Control-Software for HP 8510C VNA

[DL8RI]

Hi Everybody,

I got an 8510C some time ago.
What annoyed me, is the missing "Parameter over Time" measurement.
So, for example the ability to watch the Reflection at 10 GHz on Zero-Span and then plot this over time.

Interesting if you have rather slow settling effects like thermal or whatever.

So I started a few days ago to make a Software for that. I am going to do this in Python 3.

So far, I got a driver for the 8510C working with all functions I need ATM (Fstart, Fstop, Span, Markers and so on), this will grow over time.
My software has now the basic functionality to create a plot of a given Value on a given Frequency at a given sample-rate (e.g. every 1 s) for a given time.

I intend to let this grow into a more advanced controller... Planned features so far:
- I want the computer to output a beep which has a tune corresponding how close you are to some target-value. This is for tunig parts, so you just need to listen to the beep while turning some screws.
- I want to install a IFFT-Method, as I do not have the TD-Option in my unit and modern days computers a way more capable of the calculations.
- Maybe a 3D-style sweeping, so whole S-parameter-plots over time.
- It came to  my mind to give the 8510C more Points. This would work as follows: You do a whole SOLT-Cal for each band (e.g. 5-10, 10-15, 15-20) and safe them as Cal state 1, 2, 3. the software then controls the VNA to load the Cal-Data, make the sweep and then adds everything into one S-Paramter from 5-20 GHz with 3 times more data-points). Not sure about this one, if it's worth the hassle.

Any suggestion for additional features is welcome. Once the code has come to a "non embarrassing" maturity, it will be published on my homepage for everybody to use.

[hagster]

Before you get too much deeper into this take a look at this free software that will probably do what you want.

http://www.ke5fx.com/gpib/ssm.htm

[KE5FX]

You might be thinking of VNA.EXE:

http://www.ke5fx.com/gpib/readme.htm#VNA

It would be neat to run 50 GHz sweeps in SSM on the 8510C, but I don't think anyone would be happy with the spurious responses.

VNA.EXE won't control the network analyzer, but it is almost indispensible for saving .S1P and .S2P files and storing/recalling calibration setups.

[DL8RI]

VNA.EXE won't control the network analyzer, but it is almost indispensible for saving .S1P and .S2P files and storing/recalling calibration setups.

Yes.
I'm using your toolkit for quite some time, ever since I got my HP 3561A... I believe since 2010. Especially the plotter-emulation is my best friend since it works with more or less everything.
So thank you very much for your work there.

Because of "Christmas-Things" happening around me, I can't work on it at the moment. I have a running first version, screenshots and some test-measurements will come after the 26th.

[KE5FX]

VNA.EXE won't control the network analyzer, but it is almost indispensible for saving .S1P and .S2P files and storing/recalling calibration setups.

Yes.
I'm using your toolkit for quite some time, ever since I got my HP 3561A... I believe since 2010. Especially the plotter-emulation is my best friend since it works with more or less everything.
So thank you very much for your work there.

Because of "Christmas-Things" happening around me, I can't work on it at the moment. I have a running first version, screenshots and some test-measurements will come after the 26th.

You're very welcome!  If you make some progress toward a complete control program, let me know and I'll add a link to the readme file for the toolkit.

It is possible to emulate the time-domain security key with a relatively cheap FPGA board (Mojo V3), and I would definitely recommend doing that if yours doesn't have the key.  It's amazingly useful.  If you manage to emulate it on the PC, you'll probably want to be able to compare the results.

[DL8RI]

As promised, now that Christmas is finally over, some more input.
Setup:
I used a switchable attenuator and measured S21. So nothing too complicated.


Here you see the recording of the value at 4 GHz, sampled every second and for 30 seconds.
After a few seconds I did switch the attenuator to higher attenuation. Obviously it's of very high precision at 4 GHz.


The software is made as such:

If you start the measurement, it will set the center-frequency of the VNA to the frequency you did set, span is set to zero and a marker is put onto the frequency chosen.
Then the value is read every interval, you did chose. This does not work perfectly as the time until data-transfer is giving a significant delay. Instead of 5 data-points every 5 seconds, you get around 4.
I will check on methods to speed this up, but there seems to be limits inside the 8510 on answering data-request.

