[USRP-users] Calibrate tool not work
marcus at hostalia.de
Sat Sep 14 13:22:22 EDT 2013
sorry, I can't really make much sense of what you are telling us, so I reply what I understood.
> I do not use the loop version tool.
I did not understand that. "loop version tool" is what? and what are you using to calibrate?
> I calibrate usrp1 and wbx board.
ok, understood, USRP 1 and WBX. Versions unknown/unreported.
> since usrp1 do not provide calibrate ability
You think the USRP 1 is not able to calibrate itself.
That is correct, but also true for USRP 2/N210 and every other USRP I know of.
I/Q imbalance calibration is usually done completely in software on the host PC.
That relies on signal that leaks from the TX chain into the RX chain, see
So to calibrate right, you should _not_ plug in a spectrum analyzer _while_ calibrating.
> I modify the fpga
So you modified the FPGA image. What did you do?
> then I modify the calibrate tool let usrp output to a spectrum
The calibrate tool just uses the normal TX / RX interface but sets the antenna to "CAL", a special value that basically builds a loopback between TX and RX (if I read lib/usrp/dboard/db_wbx_simple.cpp correctly).
> no matter what the value I set the output do not affected.
I can't understand: You set which values to what?
> wonder if anyone successfully calibrate usrp and the image signal be
I can't try it right now, but to be honest: Calibration worked for me in the past, and it suppresses the image spectrum reasonably well; I'd expect your image to be more than 40dB weaker than your desired signal.
How high is the power of your image signal and how high the power of your desired signal? Can you upload a photo somewhere?
> If so then it might be my modify have mistake.
So far I don't think you made a mistake :)
> On 9/12/13, Marcus Müller <marcus at hostalia.de> wrote:
>> Um... you can't actually see anything with an external spectrum analyser
>> "during" calibration; calibration basically is based on leakage between
>> TX on RX on the same daughterboard, if I remember correctly.
>> After TX calibration, the image should be as good as gone. Since it is
>> not affected, my wild guess is that it is not actually an image due to
>> IQ imbalance.
>> However, the information you provided is a little to sparse for me to
>> take an educated guess. Supply us with
>> - device type (USRP / daughterboard incl. version)
>> - signal type
>> - center frequency
>> - amplitude difference (in dB preferably) between original and mirror
>> - theorethic/measured (by absfft plot) spectrum of transmitted samples
>>> I want to calibrate usrp i q imbalance. I use calibrate tool in uhd,
>>> but the calibrate result is not right. I watch the spectrum analyzer
>>> to observe the image signal during calibrate. but the image signal do
>>> not affected by the calibrate tool.
>>> USRP-users mailing list
>>> USRP-users at lists.ettus.com
>> USRP-users mailing list
>> USRP-users at lists.ettus.com
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the USRP-users