China Electronics Technology Instruments AV3672 Series, User Manual

Page: 192 / 424

6 Measurement Optimization
6.4 Improvement of Measurement Results of Long Device of Each Point
188
Fig. 6.5 Influence of Sweep Averaging on Measurement Results
Prompt
Reduction of random noise
➢
The average value of measurement can be independently set for each channel.
➢
The total measurement time increases along with the increase of averaging times.
➢
General noise can be minimized by setting the average value and intermediate frequency bandwidth.
➢
In order to minimize the low-frequency noise, measurement averaging is more effective than reduction of the
system bandwidth.
b) Setting of sweep average value
Menu path: [Response] > [Avg] . Set the average value by the [Average] key in the auxiliary menu bar or in the
[Average] dialog box. Enable or disable the averaging function by [Average on/OFF] . Directly enter the value or
press the arrow button to increase/decrease the averaging times according to the average factor .
Attention
Select [Restart Average] to start a group of new average sweep.
6.4 Improvement of Measurement Results of Long Device of Each Point
The frequency of the signal from the DUT may be different from the frequency of the signal into the DUT at the
given time. This may result in inaccurate measurement results sometimes. The following contents can help to
understand why inaccurate measurement results are obtained and how to compensate.
Why are inaccurate results produced as a result of delay?
➢
The source and receiver can be locked at the same time, and sweep can be done for one frequency interval
with the vector network analyzer.
➢
The signal through the DUT (DUT) represents different colors corresponding to different frequencies.
➢
If the stimulus frequency passes through the DUT and the analyzer is tuned to new frequency before the signal
reaches the receiver, inaccurate measurement results will be produced.
As shown in Fig. 6.6, if a long cable is measured by the analyzer, the signal frequency of the cable end will lag
behind the source frequency of the analyzer. If the frequency offset (typically dozens of kHz) is obvious relative to
the intermediate frequency test bandwidth of the network analyzer, measurement results may be wrong due to the
roll-down characteristics of the intermediate frequency filter.