Next steps will be:
- Include calibrated measurements (-> Load appropriate cal-set)
- Polish the overall usability (view-options, saving of data, persistent config and so on) <- That's a lot of work behind the scenes
- I did start on this beeper, that needs to be finished.

Once that is done, I will publish a first version (hopefully I manage in the next days). And then start to work on new features.

It is possible to emulate the time-domain security key with a relatively cheap FPGA board (Mojo V3), and I would definitely recommend doing that if yours doesn't have the key.  It's amazingly useful.  If you manage to emulate it on the PC, you'll probably want to be able to compare the results.

Thanks. Yes I saw this procedure on KO4BB's site. The "security"-Boards including the Option-PAL are also available on eBay once in a while.
I'm not so sure about the compatibility with my firmware-version. Maybe I get one of these FPGAs, but I'm rather lazy atm, as my other "low-freq" VNA up to 1.3 GHz can do this and for finding faults in cables the unit also okay.
But I guess, even if it's just for the sake of completeness, I will do it some time in the future
Will have to dig into the 8510 anyway to change the lithium-battery sometime 2019.

[eb3frn]

Hi,

somebody know if the power domain option is possible to activate from the pal?

thank

73! Iban
eb3frn

[jjoonathan]

Just a few days ago I put an enhanced 8510 driver into scikit-rf:

https://github.com/scikit-rf/scikit-rf/pull/591/commits/da4aeee29e506273897263e2d595e429c0b06282

It doesn't support the power domain or know how to bypass the PAL, but it does know the line format, how to speak the binary protocol for fast transfers, how to chunk sweeps to work around the 801 point limit, etc. If you're looking to script this in python, it would help.

Alternatively, I believe the 8510 achieves phase lock using the a parameter on the outgoing leg, so if you wanted to tap off a parameter from an incoming leg, you could. The output of the hexagonal mixer modules is well filtered, so I'm pretty sure you really could get away with just having a detector (see attached). Get that into a scope trace, take SWEEP OUTPUT from the source into another scope trace, and there you go. You'd have to calibrate it yourself, but if you're in the power domain, there's a very real chance you want do that anyway.

[eb3frn]

Hi

Wooooow, very interesting. It can be useful for me. I'm building a custom setup to measure noise figure using the cold source method. With the 8510c I' will measure the gain and s-parameters and I'm building a nose receiver to get the power. I wish to automate it and try to use the most parts of the 8510c to do it.

You have some example about how manage the sweeps? My idea es do a sweep then switch to the noise receiver.

73! and HNY Iban
eb3frn

[jjoonathan]

Here's a bit of code in the wild: https://github.com/jjoonathan/F3560R-YIG-Filter/blob/main/F3560R%20Data%20Gathering.ipynb

The idea is that it uses a GPP-4323 power supply to set the current through a YIG filter and then it takes a sweep. Here is the 8510 relevant code:

Code: [Select]

hp8510c = skrf.vi.vna.HP8510C('TCPIP::ad007-right.lan::gpib0,16::INSTR')   # Open Connection
hp8510c.set_frequency_ramp(start_hz, stop_hz)                              # Set up sweep
skrf_network = hp8510c.s21                                                 # Perform sweep, return results as a scikit-rf network

It uses "ramp" mode to take a fast sweep, but the source is only locked at the beginning of the sweep. Step mode lets you take a sweep that phase locks at every point:

Code: [Select]

hp8510c = skrf.vi.vna.HP8510C('TCPIP::ad007-right.lan::gpib0,16::INSTR')   # Open Connection
hp8510c.set_frequency_step(start_hz, stop_hz, npoint=10001)                # Set up sweep
skrf_network = hp8510c.s21                                                 # Perform sweep, return results as a scikit-rf network


[jjoonathan]

You'll see that I use a GPIB->Ethernet adapter and pyvisa-py. This means no VISA drivers to install and it means that the code would run just fine on a raspberry pi or something, which would make it convenient to twiddle GPIOs, SPI, etc.