开云体育

Went to great lengths to create PyVISA program to read using the GPIB the x4 parameter S-parameter data from the HP8510C. This eventually would read consecutive 1701 point sweeps in a round 5 seconds using the HP binary data format; ASCII data formate took much longer. That was for uncalibrated data. However, when i tried to read TRL calibrated data it would take over two minutes to read the data. The obvious solution then would be to just read the uncalibrated data and perform the TRL calibration in the PC. I could trawl through endless papers to find a suitable mathematical equations to code up in Python. From what i gather, mathematically its quite complicated, solving something like 12 simultaneous equation, clearly doable, but i've also got a massive list of other stuff to do. Might anyone have an algorithm to perfor the TRL calibration, or know where i might find this?

?

many thanks, N ?

Hi , you can visit the scikit-rf?

In the past, I tried some features of this software tools, but I did not go beyond a simple plot of graphs. I remember that you can define a virtual VNA, and perform a calibration starting from the measurements of the devices of a cal kit, and then apply the calibration to the raw measurements of the device under test . Good documentation and examples, Python language
Regards?
Maurizio IZ1MDJ
?

Networks

The central object in?skrf?is a N-port microwave??object. A??can be created in a number of ways, one way is from data stored in a touchstone file"

Yes, thank you. They have a selection of different calibration algorithms. That's probably a really good place to start. Cheers, N

Out of curiosity, did you have any speed optimization enabled on your IEEE-488 interface?

?

What software version was your 8510C running?

Once calibration/correction constants have been computed externally on a computer, would it be possible to upload the computed correction constants to the 8510C so it can do calibrated measurements with this externally determined correction?

If so, how? This would allow correction schemes that the 8510C not originally allows.

?

Tom

Hi Ed, i did manage to get the NI GPIB / USB to laptop interface working. In terms of running the 8510C from the laptop it seems to work quite well but only with uncalibrated data. From the Operation and Programming Manual i tried all fives forms of data transfer: FORM1 to Form5. FORM1 was a little tricky and the manual doesnt really explain for format well. After trial and error all forms were giving the same numerical values, which gave some confidence that i had the programming correct. FORM5 was the fastest, collecting data for a 801 point sweep in about 6 seconds. Slow by today's standards, but not bad for vintage 1991.

?

However, when i ask it to do calibrated data, the machine peeps indicating there's an error and still continues to show data transfer increase from 6 seconds to 100 seconds, so clearly that's not much good.

?

How do i find which software version i have running. On the display i see the information: HP 8510C Monitor C.02.16 March 9 1991, so is C.02.16 the software version number?

I've been looking at scikit-rf to do the TRL calibration and it mentions using the 'switch terms' to improve the calibration. I'm just wondering, how do i read the switch terms from the HP8510C, which appear to be for source port 1: ratio a2/b2 and for source port 2: a1/b1?

?

During measurement i can get the s-parameters ok, but it's not clear how you can measured these ratios, or measure a1, b1, a2, b2 seperately and calculate the ratios in the PC. If i can get that right i'll be able to do all calibratoin in the PC, which would be much faster than in the HP8510C.

?

Thanks, N

However, when i run from the PC and file data, the first line in the s2p file is: !HP8510C.07.14: Aug 26 ?1998, so that indicates a different software version. (well the monitor is CRT, so pretty old, so perhaps that's got a different software version, or the number was not update when other software was updated.

?

?

I note the HP8510C driver, so i should be able to use this to access the 'switch terms. Also i'm interested to see how fast it reads the s-parameters! Thank you for this, N

I see the 'switch terms' can be obtain using so it'll be interesting to see how that work and how fast this driver reads the s-parameters, thank, N.

Hi jjoonthan,

?

In using at the line in first line in the program

?

vna = skrf.vi.vna.hp.HP8510C(address='TCPIP::ad007-right.lan::gpib0,16::INSTR', backend='@py')

?

i get the error: ? AttributeError: module 'skrf.vi.vna' has no attribute 'hp'

?

any idea what i'm doing wrong?

?

many thanks, N.

Fixed, thanks Jon.

I use used scikit-rf to do a TRL and a SOLT calibration in the PC, both giving very similar results. For the SOLT i substituted the a 'quarter-wave + short', for the 'O', making it a 'S', so it was really a SSLT, so i could do both calibrations in waveguide. It's good that this works as the HP8510C is just too slow to do this internally. A nice result.?