[USRP-users] Overrun on 56MHz with B210

Marcus Müller marcus.mueller at ettus.com
Sat Apr 18 11:13:10 EDT 2015


Hi Jason,

yes, in most cases, that is a system bandwidth limitation. Many
controllers just max out; your chipset is usually on the good side[1],
but that doesn't actually mean that much.

Also, I know the noise you're talking about -- it happens to my Laptop,
too, and also on higher rates. I typically hear it when I do something
CPU intensive, though -- my blind guess is that it's a voltage converter
that supplies the CPU with power, and is driven at the edge of its
specs. You're running osmocom_fft in fosphor mode, which means you're
using openCL; I suppose you're using OpenCL as a means to get more out
of your CPU, so that would align with my CPU load assumption.

So, what you can do is first just use the rx_samples_to_file example
that comes with uhd (look into /usr/[local/]/lib[64]/uhd/examples), and
write your samples to /dev/null ; if the overflows go away, you're
CPU-bound, if they persist, you're USB-bound.

Greetings,
Marcus

[1]
http://www.ettus.com/kb/detail/usrp-b200-and-b210-usb-30-streaming-rate-benchmarks

On 04/18/2015 04:17 PM, Jason A. Donenfeld via USRP-users wrote:
> Hi folks,
>
> I'm running a B210 over USB3. I seem to be getting buffer overruns
> when asking for the maximum 56MHz of bandwidth. For example:
>
> zx2c4 at thinkpad ~/Desktop/sdr $ osmocom_fft -a
> uhd,master_clock_rate=56e6 -s 56e6 -f 420e6 -F
> linux; GNU C++ version 4.9.2; Boost_105600; UHD_003.008.002-0-unknown
> gr-osmosdr v0.1.4-16-g61184a19 (0.1.5git) gnuradio v3.7.6.1-103-g8ecfd13a
> built-in source types: file rtl rtl_tcp uhd
> -- Operating over USB 3.
> -- Initialize CODEC control...
> -- Initialize Radio control...
> -- Performing register loopback test... pass
> -- Performing register loopback test... pass
> -- Performing CODEC loopback test... pass
> -- Performing CODEC loopback test... pass
> -- Asking for clock rate 56.000000 MHz
> -- Actually got clock rate 56.000000 MHz
> -- Performing timer loopback test... pass
> -- Performing timer loopback test... pass
> -- Using subdev spec 'A:A A:B'.
> [+] Selected device: Quadro K2000M
> OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO.....[repeated enormous amounts]
>
> Incidentally, when it's overrunning, I can hear a high pitched
> capacitor noise inside of my laptop (Thinkpad W530, Intel i7-3820QM,
> C210 chipset), which is a bit funny. Is it possible that the system's
> USB controller is simply not up to task? Or is this a configuration
> issue I can address?
>
> Thanks,
> Jason
>
> _______________________________________________
> USRP-users mailing list
> USRP-users at lists.ettus.com
> http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com





More information about the USRP-users mailing list