This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Amplifiers linear and nonlinear parameters can be measured using Network and
Spectrum analyzers. For noise figure measurements, either a spectrum analyzer or Noise
Figure meter is used.
To make the measurements and interpreting the results correctly, it is important to
understand the theory of operation of these devices. This chapter gives a brief
introduction to the major components inside the measurement instruments and why
these components are important will be examined.
2.1 Spectrum Analyzer
The spectrum analyzer is used for frequency domain measurements. The spectrum
displayed on the screen of the analyzer contains the power corresponding to each
harmonic component of the signal. This information can be used for modulation,
harmonic distortion, output power, occupied bandwidth, carrier-to-noise ratio, and a
host of other measurements.
Frequency domain measurements can be made using Fourier Analyzer and swept-tuned
analyzer. The Fourier Analyzer digitizes the input signal and then performs themathematics to display the spectrum. However they have limitations in the areas of
frequency range, sensitivity and dynamic range. Swept-tuned Super-heterodyne is the
most common type of spectrum analyzer, which plots the spectrum of the signal by
translating the signal above the audio range. It has the advantage of making the
measurements over a large dynamic range and a wide frequency range. Figure-1 depicts
the block diagram of the swept-tuned spectrum analyzer.
In the analyzer, the input signal travels through an attenuator to limit the amplitude at
the mixer, and then is filtered by a low pass filter to remove the undesirable frequency
components. Then it gets mixed with the signal generated by the local oscillator (LO).The frequency of LO is controlled by the sweep generator, DC voltage ramp generator in
this case. As the frequency of LO changes, the signal at the output of the mixer gets
filtered by the resolution bandwidth filter (IF filter) and amplified by the IF amplifier.
An envelope detector, following the IF amplifier, then rectifies the signal producing a
DC voltage which is used to drive the vertical portion of the display. As the sweep
generator sweeps the frequency, a trace is drawn across the screen which shows the
Envelope Detector is used to convert the IF-signal to baseband or video signal so that it
can be viewed on the screen. The output of envelope detector is digitized with an ADC
which is then represented as the signal amplitude on the vertical axis of the display.
2.1.4 Video Filter
Video Filter is a low pass filter located after the detector and before ADC. The filter
determines the bandwidth of the video amplifier and is used to average or smooth the
trace seen on the screen. Spectrum analyzer displays signal plus noise. By changing the
VBW settings, we can decrease the peak-to-peak variation of noise [2]. With small valueof VBW, we can identify a signal embedded in noise floor. A general Rule of thumb is
VBW = RBW/100
2.2 Network Analyzer
Network analyzer is used to measure both linear and nonlinear behavior of the devices.
There are two types of network analyzers: vector network analyzer (VNA) and scalar
network analyzer (SNA). VNA can measure frequencies from 5Hz to 110GHz and can
measure the S-parameters, magnitude and phase, standing wave ratios, gain,
attenuation, group delay, return loss, reflection coefficient and gain compression [3]. A
scalar network analyzer provides fast and economical measurements of many
amplifiers. SNA can measure only the amplitude portion of the S-parameters, resulting
in measurements like transmission gain and loss, return loss and standing wave ratio.
Figure-2 shows a generalized block diagram of a network analyzer. The analyzer has an
RF signal source that produces an incident signal which is used as stimulus to the DUT.
The device responds by reflecting a portion of the incident signal and transmitting the
remaining signal. The transmitted and reflected signals are measured by comparison to
the incident signal. The network analyzer couples off a portion of the incident signal
which is used as reference signal. It then sweeps the source frequencies, resulting in a
2- Test set also separates the incident and reflected signals at the input of the DUT.
Again, couplers are used for this purpose as they have low loss and high reverse
isolation.
2.2.3 Detector
The next portion is signal detection device. The signal detection can be done using a
diode detector or tuned receiver. Diode detector converts the RF signal level to a
proportional dc level but the phase information of the RF carrier is lost in this case.
