Ok, I installed firmware 0.8.4.3 "stability test variant" with config clear. This did not resolve the problem but I made an observation. The effective error increases with each sweep iteration, growing larger over time. See the attached video showing CH0 LOGMAG and SMITH for the first 20 seconds of a continuous sweep from 140-160MHz using a shorted connector. I also attached a console log with various command outputs in case it may help.
So far, looks like my NanoNVA-H4 is the only one having issues with this firmware (outside of the heavily discussed threshold crossover anomaly). My NVA was purchased from R&L Electronics. I am unclear how many entities are actually producing H4s. I have successfully analyzed several antennas and duplexers without incident and testing after reverting back to the original firmware version is all good.
Peter
