AnsweredAssumed Answered

Error correction when ports are highly coupled

Question asked by jeffb on Oct 30, 2008
Latest reply on Nov 5, 2008 by dgun
I'm looking for some advice on calibration when measuring through microstrip conductors that are highly coupled. The conductors in question are members of a 100 ohm differential pair. For two port calibration, I used the OSL/unknown thru technique, with the non-measured trace load-terminated during reflection cal.

After calibration, when a short circuit from each trace to ground is measured in differential mode (Sdd11), instead of the desired 0 dB it is more like +3 dB at 500 MHz. Diff pair line lenth is about 1 inch.

Is it likely that 12 term calibration is inadequate with so much 'leakage' between the port conductors, and that 16 term cal would work better?

We're using a 5071C for these measurements. Does that instrument support 16 term cal, or would I need to implement the cal via post-processing?

Any comments are welcome ... thanks.  

Outcomes