Vector network analyzer from rigol ds1054z

[freebil]

Hello. Is it possible to have the functionality of a network vector analyzer using a rigol ds1054z connecting to a pc writing code in matlab? So, is it possible to see vna functionality from matlab?

[balu]

Do you think about a configuration like this - see picture?

[cncjerry]

You can get scalar results that way but no phase, or at least, no accurate phase.

You can do this more simply by using the scope's FFT functionality with a broadband noise source.  You then put the scope in peak hold mode and use the noise source as the sweep generator.

Also, a directional coupler would be replaced with a reflection bridge which will give you SWR, definitely showing the tuning.  The coupler will also show the tuning but I never thought about the math required to get actual SWR using one.

[hendorog]

You can get scalar results that way but no phase, or at least, no accurate phase.

Also, a directional coupler would be replaced with a reflection bridge which will give you SWR, definitely showing the tuning.  The coupler will also show the tuning but I never thought about the math required to get actual SWR using one.

I think balu's design would work fine for both magnitude and phase - within the limitations of the device of course. Just need to zero out any offsets from the reference channel, and then any change in phase or magnitude has to be caused by the DUT.
Since the data will be in matlab then other more powerful calibrations can be done as well.

AFAIK there would be no difference between a coupler and a bridge, the maths is the same.

[cncjerry]

I was looking at Balu's method more closely and yes, I see now he is splitting the signal into two channels and you could see the phase delta between them.  I'll have to try that and compare it to a vna.

[pascal_sweden]

Is there software available for the setup described by Balu?

Does it support Siglent SDG1050 or SDG2040X as signal generator and DS1054Z as scope?

[JohnPi]

I posted this in another thread - https://github.com/jp3141/Vector-Network-Analyzer

https://www.eevblog.com/forum/testgear/lcr-meter-accuracy/msg1796498/#msg1796498

Here's the writeup:
Use a Rigol DS1054Z Oscilloscope and Siglent SDG1025 Function Generator as a Vector Network Analyzer (VNA) to plot transfer functions or impedance vs. frequency.

This is a simple vector network analyzer using a function generator and an oscilloscope. A vector network analyzer measures the ratio between two signals -- both the amplitude and phase shift. The forcing signal is connected to Channel 1 of the oscilloscope; you can use a probe if desired. The sensed signal is connected to Channel 2 of the oscilloscope. A probe can be used here too if desired. The sync signal from the signal generator is connected directly to Channel 4 of the oscilloscope. On the SDG1025, this cannot be used above 2 MHz, so the program switches to triggering directly from Channel 1 above this frequency. This has the added benefit of allowing 500 MS/s sample rate for the oscilloscope. The program has a number of command line options:

python VNA.py -b BEGIN_Freq -e END_Freq <-p Points_Per_Decade> <-f LOGFILE_Prefix>

The program takes logarithmically-spaced steps from BEGIN to END frequencies. The nominal signal amplitude is 1 Vpp. Some options are:

  -s StepSize -- use linear steps instead of logarithmic scaling of frequency steps
  -l -- just list frequency steps to be taken; don't actually run
  -n -- don't plot the data. Program exits at the end of the run
  -f FILENAME_Prefix -- the prefix of the logfile with data in a CSV format with # comments
  -p points -- number of sample points per decade
  -z resistance -- resistance of current-sampling resistor. Plot then shows Z=(Channel2)/(Channel1-Channel2)*Z which is the complex Z of the structure under test. Note there is no calibration or deembeeding of the test fixture or probes
  -v voltage -- set the amplitude of the generator's output voltage
  -b Freq -- begin freqency of sweep
  -e Freq -- end frequency of sweep
  -h -- help & exit
Since the signals will cover a wide dynamic range (over 60 dB), the program will dynamically adjust the 'scope vertical scale on each channel to optimally use the range available. The Rigol scope can use arbitrary vertical scales (i.e. not restricted to a 1:2:5 pattern).

This VNA can be used from 1 Hz to 25 MHz, although resolution suffers at above 10 MHz. It can be used to characterize components, filters, and circuits such as voltage regulators where it can be used to measure loop gain and output impedance.

[JS]

Great! I should check that out!

I was thinking in some sort of VNA code for my DS1054Z, I guess that will come pretty handy!

JS

[RoGeorge]

Need to try it, thank you!