Here is some more information on this issue.
I am running the firmware from nanovna-firmware-0.2.2 which I downloaded from that was created by edy555. I cleared the memory using the dfu file DMR-CLEAR_MEMORY_DFU.dfu located in the files section of this user group prior to loading the 0.2.2 firmware. The nanoVNA is producing reasonably stable results by itself.
When using nanoVNA-Saver never gets bad data when collecting only 1 101-point segment between the start and stop frequencies. However, it appears to get bad data frequently at the beginning of the second segment when 2 segments are used. When continuous scanning is enable, both the first segment and the second segment frequently have bad data. This is most noticeable if you are measuring a device which has some rapidly changing amplitude and phase information such as a stop band filter. I am trying to measure a US FM-Band stop band filter (stop-band is 88-108 MHz in theory) when I am observing this issue. The frequency extent of the sweep width does not seem to matter much. It is as though the frequencies being measured are not the ones that the software assumes are being measured. This seems like a problem that Rune should look at.
--
Bryan, WA5VAH
