I think I’ve solved my problem. Here are some lessons learned that may perhaps 
help others:

**Blocking tx_streamer::send()**

I falsely assumed a send() with time_spec would result in a blocking call. But 
send only seems to block once the TX FIFO in the USRP is full. Accordingly the 
call with time_spec returns after sending the samples over the network, and 
also four more send calls return immediately in my case. Only then the TX FIFO 
is full and a subsequent send() will block. The number of initial send calls 
before blocking of course depends on the FIFO size, number of samples, etc.

**end_of_burst**

Any send() call with a time_spec after the first requires a prior end_of_burst. 
Without it, the time_spec is ignored and the USRP transmits the data right 
away. In my case I call send with end_of_burst=true and zero samples. 
Afterwards the time_spec of the following transmission is considered.

---
_______________________________________________
USRP-users mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to