Rune,
I have been using several new features in nanoVNA-Saver 0.1.3 and think that they are great additions.
I am trying to use edy555 firmware version 0.3.0 with nanoVNA-Saver version 0.1.3. There are several new features in this firmware that are beneficial to me when I cannot use nanoVNA-Saver, so I would like to keep this firmware loaded in my nanoVNA. However, I am seeing a lot more data errors with this version of firmware than I saw with 0.2.3 from edy555. It is so bad that I cannot get a reliable calibration in nanoVNA-Saver without using 5 averages and dropping 2 values in nanoVNA-Saver that are the worst. This was not the case with the 0.2.3 version of firmware. With the new firmware and software I had to do repeated scans until there were no obvious errors in the data for each termination/through condition in order to get a calibration that was reasonable. I could not use the Calibration assistant. The errors are often 0.25 to 0.5 dB in the S11 trace and quite obvious. Even data in the isolation state for through calibration had large errors in amplitude, as much as 20-30 dB. I never see these errors when I operate the nanoVNA by itself and simply use the displayed result. It appears to be either something in the transfer (timing?) or something related to scanning. Generally, there are about 3 or 4 bad samples per scan, often at exactly the same frequency. As I have observed with some previous firmware, the bad data is never in the first scan you do from nanoVNA-Saver.
I have to wondered if the by-passing on the 5V power supply may be related to this issue. Without replacing the 5V supply or improving filtering I am not sure how to isolate this potential issue. Up to this point, I have been reluctant to start modifying the hardware.
Have you been working with the 0.3.0 firmware from edy555? Has anyone else reported this issue?
--
Bryan, WA5VAH
