Negative times don't make much sense, the device keeps positive times; the -1 was probably
just mean to signifiy "immediately". So, try with, say 10.0 and 10.1 .
Best regards,
Marcus
On 14.03.24 10:00, Tim Vancauwenbergh wrote:
Hi Marcus
Thanks for your reply! I have been experimenting with the start time. With the default
being "-1.0" for both tx and rx, I set the tx time within the range of -0.9 to -0.9999
(to start later than rx) but there does not seem to be a clear difference in the qt gui
time sink.
The delay between tx and rx remains inconsistent and between 1000 to 2500us.
Is there no other way of getting tx and rx to be synchronised by a pilot pulse
for example?
Best regards,
Tim Vancauwenbergh
------------------------------------------------------------------------------------------
*From:* Marcus Müller <marcus.muel...@ettus.com>
*Sent:* Wednesday, March 13, 2024 3:42:33 PM
*To:* usrp-users@lists.ettus.com <usrp-users@lists.ettus.com>
*Subject:* [USRP-users] Re: B210: synchronise USRP Sink & USRP Source blocks in
GNU RC
Hi Tim,
that delay will vary, the way you're currently are setting things up, and
there's likely
going to be an underflow on the transmitter side, because you start both
"roughly" at the
same time, but it takes some time for RX to produce some samples, while TX
expects samples
to be ready.
Solution is to actually use the "set start time" functionality, with the RX
starting a
couple milliseconds before the TX (try 100 for a start, and reduce for as long
as you get
reliable operation without underflow indications).
Best regards,
Marcus
On 13.03.24 14:18, Tim Vancauwenbergh wrote:
> Dear
>
> I am using a B210 in a loopback configuration to achieve delay measurements.
> Currently, a basic software setup is made in GNU Radio Companion, consisting of a UHD:
USRP Sink block (TX) and an UHD: USRP Source block (RX).
>
> I generate a pulse every few milliseconds. When starting the flow, there is a clear
delay between the transmitted pulse and the received pulse when visualised in a QT GUI
Time sink. I suppose this is due to the delay between providing a sample to the USRP
sink block and the actual transmission + the delay on the receiver side.
>
>
> Is there a method to make sure both sink and source blocks start at exactly the same
time or at least can be visually synced? The delay seems to vary every time the
flowgraph is run. In this loopback configuration, the delay between TX and RX should be
0. When adding a lengthy cable it should no longer match.
>
> In the sink and source blocks I have set the sync setting to PC Clock on next PPS, and
left the start time at its default.
>
> Any ideas how to get a good sync?
>
> Best regards,
>
> Tim Vancauwenbergh
>
>
> _______________________________________________
> USRP-users mailing list -- usrp-users@lists.ettus.com
> To unsubscribe send an email to usrp-users-le...@lists.ettus.com
_______________________________________________
USRP-users mailing list -- usrp-users@lists.ettus.com
To unsubscribe send an email to usrp-users-le...@lists.ettus.com
_______________________________________________
USRP-users mailing list -- usrp-users@lists.ettus.com
To unsubscribe send an email to usrp-users-le...@lists.ettus.com
