flyoob / NanoVNA-F

NanoVNA-F, NanoVNA-Fairy, FreeRTOS version of TTRFTECH's NanoVNA.
GNU General Public License v3.0
163 stars 55 forks source link

Calibration problems at rev 0.0.4 #7

Open derfunker opened 4 years ago

derfunker commented 4 years ago

Something is wrong with the CAL routines... When doing S11 and S21 Calibration the errror terms are displayed on screen as described. But when the correction is activated only the S11 trace is corrected! Only when I have one trace (S21) on screen the correction works. In some cases the S11 sweep only works when the marker is set e.g. to center. I think here are some minor software bugs to find. In principle the device works fine! Thank you!

derfunker commented 4 years ago

I found it. The bug was sitting in front of the device :-) Now I was successful with the Cal.

flyoob commented 4 years ago

Thank you for your feedback, we will continue to pay attention to this issue.

derfunker commented 4 years ago

Thanks! One question: what is the maximum rf input level at port 2 (in case of measure of an active DUT) I would apriciate to know about any new software updates. Best regards, Klaus

BH5HNU notifications@github.com schrieb am Mi., 4. Dez. 2019, 12:49:

Thank you for your feedback, we will continue to pay attention to this issue.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/flyoob/NanoVNA-F/issues/7?email_source=notifications&email_token=AN5WSOVZRUI7PLAFQBSOJG3QW6KMVA5CNFSM4JTSUXVKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEF4X3AA#issuecomment-561610112, or unsubscribe https://github.com/notifications/unsubscribe-auth/AN5WSOXFPC4BOBINNRDKTXTQW6KMVANCNFSM4JTSUXVA .