AnsweredAssumed Answered

Transform Reference Shifting?

Question asked by mhreider on Jan 9, 2012
I know there are probably plenty of variables on the test system setup's end as to the cause of it, but I decided it'd be a good idea to see if maybe it was something internal on the VNA first.

The setup I'm using is an N5230c with an U3042AE12 12 port extender. I have calibrated the setup using an N4433A ECal kit from 10MHz Start to 20GHz stop. The equipment has each set of four ports calibrated (so, 1-4, 5-8, 9-12, 13-16 are all calibrated independently) and is then connected to an interface where the DUT is placed. Each set of four is run in balanced mode and we switch through each set over GPIB.

The issue I'm having is when we begin to use the Transform function. We take Sdd11 and Sdd22 and transform both using a Low-Pass Step. We then use the equation editor to capture the characteristic impedance (so, 100*(1+TrX)/(1-TrX) for a 100 ohm reference against rho) on the PNA's side. From the equation, we're assuming a reference impedance of 100 Ohms. However, as we test through I'll see occasional instances where the transform seems to "shift" downward about 3-5 ohms. Generally I've been able to get rid of this in the past by recalibrating the system, but I was wondering if there's something internal to the PNA that would cause this. I've attached a graph (since pictures speak 1000 words) as an example of what I'm seeing. The red lines are an Sdd11 and Sdd22 case of an instance where I've seen this happen. I could run the test again on the same DUT and get traces akin to the blue traces pictured. (X Scale is 0-1.5ns, Y is in ohms)

I've tried asking several other engineers if there was a decent explanation for this but to no avail. I would most likely guess that there's something in our setup that I'm overlooking, but I figured it was worth a shot to eliminate PNA itself as a culprit.

Any assistance with this would be greatly appreciated.