I made a quick fix which seems to work. I simply changed MAX_SAMPLES in
tx_samples.cc from 371 to 350.
BR/
Per
Quoting Josh Blum <[email protected]>:
It looks like the trailer (which is zero) is being written into the
file. The vrt::expanded_header::parse routine should be calculating
the correct length from the header. This call can be found in the
usrp2_impl.cc
I dont think the parser is wrong, nor do i think the
rx_streaming_samples is getting the length wrong. You may want to
check wireshark and verify what you are seeing over the wire.
-Josh
On 02/22/2010 08:24 AM, Per Zetterberg wrote:
Hi All,
I am transmitting using tx_samples and receiving with
rx_streaming_samples. I use basic daughterboards. The signal transmitted
is just a sequence of the value 10000+10000j. The carrier frequency is
set to 10MHz. The interpolation/decimation rate is 64. The transmitter
is repeating i.e. "-r" option is used.
The plot http://www.s3.kth.se/~perz/usrp/p1.jpg shows the absolute value
of the received data when the transmitter is using vrt code. The plot
http://www.s3.kth.se/~perz/usrp/p2.jpg shows the received signal when
the transmitter is using trunk code (checked out a few weeks ago). The
receiver is using vrt code in both cases.
The interval between the points where the signal disappears is
suspiciously close to 371. I think I also observed this "signal loss"
when looking at the signal output using oscilloscope (although it's an
old analog scope and therefore difficult to say).
BR/
Per
_______________________________________________
Discuss-gnuradio mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/discuss-gnuradio
_______________________________________________
Discuss-gnuradio mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/discuss-gnuradio
_______________________________________________
Discuss-gnuradio mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/discuss-gnuradio
