pass Sallen-Key filter, the ratio between the output
impedance of the amplifier and the value of filter component R sets the transfer
functions seen at frequencies well above cut-off. The larger the value of R, the
lower the transmission of signals at high frequency. Making R too large has
consequences in that C may become so small that the parasitic capacitors,
including the input capacitance of the amplifier, cause errors.

The best choice of component values depends on the particulars of your circuit and
the tradeoffs you are willing to make. General recommendations are as follows:

12 University of Bridgeport
1. Capacitors
 Avoid values less than 100 pF.
 Use NPO if at all possible. X7R is OK in a pinch. Avoid Z5U and other low
quality dielectrics. In critical applications, even higher quality dielectrics
like polyester, polycarbonate, mylar, etc., may be required.
 Use 1% tolerance components. 1%, 50V, NPO, SMD, ceramic caps in
standard E12 series values are available from various sources.
 Surface mount is preferred.

2. Resistors
 Values in the range of 10k_ to 200k_ are best.
 Use metal film with low temperature coefficients.
 Use 1% tolerance (or better).
 Surface mount is preferred.

2.7 Telemetry

Wireless medical telemetry is generally used in the remote monitoring of a

patient’s physiological parameters (e.g., cardiac signals) via a radio-frequency
(RF) link. The use of wireless medical telemetry allows patients greater mobility
by freeing them from the need to be connected to hospital equipment that would
otherwise be required to monitor their condition. Telemetry can also be used to
monitor an athlete’s cardiac parameters as they train.

In recent years the USA, Europe and Australia have all introduced a new
unlicensed band (ISM band) for Industrial Scientific and Medical applications.
Australia has reserved the 918-926 MHz band for ISM applications. The US has a
similar yet wider ISM band with the frequency range from 902-928 MHz reserved.
Therefore a device can be designed in Australia and be distributed in the US, with
absolutely no need for alteration.

Many Transceiver integrated circuits (IC’s) can be found on the market

today. These devices are designed especially to operate (Transmit and Receive)
within the standard ISM band.

13 University of Bridgeport
 3. Isolated ECG hardware

3.1 ECG Amplifier

In order to have a good signal to noise ratio, differential amplifiers are used.
Since the output is proportional to the difference between the two voltages, this
circuit has a fairly good common mode rejection. Unfortunately, the differential
amplifier’s performance is limited due to low input impedance. This problem has
been avoided by incorporating a dual Instrumentation amplifier in place of a
differential amplifier.

The instrumentation amplifier chosen (INA2321), is a micro Power, Single-

Supply CMOS dual device from Texas Instruments. The INA2321 provides low-
cost, low noise amplification of differential signals and boasts a micro power
current consumption of just 40_A/channel and a high CMRR of 94 dB, and states
as one of its applications the use in physiological amplifiers, in particular ECGs.

(Figure 6): Power specifications of INA2321 Instrumentation Amplifier

14 University of Bridgeport
(Figure 7): INA321 Internal diagram and basic connections

3.1.1 Setting the Gain

The ratio of R2 to R1, or the impedance between pins 1, 5, and 6, determines

the gain of the INA321. With an internally set gain of 5, the INA321 can be
programmed for gains greater than 5 according to the following equation:

The INA321 is designed to provide accurate gain, with gain error guaranteed
to be less than 0.1%. Setting gain with matching TC resistors will minimize gain
drift. Errors from external resistors will add directly to the guaranteed error, and
may become dominant error sources.
A gain of 10 was chosen for this stage of the circuit, therefore, R1 = R2 =
100kohms.

15 University of Bridgeport
3.2 Driven-Right-Leg System

A right-leg-drive system is an alternative to the grounding of a patient in

many electrocardiograph devices. In these circuits, an electrode is attached to the
right leg and connected to the output of an auxiliary op amp, where the common-
mode voltage is sensed by two averaging resistors, inverted, amplified, and
returned to the right leg. This causes negative feedback, which drives the common-
mode voltage low. The body’s displacement current in turn flows to the op-amp
output circuit, which reduces the pickup of the ECG amplifier, and effectively
grounds the patient.

The circuit also acts as an electrical safety device, by un grounding a patient

when the amplifier stops driving the right leg due to the saturation of the auxiliary
op amp caused by an abnormally high voltage appearing between the patient and
ground. This causes the large parallel resistances, Rf and Ro, to be between the
patient and ground. Although these resistances limit the current, they do not protect
the patient, as a large voltage would breakdown the op-amp transistors and large
currents would flow directly to ground.

