TVRX2 and GRC

GNU Radio
1

Hello,

We have a new TVRX2, which we are using with a N200 USRP.   The N200

has a GPSDO.

The system seems to work fine when we test it with uhd_fft.

However, we are having trouble figuring out the correct configuration
settings in for a custom GRC flow graph. The graph in question is
simple, consisting of just the USRP and an FFT GUI, as shown in the
first attached screen shot.

 As the second attached screen shot shows, however, the system does

not run properly, producing a string of “DDDDD’s”, with the FFT window
going dark. The warning that reads “Sensor ‘lo_locked’ failed to lock
within timeout on channel 0” seems like it might be related.

 The 3rd and 4th screen shots show the setup menus for the USRP and

the FFT GUI.

 Any advice on what might be going wrong would be greatly

appreciated.

Sincerely,
Dan Marlow

2

Sincerely,
[email protected]
Discuss-gnuradio Info Page
Unfortunately, those images are too small for my aging eyes to glean
useful data from them.

3

Sincerely,
[email protected]
Discuss-gnuradio Info Page
If you put your flow-graph itself (.grc file) somewhere, or attach it,
we can take a better look.

What kind of comptuer? (OS, CPU/MEM specifics, etc).

4

Dan,

If you execute the various “sudo sysctl …” lines shown in the
warnings, does it work any better? You can make these permanent.

man sysctl.conf

  • Jeff
5

On Oct 3, 2014, at 7:30 PM, Marcus D. Leech wrote:

Shirleys Bay Radio Astronomy Consortium
http://www.sbrac.org

Hi Marcus,

Thanks.   The .grc file is attached.      Other details:

o Intel(R) Core(TM)2 Duo CPU     E8400  @ 3.00GHz
o Memory 3GB
o OS 14.1 Ubuntu

Cheers,
Dan

6

–
o Memory 3GB
o OS 14.1 Ubuntu

Cheers,
Dan

Well, for one, your FFT block disagrees rather dramatically with your
UHD source block about the same rate, which you’ve selected as 10Msps.
I’d
go for 6.25Msps, and make certain that your source block and FFT
block agree on sample rate. And if you’re really wanting the 1.7MHz
that
you’ve programmed in the source to 1.7MHz, you don’t need more than
about 2Msps sampling rate. Very few daughtercards have variable
analog bandwidth, the TVRX2, DBSRX2, and perhaps XCVR are the only
ones with variable analog baseband/IF bandwidth.

7

On Oct 3, 2014, at 7:49 PM, Jeff L. wrote:

Dan,

If you execute the various “sudo sysctl …” lines shown in the warnings, does
it work any better? You can make these permanent.

man sysctl.conf

  • Jeff

Hi Jeff,

 Thanks for the suggestion.    We will give that a try.

Cheers,
Dan

8

–
o Memory 3GB
o OS 14.1 Ubuntu

Cheers,
Dan

I’ve attached updates to your graph that should make it behave more
rationally:

2Msps sample rate – consistent with your programmed 1.7MHz analog
bandwidth
Deleted antenna spec in source block–it contained invalid specification
Made the device address spec valid
Moved things into user-changable variables: freq and gain
Made both source and FFT use the “samp_rate” variable, so changing that
variable will change both source and FFT sink
Made FFT sink turn on averaging, and use “freq” variable to set center
frequency

I’d suggest looking it over carefully, and comparing against what you
originally had.

9

On 10/03/2014 11:23 PM, Daniel M. wrote:

 As the second attached screen shot shows, however, the system does

not run properly, producing a string of “DDDDD’s”, with the FFT window
going dark. The warning that reads “Sensor ‘lo_locked’ failed to lock
within timeout on channel 0” seems like it might be related.

On top of what Jeff + Marcus have said, that warning was a result of an
over-zealous checking, which is removed in the latest version.

Cheers,
M

10

On Oct 4, 2014, at 6:39 PM, Marcus D. Leech wrote:

Marcus L.
o OS 14.1 Ubuntu
Made the device address spec valid
Principal Investigator
Shirleys Bay Radio Astronomy Consortium
http://www.sbrac.org
<TVRX2.grc>

Hi Marcus,

 Many thanks.   We will give that a try.

Cheers,
Dan