Experiment 6

CHARACTERISTICS OF A MICROWAVE AMPLIFIER

Experiment 6

Objective:
 To study the small signal S-parameters of an amplifier.
 To understand different classes of amplifiers.
 To study the linearity and power added efficiencies of a class A linear
amplifier and a class AB power amplifier.

(a) (b)
Fig. 1 Photographs of a (a) class A linear amplifier designed for maximum gain, (b) class
AB power amplifier using the same transistor.

(a) (b) (c)

Fig. 2 Measured (a) gain vs. input RF power, (b) low power gain, and (c) PAE vs. input
RF power of a class A amplifier implemented for maximum gain at 2.4 GHz.

Theory:
An RF amplifier is utilized to amplify the input RF signal. Fig. 1 illustrates two RF
amplifiers operating in class A and class AB modes. The first amplifier is designed for
maximum gain, while the second amplifier employs the same transistor but is optimized
for maximum efficiency.
The required characteristics of an RF amplifier vary widely depending on its
application. In an RF receiver chain, the first component after the antenna must amplify a
very weak signal without significantly degrading its signal-to-noise ratio. An amplifier
increases the power of both the signal and the noise present at its input. To minimize
additional noise and maintain good overall noise performance in the receiver chain
(measured as the noise figure), the first component, known as the low noise amplifier
(LNA), is designed to minimize noise generation. LNAs mostly operate in class A, with
the transistor always ON.
On the other hand, in an RF transmitter, the last component just before the antenna
must generate high power without significant signal distortion. As a result, this amplifier
operates at the 1 dB compression point, which is the universally agreed boundary
between the linear and nonlinear regions of operation. This type of amplifier is called a
power amplifier. RF designers aim to extract the maximum possible RF power without
damaging the device, making the power amplifier have the highest DC power dissipation
of any RF component. To minimize heat generation and DC power consumption, the DC
to RF power conversion efficiency must be maximized, measured as power-added
efficiency (PAE). PAE varies with amplifier classes and input RF power. Higher classes,
like class C and class F operations, provide higher PAE but at the cost of higher
nonlinearity since the transistor remains OFF for most of the RF cycle. There is a
common trade-off in amplifier design between efficiency and linearity.
Sometimes, additional linear RF amplifiers, mostly designed for maximum gain
and operating in class A, are used as gain blocks. Fig. 2 displays the measured
characteristics of an amplifier working in class A mode. During the implementation of an
RF amplifier, the overall circuit must be checked for stability, otherwise, an amplifier
may become an oscillator, resulting in much lower gain than expected and generating one
or more arbitrary RF frequencies.

Fig. 3 Conduction angle and efficiency of various classes of amplifiers.

Different classes of amplifiers:
RF amplifiers are divided into several classes, based on how much of the cycle of
the sinusoidal signal the transistor is conducting current. Most popular classes are class
A, and class AB, which are considered the linear amplifier classes in which the active
device is used as a controlled current source while class C is a nonlinear class in which
the active device is used almost like a switch. The bias at the input of the active device
determines the class of the amplifier. Fig. 3 shows how the PAE varies with the
conduction angle.
Noise and Noise Figure (NF):
Noise power is a result of random processes such as the voltage generated between
the terminals of a device or component due to the random motions of charges or charge
carriers in the device. All components in a microwave system generate noise. Noise also
enters from external source. Among different types of noise, the thermal noise due to the
thermal vibration of bound charges at a temperature above absolute zero is the most basic
source of noise. The noise level of a system sets the lower limit on the strength of a signal
that can be detected in the presence of the noise.
The thermal noise is modeled with a Thevenin equivalent circuit consisting of a
noiseless resistor R and a generator with a rms voltage vn = √(4kTBR), where k = 1.380 ×
10−23 J/K is the Boltzmann’s constant, T = the temperature in degrees kelvin (K), B = the
bandwidth of the system in Hz. Then, maximum power delivered from the source to a
matched load is kTB, independent of the resistance value. An antenna when connected to
a LNA delivers signal as well as noise. Noise is also generated inside the LNA. Thus,
input SNR of a LNA is more than the output SNR.
Noise figure (NF) and noise factor are measures of degradation of the SNR caused
by a component. The noise factor is defined as the ratio of actual output noise to that
which would remain if the device itself did not introduce noise, i.e.
Input SNR
Noise factor  F  = .
Output SNR
While measuring the input SNR, the input noise is attributed to thermal noise produced at
a standard noise temperature T0 = 290 K. Noise figure is the noise factor in decibel scale
i.e. Noise figure  NF  = 10 log10  F   SNRin  dB   SNRout  dB  .

