Weirdness using osmocom_source in GRC

Discussions related to embedded firmware, driver, and user mode application software development

Moderator: robert.ghilduta

Post Reply
wpats
Posts: 9
Joined: Tue Aug 18, 2015 8:13 pm

Weirdness using osmocom_source in GRC

Post by wpats » Thu Sep 17, 2015 8:46 pm

Hi,

I created a GRC flow-graph using the osmocom-source connected to a scope sink. In running this the device seems to get into a weird state with a wide I to Q dc offset. If I unplug and replug the device and then run osmocom_fft in scope mode with manual dc osffet and iq balance correction then it shows almost equal I/Q values, modulo noise. Jon on the forum suggested, in repsonse to some other issue, that the cause for this might be setting the DC offset mode and IQ balance mode to off in the osmocom-source since this overwrites the calibration settings. He suggested setting the values to "Manual" to prevent this. However, even this does not work. Every time I run the GRC flow-graph after a fresh power cycle it goes into the "weird" state. Running osmocom_fft with exactly the same arguments after this shows the same large dc offset. Repeating the power cycle and running osmocom_fft shows reasonable values. I have looked at the osmocom_fft source and checked that the source block settings are identical to that in the python file generated by the GRC flow-graph but it still doesn't work.

What could be happening ?

Thanks for any help/suggestions.

--Patrick

wpats
Posts: 9
Joined: Tue Aug 18, 2015 8:13 pm

Re: Weirdness using osmocom_source in GRC

Post by wpats » Thu Sep 17, 2015 9:07 pm

Ok, this was with the version of libbladerf from last month and gr-osmosdr from May of this year that I had downloaded and built from source. I retried the same scenario now after updating the bladeRF package yesterday and the issue no longer repros. Don't know what fix has changed the behaviour.

Thanks, this is no longer an issue.

--Patrick

jynik
Posts: 455
Joined: Thu Jun 06, 2013 8:15 pm

Re: Weirdness using osmocom_source in GRC

Post by jynik » Fri Sep 18, 2015 7:44 am

Hi Patrick,

Sorry for the delayed response - I was just about to take a look at this. Would you mind just listing the versions you're now using, in case others run into this.

In the future, it might be helpful to show screenshots (even better, with some sampeles posted to dropbox.com) at each step in the procedure you described -- a picture's always worth a thousand words. :)

Happy SDR'ing!
Jon

wpats
Posts: 9
Joined: Tue Aug 18, 2015 8:13 pm

Re: Weirdness using osmocom_source in GRC

Post by wpats » Wed Sep 23, 2015 2:14 pm

Hi Jon,

Here are the software versions I am using now:

bladeRF-cli version: 1.2.1-git-01940e6
libbladeRF version: 1.4.3-git-01940e6

GNUradio:

linux; GNU C++ version 4.8.2; Boost_105400; UHD_003.009.git-186-geb9cfef4
gr-osmosdr v0.1.4-48-g86ad5842 (0.1.5git) gnuradio 3.7.8

--Patrick

Post Reply