W dniu 25.02.2018 o 11:16, Piotr Krysik via USRP-users pisze:
> W dniu 24.02.2018 o 23:10, Marcus D. Leech via USRP-users pisze:
>>
>> What if you use the time-stamps on the streams to time-align, and THEN
>> cross-correlate?
>>
>> Remember that multi-usrp does this as a first step when it's dealing
>> with multiple notionally-time-aligned streams from multiple devices.
>>
>> But that's only *WITHIN* a single multi-usrp object, and you
>> necessarily have more than one multi_usrp object
> To check this I connected two tag_debug blocks to usrp sources. They
> caught following tags at offset 0 (first sample of the stream):
> ----------------------------------------------------------------------
> Tag Debug:
> Input Stream: 00
> Offset: 0 Source: gr uhd usrp source2 Key: rx_time Value: {9
> 3.86875e-05}
> Offset: 0 Source: gr uhd usrp source2 Key: rx_rate Value: 1e+06
> Offset: 0 Source: gr uhd usrp source2 Key: rx_freq Value: 1e+09
> ----------------------------------------------------------------------
>
> ----------------------------------------------------------------------
> Tag Debug:
> Input Stream: 00
> Offset: 0 Source: gr uhd usrp source1 Key: rx_time Value: {9
> 3.86875e-05}
> Offset: 0 Source: gr uhd usrp source1 Key: rx_rate Value: 1e+06
> Offset: 0 Source: gr uhd usrp source1 Key: rx_freq Value: 1e+09
> ----------------------------------------------------------------------
>
>
> There are two rx_time tags with the same value. So they should be
> aligned in time, but they aren't.
>
> What is a bit unexpected for me is the non-zero fractional part.
>
For X310 there is also constant fractional part in rx_time tag. So this
probably won't lead to the source of the problem.
_______________________________________________
USRP-users mailing list
USRP-users@lists.ettus.com
http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