Modern network analyzer use the tuned receiver approach. The tuned receiver uses a
LO to down convert the signal from RF to intermediate (IF) frequency. The IF signal is
later bandpass filtered that narrows the bandwidth of the receiver and greatly improves
the sensitivity.
2.2.4 Processor/Display
This block process the reflection and transmission data and display the results. Most
network analyzers have similar features such as linear and logarithmic sweeps, linear
and log formats, polar plot and smith chart, etc.
2.3 Noise Figure Meter
Noise figure of an amplifier characterizes its ability to process low level signals. It is a
key parameter that differentiates an amplifier from another. The noise figure of an
amplifier can be measured using a spectrum analyzer or with a Noise Figure (NF) meter.
Noise figure measurement with the spectrum analyzer uses the Gain method and is
useful for measuring very high noise figure. But it will not be discussed in this report as
the spectrum analyzers needed for making noise figure measurement should be able to
provide very high resolution bandwidth and noise floor in the order -130dBm [4]. Most
of the noise figure measurements are done using a noise figure meter, which gives
accurate results for small NF measurement. Many types of noise figure measurement
equipment uses Y factor method, which will be illustrated later in the report.
A block diagram of the HP 8970A noise figure meter is shown in Figure-3. Themicroprocessor controls the input and IF attenuators and the first LO, reads the analog-to-digital
converter, provides output data for digital storage circuits, and turns the noise source on and off
This chapter contains the methods for making the measurements of an amplifier. Bothlinear and nonlinear measurements are discussed in detail. Network analyzers are
traditionally used for making both linear and nonlinear measurements, while some of the
nonlinear measurements are done using a spectrum analyzer
3.1 Linear Measurements
Linear measurements include the determination of gain, reverse isolation, return loss and
noise figure.
3.1.1
Gain MeasurementAmplifier small signal gain is the gain in linear region and is defined as the ratio of the
amplifier’s output power delivered to the load Z0 to the input power delivered from a Z0
source, where Z0 is the characteristic impedance in which the amplifier is used [3]. The
gain can be expressed as follows:
( ) ( ) ( )out in
Gain dB P dBm P dBm
Before the measurements started, it is important to know the input and output power
levels of the amplifier under test (AUT) and the type of calibration required. Given the
approximate small signal gain and output power at 1dB compression, the input power
level can be estimated with the following formula.
1( ) ( ) ( ) 10in dBcompressionP dBm P dBm Gain dB dB
It is also important to estimate the output power from the AUT to avoid overdriving or
damaging the test port of the network analyzer.
Once the appropriate measurement parameters are selected, the next step is to perform
two port calibration of the network analyzer. If the attenuators are used at the output of
AUT they should also be included in calibration. If any change is made in themeasurement setup, it is necessary to calibrate the network analyzer again.
The block diagram of the measurement setup for small signal gain is shown in Figure-4.
The gain is measured at constant input power over a swept frequency range and it is
enough to measure the S21. The result can be displayed on a logarithmic scale or on a
smith chart. Figure-5 depicts the gain measured with a network analyzer. Markers can
be used to determine the gain at a particular frequency.
Figure-4: Basic Setup for amplifier measurement [3]
Figure-5: Gain Measurement [3]
3.1.2 Return Loss Measurement
Return loss is a measure of the quality of the match of the input and output of the
amplifier, relative to system impedance [3]. Reflection coefficient contains the
magnitude and phase of the reflected signal. Return loss considers only the magnitudeof the reflected signal. The formula for the reflection and return loss are as follows:
20log( )
reflected
incident
V
V
RL
Measurement setup shown in Figure-4 can be used for return loss. The network analyzer
feeds RF power to port 1 of the AUT and measures the reflection at the same port. Plot
the S11 versus frequency and using the formula 20log(S11) gives the input return loss.
Figure-9: Amplifier 1dB compression point measurement [3]
3.2.2 Harmonic and Intermodulation Product Measurement
Harmonic and intermodulation products of an amplifier can be measured using a
spectrum analyzer. When the input signal is small, the amplifier is operating in the
linear region and harmonic components are almost non-existent. But increasing the
input signal level drives the amplifier to the nonlinear region and products of second
and third order of fundamental frequency are generated. The generation of harmonic
components reduces the gain of the amplifier as some of the output power at the
fundamental frequency is shifted to the harmonics. Harmonic products can be
visualized on a spectrum analyzer by selecting appropriate span.