(Figure 8): Schematic showing Right-Leg-Drive, Defibrillator protection, and Instrumentation amplifier

16 University of Bridgeport
 The right-leg-drive circuit can be seen in the lower section of Figure 8. It
consists of a buffer (which gets its common mode input from the Instrumentation
Amplifier), followed by a common Driven-Right-Leg circuit.

Using Kirchhoff’s Voltage Law, solve VCM :

VCM should be made very small. This can be done by making Rf >> Ra, and Rf =
Ro.
Typically, Ro should be quite large, (around 5Mohms).
When Rf = 5M_, Ra = 25k_, then typically id = 0.2_A. This then sets VCM at
2.5mV, which is very small.

17 University of Bridgeport
Choose real and available resistor values close to these:
Rf = Ro = 4.7Mohms
Ra/2 = 10kohms
An unpopulated resistor R35 was added in the design for an alternative right-leg-
drive circuit that was not used.

3.3 Defibrillator Protection

As discussed in chapter 2, ECG devices use neon glow lamps across the
input lines, to provide defibrillator protection. These neon lamps are used to bypass
large voltages. The maximum strike voltage (firing potential) of the lamps being
used in the design is 90V. A 330k_ resistor is placed in series on each input line,
which limits the current flow. 1206 Resistors have been used here, as they have a
maximum voltage rating of 200V, and a power rating of .125W, which far exceeds
the ratings of smaller resistors, which are being used for the remainder of the
circuit due to size restrictions.

3.4 Filtering and Gain stage

ECGs have a frequency range from between 0.1 to 250 Hz. Due to noise at
50Hz from the power mains, the acquired signal requires low-pass filtering below
50 Hz. For these devices, the corner frequency has been selected at 34 Hz. To
block any wandering DC effects, a high-pass filter with a very low cut-off
frequency has also been implemented.

3.4.1 High Pass Filter Design & Calculations

To nullify any wandering DC effects, a simple RC high pass filter set to

approximately 1.6Hz has been implemented on the ECG Hardware directly
following the Instrumentation amplifier stage.
To calculate the filter characteristics, the High Pass filter equation can be used.

18 University of Bridgeport
3.4.2 Low Pass Filter Design & Calculations

An LT1496 quad channel op-amp from Linear Technologies has been chosen for
use in the positive feedback amplifier. Two cascaded second order sallen-key
filters have been used for the Low Pass filter. By cascading the filters, sharper roll-
off and better attenuation is achieved.

Step 1: Calculate the gain of the filter from the ideal Quality factor.
fc = 34 Hz. Q=0.707 (flat response, sharp falloff).

19 University of Bridgeport
Step 2: Calculate the resistor divider circuit from the known gain.

Step 3: Choose values for R3 & R4

Choose R4 = 30kohms.

Step 4: Calculate the filter requirements

20 University of Bridgeport
3.4.3 Final Gain Stage

The final gain stage has been designed around a simple non-inverting
amplifier circuit similar to the one seen in Figure 9.

(Figure 9): Non-inverting operational amplifier circuit

To calculate the final gain, the gains from all the previous stages must be
calculated.
ECG expected input: 2.7 mV

21 University of Bridgeport
Instrumentation Amplifier: 10
Low pass filter: 1.588
Maximum voltage output: 1.3V

Choose R1 = 560ohms; and R2 = 33kohms.

3.5 A/D conversion

Instead of using the onboard Analogue to Digital (A/D) converter on the

MSP430 microcontroller on the ECG device which would use up quite a lot of
power, two micro power A/D converters from Linear Technology (LTC1285) were
chosen for the following reasons: their typical supply current is a low 160_A and
there is an Auto shutdown feature which typically draws only 1nA. This is an
extremely useful feature to have in a micro power device. For the majority of time,
the A/D converter will operate in the shutdown mode, only being turned on when a
conversion is needed. This saves a great deal of power and in a situation where
power needs to be minimal this is essential.

22 University of Bridgeport
3.6 Isolation

For the isolation of signal lines, coming out of the A/D converter, very low-
power opto couplers from Agilent (HCPL0730A) can be used. Each channel can be
driven with an input current as low as 40 μA. These high gain couplers use a LED
and an integrated high gain photo detector to provide an extremely high current
transfer ratio between input and output.

