Hello Derek,

so my code looks like this:


now = self.uhd_usrp_sink_0.get_time_now()

        cmd_time_sink = now + uhd.time_spec_t(0.1)
        cmd_time_source = now + uhd.time_spec_t(0.2)

        self.uhd_usrp_sink_0.set_start_time(cmd_time_sink)
        self.uhd_usrp_source_0.set_start_time(cmd_time_source)

        self.uhd_usrp_sink_0.clear_command_time()

I do not understand the sphinx docu.
Can you tell me what is the difference between set_start_time and 
set_command_time?

For me it would be the best to START timer when the first sample is transmitted 
and STOP it when the first sample is received. Then I would be able to 
calculate the time delay.
Or is there a possibility to have a constant delay?

Or could you just recommend me which commands should I use? I will manage the 
rest.
Thank you

Petr


________________________________
Od: Derek Kozel <[email protected]>
Odesláno: čtvrtek 11. dubna 2019 9:36:38
Komu: Horký Petr; [email protected]
Předmět: Re: [Discuss-gnuradio] USRP time synchronization

Hi Petr,

You'll have to use timed commands for starting your transmissions. This can be 
done in GRC, but requires writing some code to issue the commands. Usually I 
find it easier to add the commands to the generated Python if I'm not changing 
the application very often.

Here's a C++ example that shows timed stream commands. You'd need to use these 
for both the transmit and reception side.
https://github.com/EttusResearch/uhd/blob/master/host/examples/test_timed_commands.cpp#L111

Regards,
Derek

On 10/04/2019 14:18, Horký Petr wrote:

Dear GNU Radio users,

I have a simple program,  Signal source (square) -> USRP SINK     USRP SOURCE - 
> TIME SINK.

If I plot Tx/Rx signals in a Time sink, I see, that there is a time differnce, 
which is OK. But each time I start the program, the time difference changes (As 
shown in a picture in attachements).

What can I do? I tried to change Sync, Clock rate, Clock Source, Time 
source....everything. But still the delay is changing each time I execute my 
program.


If the delay was not random I could just substract that. This causes me 
problems in every radar program, since I need precise time measurements.


Regards

Petr







_______________________________________________
Discuss-gnuradio mailing list
[email protected]<mailto:[email protected]>
https://lists.gnu.org/mailman/listinfo/discuss-gnuradio


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

Reply via email to