TRL with Keysight E5080B

[Ziad Hatab]

Hi!

So, recently I got access to Keysight 14GHz ENA (E5080B). I have a simple microstrip TRL design. I gathered raw measurements (with switch terms) and did a first-tier TRL calibration offline in python. So, my problem is that I get a lot of oscillation after calibration (which are the standing waves of the coaxial cables, which are supposedly calibrated by TRL). The plot below is the attenuation per-unit-length (alpha).

From my experience in (m)TRL calibration, such phenomenon happens when switch terms are not considered in the calibration (which I did consider). So, in the plot above I also ignored the switch terms and I still get the same result. So, my question, does any of you have experience with Keysight ENA? Are there any settings that I should change? I made sure that no previous calibration was active. My colleagues use this VNA and they do SOLT calibration just fine without any problems (they do it online using the instrument software). 

I will do some checking in the lab tomorrow, but I thought maybe someone already experienced such thing, or maybe with a different VNA vendor?

You can try my TRL data with code: https://cloud.tugraz.at/index.php/s/jGmS99TeL3EXMLL 

From my experience, I do on-wafer mTRL with Anritsu VectorStar. You can see in the plot below the difference when switch term is considered or not:

Thanks!

Ziad

[Ziad Hatab]

Ok, after few hours in the lab, I figured the cause of the problem and now TRL calibration works as expected. The problem was a setting in the VNA. It turned out that the factory calibration is on by default, even if the correction option is off. So, the solution is to turn the factory cal off.

Ziad