23 University of Bridgeport
 4. Antennas

4.1 Antenna Analysis & Design

For RF designers developing low-power radio devices for short-range

applications, antenna design has become an important issue for the total radio
system design. Taking the demand for small size and low cost into account in the
development of such radio modules, a small-tuned loop antenna on the same
printed circuit board as the radio module is a good solution.

The definition of an electrically small loop antenna is an antenna such that

the circumference is less than one-tenth of a wavelength. These antennas have field
patterns similar to those of infinitesimal dipoles, null perpendicular to the plane of
the loop and with its maximum along the plane of the loop.

This section describes the geometry and the electrical equivalent circuit for a
rectangular loop antenna. Physical dimensions for the antenna are used to calculate
the components in the antenna equivalent circuit and the antenna efficiency. The T
matching method is presented in order to match the impedance of the antenna to a
transmitter/receiver.

4.1.1. Loop antenna physical parameters

The loop antenna physical parameters used in the calculations are

a1 = loop antenna width [m]
a2 = loop antenna length [m]
b1 = thickness of loop conductor [m]
b2 = width of loop conductor [m]

For a loop antenna fabricated on a printed circuit board (PCB), the thickness
of the loop conductor b1 means the thickness of the copper layer on top of the
substrate. In order to calculate the electrical parameters of the antenna, the
rectangular loop has to be modeled as an equivalent quadratic loop, and the planar
loop conductor has to be modeled as a wire conductor with an equivalent circular
radius.

From the parameters above the equivalent quadratic sides of the loop are
given by:

24 University of Bridgeport
The calculated equivalent quadratic sides are used in the formulas below for the
loop area A, and the inductances LA and LI.
The loop area is given by:

The equivalent electrical circular radius of the loop conductor is given by:

In electrostatic, the equivalent radius represents the radius of a circular wire whose
capacitance is equal to that of the noncircular geometry.

Directivity of antenna:

25 University of Bridgeport
The maximum effective aperture is calculated by:

Electrically, the loop is about 13 times larger than its physical size. To be effective,
a small loop must be larger electrically than its physical size.

4.1.2. Loop antenna electrical equivalent circuit

The input impedance of the loop antenna needs to be calculated in order to

estimate the capacitor CP used to resonate the antenna. Radiation resistance, loss
resistance of the loop conductor and other ohmic losses has to be determined in
order to estimate the antenna efficiency.

The equivalent circuit for the input impedance of a small loop when the loop
is used as a transmitting antenna is shown in Figure 10.

26 University of Bridgeport
(Figure 10): Loop Antenna Equivalent Circuit (Transmit mode)

The loop antenna input impedance Zin is given by:

27 University of Bridgeport
c is the speed of light equal to 3×108 m/s
f0 is the resonance frequency in Hz.
f0 = 915MHz

28 University of Bridgeport
The loss resistance of the loop inductor is given by

The inductive reactance XA of the loop is computed using the inductance LA of

Square loop with sides a and wire radius b:

29 University of Bridgeport
The reactance Xi of the loop conductor can be computed using the inductance Li of
the loop. For a single turn this can be approximated by:

The additional ohmic losses that is introduced mainly because of ESR (Equivalent
Series Resistance) of the capacitor Cp is given by:

30 University of Bridgeport
 As the above expression indicates, the main determinate of the maximum
quality factor (Q) of a loop antenna is the ESR (i.e. the quality factor) of the
capacitor Cp. The addition of a resistor Rq in parallel with Cp can be used to
control the Q-value of the antenna at the same time as reducing the input
impedance of the antenna.

The capacitor Cp is used in parallel to Zin to resonate the antenna. It can be

shown that the parallel capacitor Cp at resonance is given by:

The antenna efficiency can then be expressed as:

31 University of Bridgeport
4.1.3. Antenna impedance and Q-value with chip capacitors in the loop

The antenna impedance is dependent of both feed length and Q-value (read
parallel resistor RQ). The Q-value is independent of the impedance of the antenna,
which means that one chooses a Q-value and then chooses the feed length.

In order to ensure that the transmitter power radiation and receiver

sensitivity are always identical, the Q-value of the antenna has to be exactly
matched to the capacitors that will be used to tune the antenna to the correct
resonant frequency. The Q-value of the loop can be determined from the following
formula:

The tol variable is the percentage tolerance of the capacitors. The equation
assumes that the variation in radiated power due to capacitor variations should be
lower than 3dB. If the tolerance is 4% the Q-value will be 50. If a higher Q-value
is chosen, each antenna has to be tuned to keep the variation in radiated power
lower than the assumed 3dB.

