[Discuss-gnuradio] IQ changing with restart

[SOUTHCOTT, MARK A CIV USAF AFMC AFRL/RITC]

I've done my homework on this one, crawled through the web & talked to
colleagues. If I am missing something obvious please point it out - it's not
for lack of effort on my part! Also, system admin rules here prevent us from
upgrading every time an update comes out, so if this is fixed in an existing
release please let me know so I can push for the update.

Procedure:
1- Start continuous transmitter (transmits qpsk symbols continuously,
cycling through one of each symbol followed by 4 "zeroes", 7 "ones", 4
"twos", and 4 "threes")
2- Run receiver & note the IQ data (.png attached, straight from USRP, no
processing done)
3- Shutdown & restart receiver, noting IQ data (.png attached, same
location)

The procedure can be change to have the rx run continuously while the tx is
run and reset.

Problem:
If you look at the IQ data in the attached .png files you'll see that the
two files are quite different in the progression from symbol to symbol. The
resultant IQ values (and therefor demodulated symbols) seem to change
randomly with each run of either tx or rx.

Steps I've Taken:
1- I have let the rx run continuously and the constellation (and IQ from the
USRP) is consistent with the starting values (ie. A 90 degree error at
startup stays a 90 degree error through the test)
2- I have tried two different USRP's as the receiver, both exhibit the same
issue
3- I have run the tx and rx off the same rubidium clock with no change

System:
N210 with SBX daughterboad for tx and rx operating at 2.49 GHz
GRC 3.6.5
Ubuntu 12.04 LTS 64-bit

Thank you for your time and effort on this.

Respectfully,

Mark A. Southcott

USRP_in_1.png
Description: PNG image

USRP_in_2.png
Description: PNG image

smime.p7s
Description: S/MIME cryptographic signature