USRP2 transmitter spectrum measurements

GNU Radio
1

Hi everyone!

When transmitting with the USRP2 and the RFX400 daughterboard, we get a
very significant frequency spike at the tuned center frequency. The
level of the spike is almost as high as the signal level, and it will
appear even when transmitting a 0-amplitude sine wave signal. Does
anyone know if this is some LO leakage or what this effects are caused
by?

Regards
//Ulrika and Patrik

2

Ulrika U. wrote:

Hi everyone!

When transmitting with the USRP2 and the RFX400 daughterboard, we get a
very significant frequency spike at the tuned center frequency. The
level of the spike is almost as high as the signal level, and it will
appear even when transmitting a 0-amplitude sine wave signal. Does
anyone know if this is some LO leakage or what this effects are caused
by?

Regards
//Ulrika and Patrik

I am having this same problem, I am using Flex400 board

3

has anybody else experienced this problem? I am working on receiving an
RF signal, performing processing on the data, and then retransmitting
it, using a pair of RFX400s. On the TX side, I always get this big
carrier spike, even with no input to the RX side. Ive tried just
straight feeding the RX to the TX, same problem. Sometimes if I stop
and restart the program, the spike disappears.

I have noticed a slight DC offset on the RX data… Ive tried saving the
RX data to hard drive and removing the offset in Labview, then
transmitting the new data, but the spike is still there

I can post spec-analyzer screenshots if needed

4

Hi!
We have encountered the same phenomena. The spikes at the carrier
frequency seems to be especially large on the RFX400, but is also there
on the RFX2400. We haven’t been able to get rid of it, it’s always there
when transmitting. We think it might be an LO-effect.

Regards
//Ulrika and Patrik

-----Ursprungligt meddelande-----
Från: discuss-gnuradio-bounces+ulrika.uppman=removed_email_address@domain.invalid
[mailto:discuss-gnuradio-bounces+ulrika.uppman=removed_email_address@domain.invalid] För Jamin
Chan
Skickat: den 24 juni 2009 23:02
Till: [email protected]
Ämne: [Discuss-gnuradio] Re: USRP2 transmitter spectrum measurements

has anybody else experienced this problem? I am working on receiving an
RF signal, performing processing on the data, and then retransmitting
it, using a pair of RFX400s. On the TX side, I always get this big
carrier spike, even with no input to the RX side. Ive tried just
straight feeding the RX to the TX, same problem. Sometimes if I stop
and restart the program, the spike disappears.

I have noticed a slight DC offset on the RX data… Ive tried saving the
RX data to hard drive and removing the offset in Labview, then
transmitting the new data, but the spike is still there

I can post spec-analyzer screenshots if needed

Posted via http://www.ruby-forum.com/.

Discuss-gnuradio mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/discuss-gnuradio

5

Mattias K. wrote:

Hi,

I think that a similar question was raised a while back, and that the
ticket for fixing it is still open.

http://www.mail-archive.com/[email protected]/msg18067.html

BR
//Mattias

So there is a temporary workaround? Im not sure if i understand the
code in that post, can you clarify?

6

Hi,

I think that a similar question was raised a while back, and that the
ticket for fixing it is still open.

http://www.mail-archive.com/[email protected]/msg18067.html

BR
//Mattias

7

Hi,

Assuming I have understood the problem correctly, and you are writing
your application in c++, you can use the workaround in the post.

For a closer description, I suggest having a look at the API-
documentation. write_io(int, int, int) is defined in class
usrp_basic_tx.

//Mattias

8

On Thu, Jun 25, 2009 at 01:57:23PM +0200, Mattias K. wrote:

Hi,

I think that a similar question was raised a while back, and that the
ticket for fixing it is still open.

http://www.mail-archive.com/[email protected]/msg18067.html

ticket:348 was fixed in [10649].

http://gnuradio.org/trac/ticket/348

Eric

9

After uprading from 3.1.3 to 3.2.2 and also switching from mac osx to
ubuntu 9.04 (significantly easier gnuradio install), i still have the
spike problem.

I was using ctrl+C to end the program and the code looked like this:

if name == ‘main’:
try:
my_top_block().run()
except KeyboardInterrupt:
pass

I tried changing it to use the run functions in my_top_block like this:

if name == ‘main’:
ss=my_program()
ss.start()
raw_input ('Press Enter to quit: ')
ss.stop()

The spike is still there either way.

10

I found out today the spikes i am getting have at least 2 causes…
seems when i configure the RX side for a frequency of a multiple of 2
MHz I get a spike. If I set the RX freq to 450.01 MHz, i will get a
spike at the center freq and at -10 kHz offset. I only see a center
spike if i set the rx freq to 451.01 MHz (not near a multiple of 2MHz)

i still dont know what is causing the first spike at the center freq. I
have tried just transmitting zeros and the gr.null function, and i get
no spikes on the TX side, so i am certain they are being caused on the
RX side.

maybe ill start a new thread, since it seems to have a different cause.
i am also using USRP1 currently. i did get a USRP2 in the mail earlier
this week, maybe ill have better luck with it