Capacitors with a tolerance of approximately 5% have been chosen for the

antenna, hence the Q-value of the antenna must be:

32 University of Bridgeport
4.1.4. Range calculations

Unlike optical IR systems, RF systems operations in the UHF band are not
restricted to line-of-sight coverage. This is because of the diffraction and reflection
of radio waves at edges and conductive surfaces as well as the capability of RF
waves to penetrate dielectric materials.
The range calculation parameters are:

 transmitter output power, PRF [dBm]

 transmitter and receiver antenna efficiency,
 antenna range, R [m]
 free-space loss, Lp [dB]
 additional propagation losses other than free-space loss, Lx [dB]
 receiver sensitivity, S [dBm]

When the transmitter and receiver have a clear, unobstructed line-of-sight path
between them, the free space propagation model is used to predict received signal
strength.

The free-space loss factor Lp is given by:

This loss factor accounts for the losses due to the spherical spreading of the
energy by the antenna. From the equation it can be seen that the received power
strength is dependent upon the square of the transmitter-receiver separation
distance. This implies that the received power decays with distance R at a rate of
20 dB/decade.

When reflection and polarization-matched antennas are aligned for

maximum directional radiation and reception, the communication range can be
calculated by the following formula. PRF is the output power and S is the
sensitivity of the antennas.

33 University of Bridgeport
 This equation assumes that the two antennas are separated by a distance

where D is the largest dimension of either antenna. The operation

range may increase due to the wave guidance which sometimes occurs along
conductive surfaces. As previously mentioned, the free-space path analysis applies
to line-of-sight propagation, hence the other propagation losses Lx such as signal
reflection, diffraction, scattering and polarization losses will need to be considered.
With the inclusion of these losses, the communication range can now be
calculated:

Given the required range R, assumed losses Lx, sensitivity S and equal TX/RX
antennas, the necessary output power PRF is given by:

34 University of Bridgeport
4.1.5. Selection of Capacitors and Resistor

The antenna is designed by placing three series capacitors in parallel with a

resistor. This resistor determines the Quality factor of the circuit.
Capacitors in series:

35 University of Bridgeport
The manufacturer of the transceiver chip (Nordic VLSI) recommends the use of an
18kohm 1% resistor for R q.

36 University of Bridgeport
4.2 Wireless link hardware implementation

The Transceiver and Receiver circuits are very similar. To have a matched
wireless link, the antennas need to be identical, and the Transceiver chip’s external
circuitry remains the same. The majority of the external components were taken
directly from the nRF903 Transceiver datasheet, with the exception of the antenna
components which were calculated in the Antenna design section (4.3) of this
project

4.2.1. Transmitter

The Transmitter circuit begins with a rotary locking FPC connector along
which the power and four output lines from the primary ECG device are input. One
of the data lines serves to power up the wireless board, by switching on a P-
channel MOSFET. When there is no signal, the MOSFET is off, because the data
line is high at 3.3V and no current is drained across the resistor. When the line is
switched to an active high, the data line will read low at 0V, and the MOSFET will
turn on, thus providing power to the Transmitter circuit.

The three other data lines are used in the following way: Data line for the
transmission of serial data; a Clock line to read the data in from; and a Transmit
enable line which when high, tells the on-board microcontroller (MSP430F1232)
to put the Transceiver chip into transmit mode, by passing along a high to the
TXEN pin.

The microcontroller (MSP430F1232) is the link between the input data lines
and the Transceiver chip. It initiates and configures the nRF903 and is responsible
for the passing of ECG waveform data to the wireless link.

4.2.2. Receiver

As previously mentioned, the Receiver hardware is not unlike the

Transmitter hardware, except for the addition of a USB circuit. Data is now input
via the RF link which passes the data to the onboard microcontroller
(MSP430F1232 - same model as the Transmitter microcontroller).

37 University of Bridgeport
38 University of Bridgeport
 5. Simulation

This application notes describes a RF transmitter system simulation using

Agilent Advanced Design System (ADS). When using RFIC chipset to design a RF
transmitter system, it is important to select a suitable set of chips, which match the
required performance. ADS gives us a visual lab for selecting RFIC chip. In this
application notes, the key steps in building the system schematic and computer
simulation results using ADS are presented.

