
I just measured a bandpass filter on 2.4 GHz with NanoVNA V2_2. With only 1 marker the frequency values are correct, but adding more markers, the values show incorrect frequencys with a "-" sign. (e.g. -1895.967MHz instead of 2400MHz).
Firmware is: git-20200617-1a9a11d.
At 1.3 GHz everything is ok. I found out, that this fault happen exactly at 2147 MHz. Moving the marker bejond that value, it switches to a "-" sign and couts backward.
A signed 32bit integer can count to max. 2 147 483 647, as Chirstian Zietz explaind in nanovna-users@groups.io. So this may be a SW-bug. He recommended to use unsigned integer.
73, DK5DN