Forum: GNU Radio no packet reception when using uhd_tune_request

Hi All,


I have a couple of questions this time.
1) I have a setup in which I transmit gmsk packets from a tx node which 
are then received successfuly at an rx node (both nodes are USRP N200 
radios and are being operated by GRC). However, when I turn off the 
cordic at two nodes by putting this line in uhd usrp sink at tx node and 
in uhd usrp source at rx node, I do not receive any packets at all at rx 
node.
uhd.tune_request(915e6, 
dsp_freq=0,dsp_freq_policy=uhd.tune_request.POLICY_MANUAL)
I have verified (by means of get_center_frequency() method) that the two 
nodes in this case are set to the same center frequency which is 914.285 
MHz, so, there is no center freq mismatch between tx and rx nodes. In 
this case, I see some underflows at tx node and if I put a scope at rx 
node to see the raw received samples, then I see that the samples are 
not being received continuously (i.e., I see received gmsk samples for a 
few seconds, then noise samples for next few seconds and so on) even 
though the tx node is sending packets back-to-back.

2) Previously, the frequency of DUC/DDC cordic, target frequency, actual 
frequency used to be printed out while running any flow-graph involving 
usrp sink/source but it is not the case anymore. What is the easiest way 
to print all this info while within GRC?


Thanks,
Yu.

On 10/31/2012 01:10 PM, Zing Yu wrote:
> dsp_freq=0,dsp_freq_policy=uhd.tune_request.POLICY_MANUAL) I have
> verified (by means of get_center_frequency() method) that the two
> nodes in this case are set to the same center frequency which is
> 914.285 MHz, so, there is no center freq mismatch between tx and rx
> nodes. In this case, I see some underflows at tx node and if I put a
> scope at rx node to see the raw received samples, then I see that the
> samples are not being received continuously (i.e., I see received
> gmsk samples for a few seconds, then noise samples for next few
> seconds and so on) even though the tx node is sending packets
> back-to-back.
>

Every thing you are doing sounds correct.

Am am a bit confused about underflows. Are you saying that the transmit
underflows, but only with the special tune request? Normally it does 
not?

I suggest that you check that daguhterboard LO is locked after the tune.
print usrp_source.get_rx_sensor("lo_locked").to_bool()

What daughterboard are you using?

> 2) Previously, the frequency of DUC/DDC cordic, target frequency,
> actual frequency used to be printed out while running any flow-graph
> involving usrp sink/source but it is not the case anymore. What is
> the easiest way to print all this info while within GRC?
>

You can print the result of the set_center_frequency(), it returns a
tune_request struct with RF and DSP frequency fields.

-josh