System Schematic and Parameter Setting

The smart antenna transmit system consists of one base band up-converter,
one IF amplifier, two IF filters, one RF up-converter, two RF filters, one RF pre-
amplifier, and one RF power amplifier. We need to make sure that the system will
have enough output power, linearity and sufficient attenuation on unwanted
frequency spectrum.

The entire component models can be selected from RF/Analog library.

Amplifier and mixer models can be obtained from the system amplifier and mixer
library. Filter models are acquired from band pass filter library. The two local
oscillator models are obtained from sources in frequency domain library. P_1 tone
source in harmonic balance library is selected to simulate the base band input
signal. Finally, a 50 Ohm terminal in harmonic balance library is used to terminate
the transmitter output.

Because all the RFIC chips have detailed specifications, the parameter
settings in each corresponding model should be set up as close as what can be
gotten from the specifications. In setting parameter to an amplifier, it is important
to use correct parameter such as noise figure, gain, power saturation point (Psat),
three order intercept (TOI) power. The gain compress characteristic parameter of
amplifier is fundamental to system nonlinearity characteristic simulation. While
setting parameter of a mixer, P lomin should be taken care of. The meaning of Plomin
is the minimum Local oscillator input power before mixer starved. But for some
unknown reason, it had better to leave it blank. Otherwise, an error convention
gain would occur.

Filter is the only kind of component determining the system band pass
characteristic; correct filter models should be selected to emulate real filter
performance. In the design, all the filters selected are Butterworth type. All the

39 University of Bridgeport
parameters of filters, such as center frequency, insertion loss, pass band width, stop
bandwidth, pass band ripple, and stop band attenuation, etc., are adopted from the
real filter specifications.
PLL frequency synthesizers are used as the first local oscillator and the second
local oscillator for up convention, in order to achieve enough short term and long-
term frequency stability. The phase noise of the two synthesizers is derived from
the performance of selected reference frequency and VCO. Both of the two local
sources use oscillator with phase noise in frequency domain source library to
simulate. The phase noise parameter of the two models depends on the derived
result.

Simulation Result

The simulation setup shown in Fig.2 is used to simulate the system

performance.

Harmonic Balance simulator HB1 is used to investigate the performance of

the system output spectrum. While doing the harmonic balance simulation, the
base band frequency is 40MHz. The frequencies of the two local oscillators are
160MHz and 1750 MHz. The simulation result is shown in Fig. 3. Large signal S
parameter simulator HB2 is also used to simulate the system band pass
performance by using parameter sweep controller to control the base band
frequency sweep from 20MHz to 200MHz. The system band pass performance is
presented in Fig. 4. In order to find the output power performance and system
saturation performance versus input base band power variation, a LSSP simulator,
HB3 that incorporates a parameter sweep controller sweep2 to control the input
power sweep is used. In conducting these simulations, the gain compression
calculator is used to calculate gain compression. Throughout the simulation, the IF
amplifier gain is set to 10dB. Fig.5 and Fig. 6 show the performance curve of the
simulation. Fig.7 illustrates the power at output port in dBm versus the input base
band power variation. Finally, the sweep variable parameters in sweep2 are
changed to IFAGC, which is the gain of the IF amplifier. The system gains
performance versus IF gain variation is presented in Fig. 8.

Conclusions

The RF system for smart antenna transmitter presented in this application

notes is a typical communication up convention channel. All the components work
in large signal condition. The main tasks of the system simulation are (1) to ensure
the system has enough linear output power, (2) to find out the system AC

40 University of Bridgeport
parameter limits, and (3) to tune the component parameter to keep the system in
the best working status. The system simulated is an experimental transmitter, in
which many practical factors are not considered, such as the fast frequency
hopping and large amount user etc. More simulations, such as ACPR
characteristics, PLL time domain and frequency domain responding characteristics,
and inter-modulation characteristics, etc., should also be carried out in designing
commercial products.

Figure.1 System Schematic

41 University of Bridgeport
Figure.2 Simulation Tools used in ADS

42 University of Bridgeport
43 University of Bridgeport
 6. References

1. Portable ECG Logger by Jessica Lambourn

2. System Simulation of a RF Transmitter System for Smart Antenna using Agilent

Advanced Design System by Huang Yong.

44 University of Bridgeport