Re: [Discuss-gnuradio] How to check at which GPS time correspond the pps edge of set_time_next_pps()

[Nick Foster]

On Thu, 2011-08-04 at 21:11 +0200, Bastien Auneau wrote:
> Hi
> 
> 
> USRP N210, UHD 3.2.1, GPSDO Jackson Labs installed
> windows7
> 
> 
> Using set_time_next_pps()
> How can I be sure at which GPS time (in seconds from 01/01/1970) the
> pps egde correspond to
> 
> 
> So far, here is how I do :
> _ get GPS time (get_mboard_sensors("gps_time")), let it be gps_t_0
> _ set_time_next_pps(0, 1, 54000000)
> _ add + 2 sec to gps_t_0. +1 because time is set at next PPS, +1
> because this seems to be time needed to execute the sensor query on
> GPS (practically, I was always 1 sec behind, before this last +1)
> 
> 
> This has shown to be stable in development lab, for the last few days.
> But is there another mechanism to map the "next pps" to a GPS time (in
> seconds from 01/01/1970) ?
> What bothers me, it's that the second +1 I add is based on the
> get_mboard_sensors("gps_time") execution time (approx 1 sec). But this
> may change randomly (depending on load, HW etc..)

The extra second is required because the GPSDO only outputs a report
once per second. It returns immediately upon receiving a message, so as
long as you call set_time_next_pps soon after
get_mboard_sensors("gps_time"), you should be fine.

--n

> 
> 
> Regards
> Bastien
> _______________________________________________
> Discuss-gnuradio mailing list
> address@hidden
> https://lists.gnu.org/mailman/listinfo/discuss-gnuradio