Strange side-band when transmitting using WBX on USRP1

Hi All,

We’re having some problems with transmitting using the WBX board on a
USRP1 (not an old one).
We have a flow graph with a NULL source going into a NBFM, then
through a multiply_const and into the USRP.
At low frequencies (100MHz) the signal looks ok, but when we get up
passed 500MHz we can see/hear that what should be an unmodulated
carrier also has a single side-band (and is audible as a tone when
demodulated) . And by increasing the const in the multiply_const, we
can see our wanted signal getting stronger while the side-band remains
there at a constant level.

We’ve had a play with manually adjusting the DC offset of the USRP
sink, which seems to change the size of the side-band a bit, but we
don’t really know what to do with that.

We’ve tried several USRPS and several WBX boards and seen the same
problem on all of them.
We don’t see it on the USRP2.
We don’t see it with the flex400 board.

This picture IMG_1169 | happytimes54 | Flickr shows the side-band on the
left and the wanted signal on the right.
The grc file is also attached.

Any help you can offer would be greatly appreciated thanks!

Drew


Andrew Read +64 (03) 357 0787
Test Analyst
Design Verification and Validation Team
Tait Electronics Ltd

=======================================================================
This email, including any attachments, is only for the intended
addressee. It is subject to copyright, is confidential and may be
the subject of legal or other privilege, none of which is waived or
lost by reason of this transmission.
If the receiver is not the intended addressee, please accept our
apologies, notify us by return, delete all copies and perform no
other act on the email.
Unfortunately, we cannot warrant that the email has not been
altered or corrupted during transmission.

On Oct 13, 2010, at 8:45 PM, Drew R. wrote:

We’re having some problems with transmitting using the WBX board on a
USRP1 (not an old one).
We have a flow graph with a NULL source going into a NBFM, then
through a multiply_const and into the USRP.
At low frequencies (100MHz) the signal looks ok, but when we get up
passed 500MHz we can see/hear that what should be an unmodulated
carrier also has a single side-band (and is audible as a tone when
demodulated) . And by increasing the const in the multiply_const, we
can see our wanted signal getting stronger while the side-band remains
there at a constant level.

That sounds (and looks) like an issue that I encountered yesterday with
my USRP + WBX. I was using it to generate some test signals in the
1.5-1.7 GHz range, and I found that there was a constant spur about 1-2
kHz above the transmitter’s center frequency. I also found that I could
reduce its effects by playing with the signal level, transmitter gain
and an external attenuator. For now, I think I can work around it by
simply generating my test signals at an offset from the transmitter
center frequency and adjusting the center frequency to put them back
where I want them, thus moving the spur out of the middle of my test
signal.

We’ve had a play with manually adjusting the DC offset of the USRP
sink, which seems to change the size of the side-band a bit, but we
don’t really know what to do with that.

I also found a bit of the transmitter center frequency bleeding through
when I modulated it with no DC component. I haven’t tried to mitigate
this yet, but my workaround for the spur should also let me ignore this
little bit of unwanted carrier for the time being. I presume that it’s
due to a small DC offset error in the DACs, based on the my observation
of what appears to be a small (sub-LSB) DC error in the receiver ADCs.
In my receive path, I found there to always be a received component at
the tuned receiver frequency, and I found that I could null it out by
adding a small (magnitude < 1.0) complex constant to the USRP source
output before it passed to the rest of my receive chain. Over the course
of a few hours and a power-cycle, I found that I had to readjust the
constant once or twice. This little error is pretty negligible when
receiving larger signals, but it was annoying while I was trying to look
at some pretty weak signals. I may be able to swamp it out with some
external gain, but I didn’t have an LNA handy yesterday to try that.

For the time being, I’m going to try to ignore these DC components in
both the receiver and transmitter by simply offsetting the LO a bit from
the signal that I want to transmit or receive. I have to do some more
tests tomorrow using my USRP as a signal generator, so I hope this will
work.

In the longer term, I’m interested in looking into whether there’s a
better approach to trimming out any DC offsets in the DACs and ADCs. I’m
also interested in learning more about that unwanted spur in the
transmit output (particularly since it’s quite a bit larger than the
component that appears to be caused by a small DC error).

We’ve tried several USRPS and several WBX boards and seen the same
problem on all of them.
We don’t see it on the USRP2.

Interesting.


Mark J. Blair, NF6X [email protected]
Web page: http://www.nf6x.net/
GnuPG public key available from my web page.