Power added Efficiency (PAE):

The PAE is defined as the ratio of RF output power with addition of RF input power to
the DC input power of a power amplifier. It denotes the percentage of DC power
converted to effective AC power by an amplifier. As a mathematical expression, it can be
defined as
PAE = (PRFout - PRFin)/ PDC,
where PDC is the DC power consumed by the amplifier.

P1dB:
P1dB or 1 dB compression point is a measure of a device's ability to withstand
and maintain a certain level of signal power without distortion. It is defined as the output
RF power level at which the gain decreases 1 dB from its constant value or small signal
value. Beyond this point, any increase in input power causes a greater reduction in output
power. Devices with a higher P1dB are generally considered to be better suited for high-
power applications. Above this point, the power lost in the harmonics increases, and the
waveform deviates significantly from the input waveform shape.
Know your amplifier:
Absolute maximum rating of the transistor BFP650:
 Device current ICC = 150 mA, IBB = 10 mA
 Device voltage VCE (+V) = 4 V
 Total power dissipation at TS = 120°C, Ptot = 500 mW.
 Junction temperature Tj = 150°C.
DC operating points for the implemented amplifiers @ 2.4 GHz:
 Class A mode: ICC = 70 mA, IBB = 400 µA, and VCEQ = 2.9 V, VBEQ = 0.75 V
 Class AB mode: ICC = 70 mA, IBB = 370 µA, and VCEQ = 3.9 V, VBEQ = 0.85 V

Procedure:
Identify the transistor, input and output matching circuits, input and output coupling
capacitors, the bias Tee (used for RF-DC isolation) and the SMD resistor RC connected
between the bias Tee and the bias point VCC. Connect the circuit and apply a biasing
voltage of VCC = 5 V (absolute maximum). Wait for 10-15 minutes before turning ON the
RF source.
Scattering parameters:
1. Set the VNA power level to -20 dBm or “low level” (Caution: An amplifier
provides gain. Therefore, |S21| dB is positive. The VNA may be damaged if you
don’t set the VNA power to a low level).
2. Calibrate the VNA over 1.0-4.0 GHz.
3. Connect the amplifier to VNA.
4. Switch on D.C. supply of the amplifier and wait for 10 min.
5. Measure and plot the S-parameters (|S11|, |S22|, |S21|, and ∟S21) over 2.0-2.8 GHz.
P1 dB compression point and PAE:
1. Connect the amplifier input to a RF source (Signal Generator).
2. Connect the amplifier output to a Spectrum Analyzer.
3. Set the signal frequency at 2.4 GHz or as suggested.
4. Switch on the amplifier D.C. power supply and wait for 10 min.
5. Vary the input power level PRFin from -80 dBm to +10 dBm or as suggested. Note
down both input and output powers for 10-15 steps.
6. Also using a multimeter, measure and note down the DC voltage drop VR across
RC for each and every value of PRFin. Then, VCEQ = VCC – VR, and ICQ = VR /RC.
Now, use the relationship for the PAE. Note that the DC power dissipation
ICQ×VCEQ does not vary with the RF cycle but the PRFin.
7. Plot the gain variation with PRFin and compare the results with the gain obtained
from previous step. (Caution: Do not increase the input power level
unnecessarily, otherwise you will burn the device).
NF measurement (approximate method):
1. Fix the centre frequency of measurement as 2.4 GHz. Set the resolution
bandwidth of the Spectrum Analyzer to minimum available value (1 Hz), the
frequency span between 1 and 10 kHz.
2. The output of the LNA is connected to the input of the spectrum analyzer with its
input terminal terminated with a matched load.
3. Measure the output noise power (PNOFF dBm) by using the averaging option in the
Spectrum Analyzer without the D.C. bias*.
4. Now, switch on the D.C. bias and again measure the noise power (PNON dBm).
5. Then, NF = PNON -PNOFF dB.
6. Calculate the theoretical noise power as below and compare it with the measured
PNOFF.
PNout  174 dBm/Hz  10 log10  B.W.  Gain.
*N.B.: The averaging option may not work if the sensitivity of the spectrum
analyser is not below the calculated PNout. Then, collect the measured data and
obtain the averaging in MATLAB. Accurate measurement of NF is possible by
using a calibrated noise source.