Data aquisition and plotting FFT

I am trying to transmit a 1 KHz sine wave using a USRP2 with carrier
2.48
GHz. I use,

usrp_siggen.py –f 2.468G –x 1k --sine -A 0.5 -v

Verbose gives:-

Interface: eth0 MAC Address: 00:50:c2:85:33:fa D-Board ID: 0x2b
USRP interpolation rate: 16
USRP IF bandwidth: 6.25MHz
Set TX gain to: 0.0
Set center frequency to 2.48GHz
Tx baseband frequency: 2.49375GHz
Tx DDC frequency: -13.75MHz
Tx residual frequency: -4.65584mHz
Set baseband modulation to: Complex Sinusoid
Modulation frequency: 1kHz
Initial phase: 0
TX amplitude: 0.5

Then with a second USRP2 using a different host I try to receive 1M
complex
samples at a centre frequency on 2.48GHz and write them to a .dat file.
I
use,

sudo ./usrp2_rx_cfile.py -e eth1 -d 16 -f 2.48G -g 35 -N 1000000 -v
out.dat

Network interface: eth1
USRP2 address: 00:50:c2:85:33:f7
Using RX d’board id 0x0027
Rx gain: 35.0
Rx baseband frequency: 2.48125G
Rx DDC frequency: 1.25M
Rx residual frequency: 4.65679m
Rx decimation rate: 16
Rx sample rate: 6.25M
Receving 1M samples
Writing 32-bit complex floats
Output filename: out.dat

Finally I try to read the binary .dat file created and plot the signal
over
time and the FFT. I use,

sudo ./gr_plot_fft.py -B 1000000 -R 6250000 out.dat

In the FFT with a block size of 1M the peak is not at 1KHz (verified
using
matlab) and in time I also receive high frequency signals. I basically
get
some sort of AM modulation and a fixed message frequency shift. Can
anybody
tell me why these effects are present. Transmitting signals in the MHz
range
the frequency offset is around 190KHz independant of the actual message
frequency (1MHz, 2MHz…etc…).

Best regards,
Marc.

http://old.nabble.com/file/p27701654/rx.jpeg

View this message in context:
http://old.nabble.com/Data-aquisition-and-plotting-FFT-tp27701654p27701654.html
Sent from the GnuRadio mailing list archive at Nabble.com.