AnsweredAssumed Answered

What causes these "spikes" when measuring a short?

Question asked by drkirkby on Jan 8, 2015
Latest reply on Jan 13, 2015 by Dr_joel
Someone asked me to measure a Maury Microwave 3.5 mm short on my VNA. To make things as accurate as I could, I used a very narrow IF bandwidth (10 Hz), and put an NMD to 3.5 mm adapter on the test port, so there were no flexible cables. The results are puzzling me. 

If you look at the graph, there are some spikes which appear. I've used two very similar cal kits (85052B and 85033E) and get the same spikes in the same positions. If I re-measure the calibration short, I don't see these. The fact that |S11| > 0 dB is not a good sign, but I know that no instrument is perfect. 

I don't for one minute believe the short actually has these characteristics. 

I get virtually the same if I use test port cables. 

Any ideas what it could be? Any ideas if it can be resolved, or is this just a limitation I have to live with? 

The instrument was calibrated by Keysight last year, so is well within its calibration period. BUT there's one change I made to the VNA a few months ago, which *might* possibly be the reason for this. I have added option 012, which brings the A and B receiver inputs, and coupler outputs, to the front panel. So adding that involved removing a 2.4 mm cable between the port 1 coupler and the A receiver, and replacing it with a pair of 2.4 mm to SMA cables and an SMA-SMA jumper. Likewise on port 2/B-receiver.  Although the SMA-SMA jumper is the right Agilent part, the two 2.4 mm to SMA cables I had made up by a specialist cable manufacturer as Keysight did not have them. 

I've attached a .s1p file. There are tons of comments in the file, so if anyone wants to see what settings were used, they are all there. 


Edited by: drkirkby on Jan 9, 2015 12:32 AM