Hi Bryan,
sweeping currently works as follows:
1) The requested frequency range is set (start and stop frequency)
2) A delay of 300 ms
3) The frequencies are read back by sending the text "frequencies", waiting
50 ms, and then reading them
4) The S11 data is read (again, 50 ms delay included)
5) The S21 data is read
I decided on 300 ms after testing a few times, and figuring that it was
"probably fine". Setting no delay sometimes let to errors, and more than
250 ms seemed to be fine, so I set it to 300. Maybe it's too low? But
then I was using Hugen's firmware.
If you can make a non-continuous sweep that produces a noticeable error,
you can try exporting the S1P/S2P Touchstone file, and looking at the
frequencies. The Touchstone export reads the exact frequencies reported
back by the NanoVNA, so it should tell you what frequencies it thinks it's
reporting data for.
Sadly I don't think it's something I can really address in the application
- I believe it has to be accounted for in the firmware.
Thanks for the work you're doing on this issue, though! Good luck with it!
--
Rune / 5Q5R
On Tue, 1 Oct 2019 at 16:40, bryburns via Groups.Io <bryburns=
[email protected]> wrote:
Rune,
Thanks for taking a look at this thread!
My biggest concern is that with the firmware 0.2.2 from edy555 I always
see incorrect data in the second scan when using 2 scans between 80 and 120
MHz. With a single scan there are no errors over the same frequency
range. The incorrect data is always wrong in the same way. I.E. it is
actual measurements but they are from the wrong frequencies. If I do 5
scans, it appears that the first part of scans 2 through 5 is almost always
incorrect data. If it were occasional errors, I could work with it;
however, it is repeated with my nanoVNA and this firmware. When I request
a single scan over the exact same frequencies as the 2nd or 3rd or 4th or
5th scan, I very rarely, if ever, see an error in the nanoVNA-Saver data,
even in continuous sweep mode. What is different about continuous sweep
mode? Is there a pause between the scans?
I very rarely, if ever, see it with firmware from hugen dated August 2,
2019 using the exact same scan settings 2 or 3 or 4 or 5... Yes, I will
occasionally see errors with the hugen firmware but they are quite rare and
usually pretty obvious. Repeating a sweep or measurement will clear the
errant data.
I don't know much about the transfer protocol between the nanoVNA and the
PC. From your comment above, I would have to assume that you do not get
the frequency information in the same message as the data so there is no
way to throw out data based on the fact that the frequencies you requested
are not the frequencies sent from the nanoVNA. Is that correct?
A few questions, mostly thinking out loud. :>) I wonder if data is being
sent from the firmware before the hardware has actually stabilized at the
new frequencies you have requested? Would it help to pause a bit after
requesting a new set of frequencies?
I will continue looking for more clues to this issue.
--
Bryan, WA5VAH