Intermodulation distortion occurs when several signals are applied to an AUT when
it is operating in nonlinear region. Third order intermodulation products are most
problematic as they have frequencies close to the useful signal. Figure-10 shows how
the levels of useful signal and intermodulation products are related.
Figure-11 shows the measurement setup that consists of two signal generators, a
coupler to combine two signals and the spectrum analyzer. The frequency of the twosignals should be close to the centre frequency of the amplifier and the amplitude
should be equal. Select the appropriate reference level and attenuation, if needed. The
resolution bandwidth of the spectrum analyzer, in this case, must be carefully chosen
so that the signals can be separated in frequency domain. Figure-12 depicts the
intermodulation products of an amplifier whose centre frequency is 1.5GHz and the
frequency of two tones is 1.49995GHz and 1.50005GHz. 1MHz span in this case
would be sufficient to analyze the frequency products.
Span: It is the range of frequencies that are displayed on the screen of a SpectrumAnalyzer or Network Analyzer.
Resolution Bandwidth: It is the bandwidth of the internal intermediate frequency filterof the Spectrum Analyzer. It is the minimum resolution that a Spectrum Analyzer canhave in terms of spacing between adjacent peaks. Therefore, two peaks will appeardistinct in a Spectrum Analyzer only if their distance on frequency scale is greater thanthe resolution bandwidth.
Video Bandwidth: It is the bandwidth of the low pass filter located after the detectorand before ADC. It is used to average or smooth the trace seen on the screen.
Sweep Time: When we narrow the resolution bandwidth, we must consider the time ittakes to sweep through them. Narrower bandwidth requires a long time. When sweeptime is very short, the RBW filter cannot fully respond, and displayed response becomesun-calibrated both in amplitude and frequency [2].
Reference Level: It is used to describe how much level is present in dB above or belowthis reference.
Attenuation: refers to the attenuation, and used to attenuate the signal. The attenuatedsignal is required at the input of mixer as if the amplitude is too high it could createdistortion and may damage the mixer.
Measurement Un-calibration: is related to sweep time. When the sweep time is veryshort, the RBW filter cannot fully respond both in amplitude and frequency due to thedelay in filter and measurement becomes un-calibrated. If MEASU UNCAL warningappears, try to change the sweep time or use wider filter.
Network Analyzer:
Calibration: Before we can actually use the VNA, we need to calibrate it. We can
calibrate the VNA using 4 knows standard terminations: Open, Short, 50ohms Matchedand through. Open the calibration menu on the VNA and it will progressively ask toconnect one of the terminations to the RF OUT port and then calibrate the losses of theRF cables. For the Through calibration, we connect the two RF cables to each other usingthe through calibration standard. We may calibrate both terminals of the VNA. Thepurpose of calibration can be explained from the following block diagram.
For the user, the plane for RF measurements is at the terminals of AUT, however, for the
VNA, the plane for measurements is at its own terminals. In other words, in the default
case, neither the user nor the VNA is taking into account the RF cables (and the losses or
attenuation associated with them) connecting the AUT to the VNA. These cable losses
can considerably alter the RF measurement results and hence, to move the plane of
measurement for the VNA beyond the cables, we perform calibration. Now that the
cable losses are stored in the memory of the VNA, this set-up should not be disturbed,
meaning, the same RF cables, attenuators, connectors and adapters have to be used for
the entire measurement.
Directional Coupler: It is a transmission coupling device for separately sampling(through a coupling loss) either the incident or the reflected wave in a transmission line.
Splitter: Splitters are used to divide the input signal into two or more output signals.The performance of a splitter is evaluated according to loss and isolation. Loss is theamount of attenuation that a signal receives as it passes through input to output.Isolation means that the two input signals do not mix up.