unexpected measurements from differential test fixture

Have a loaner differential test fixture for evaluation purposes. Made by picoseconds pulse labs. pspl 5310A. I was expecting to see a short on the VNA smith chart when I shorted the two leads, an open on the VNA smith chart when the leads were unconnected instead i saw the following:

I made sure my VNA was calibrated prior to hooking up the test fixture.

Is this what I should expect to see from this this differential test fixture? Looks broken to me.



75

open

short

not familiar with that exact model, but basically baluns made to perform differential tests have:
1) a lot of loss (maybe as high at 9 dB)
2) not necessarily a good impedance match on the differential port
3) There can be significant electrical length in the device.

The short/open measurement thru this device, as you show, is not at the smith chart edge. That is most likely due to the balun insertion loss.

Apart from minor deviations, the main problem seems to be that the VNA is not calibrated for the text fixture's reference plane. You should at least perfom a port extension to the correct delay, or better a full SOL calibration with text fixture.