Some more details
_ when Jackon Labs GPS detected as Generic NMEA GPS device, no need to power
cycle USRP to get back to normal. It just happen randomly to be 3 times in a
row when I tested it, but it's as random as other problem with gps_time and
gps_locked
_ when some problem occurs (GPS device detection, error when query gps_time
or gps_locked), it is very often that few query in a row, or close to each
others have such error
The common point I see from these calls is that data is fetched from GPSDO
serial. So there must be some problem with serial connection ? Or with
string manipulation coming out of serial ? Or maybe the way it's sent to
UHD-host ?
We have many USRP in our lab, but only one GPSDO. Nick, do you
have opportunities to try these sensor calls on some other devices ? Then
we'll see if it is related to our HW or the way GPSDO has been installed (I
followed strictly the procedure on the wiki, as far as I know...) ? Or if it
is SW problem
Regards
Bastien
On 2 August 2011 17:44, Nick Foster <n...@ettus.com> wrote:
> Bastien,
>
> Does gps_time work correctly 100% of the time for you?
>
> --n
>
> On Tue, 2011-08-02 at 17:08 +0000, Bastien Auneau wrote:
> > Hi All
> >
> > On USRP N210, UHD 3.2.1
> > GPS does lock, because it gets the right time (not the 01/01/2006 set at
> > init)
> > using get_mboard_sensors("gps_locked") fails sometimes
> >
> > 1.
> > It may print on console :
> >
> > UHD Error:
> > get_nmea(): unsupported GPS or no GPS detected
> >
> > while at the start of the code, it always prints :
> >
> > Found a Jackson Labs GPS
> >
> > At this pont, I have to reboot the USRP. I have seen this error twice on
> > 20 to 25 runs
> >
> > 2. Or :
> >
> > UHD Error:
> > get_token: too few tokens in reply $GPGGA, [etc... then there are
> > more or less args after that, randomly)
> > GPS lock status: locked
> >
> > This second error occurs 1% of the time, but following
> > get_mboard_sensors("gps_locked") succeed (no need to restart USRP)
> >
> > 3. Which leads me to :
> > How do you check if a get_mboard_sensors(...) fails or not ?
> > Because even when it fails, the value returned seems ok. And there is no
> > error field.
> > Or should this not happen ?
> >
> > Thanks and Regards
> > Bastien
> >
> > _______________________________________________
> > Discuss-gnuradio mailing list
> > Discuss-gnuradio@gnu.org
> > https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
>
>
>
_______________________________________________
Discuss-gnuradio mailing list
Discuss-gnuradio@gnu.org
https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
