RE: GPSD Not Using /dev/pps0 But ppstest Shows /dev/pps0 is OK

[Joshua Quesenberry]

Thanks that helps! A few more questions.

Is the fix time modified at all when PPS present? Or does it remain what was
in the NMEA string?

Would (pps.clock + (gps fix time - pps.real) be an appropriate calculation
to relate the current GPS data to the actual system time at the point the
GPS sent out the NMEA string?

At this point toff.clock seems to always equal pps.clock, does that imply
there's minimal delay from the time the GPS sends the data till the time
GPSd processes it?

-----Original Message-----
From: gpsd-users <gpsd-users-bounces+engnfrc=gmail.com@nongnu.org> On Behalf
Of Gary E. Miller
Sent: Wednesday, December 2, 2020 8:18 PM
To: gpsd users <gpsd-users@nongnu.org>
Subject: Re: GPSD Not Using /dev/pps0 But ppstest Shows /dev/pps0 is OK

Yo Joshua!

On Wed, 2 Dec 2020 20:03:07 -0500
Joshua Quesenberry <engnfrc@gmail.com> wrote:

> I just tried ntpshmmon again, this time while the command you 
> recommended for creating the log and it IS outputting information now. 
> Narrowing it down, it looks like `-n` is making the difference here. 
> My application is also now seeing gps_data_t's toff and pps being 
> populated. So now that I have the data coming through, I have a few 
> questions about what I'm seeing.

I wish -n was the default, but having it off may save someone a few mW of
poer.

> What is the difference between toff and pps?

toff is the time from the serial stream.  Can be off several seconds.
Since you are using tcp, could be way off, which can confuse PPS.

> Should toff be equal to the time in gps_fix_t's time variable?

No.  toff is when the fix was actually received, not the time that was in
the fix which may be delayed in various ways.

pps is the time from KPPS.  Since you have KPPS.

> What is the difference between real and clock?

Notice the tv_nsec is always zero on pps real?  That is the "real" time that
PPS tells you.  Clock is what the CPU clock said when the PPS interrupt was
handled by the kernel KPPS driver.

> It looks like all four are aligned more with the GPS time than the 
> System time,

Not really.  real is the KPPS and GPS time.  clock is the System time.

If ntpd is doing its job they should be close.

> I was expecting them to be relating to the System time since the PPS 
> timestamp comes from the Kernel?

Yes, "clock" is the time as seen by the kernel.

> Is there some
> conversion happening to base them off GPS instead of System timestamp?

Nope.

> The tv_nsec portion of each isn't changing that often, I'm not sure 
> what that means, seems like it should be incrementing by 100,000,000 
> each update since I have my GPS running at 10Hz?

It means your system clock is either stable, of being disciplined by ntpd.

RGDS
GARY
---------------------------------------------------------------------------
Gary E. Miller Rellim 109 NW Wilmington Ave., Suite E, Bend, OR 97703
        gem@rellim.com  Tel:+1 541 382 8588

            Veritas liberabit vos. -- Quid est veritas?
    "If you can't measure it, you can't improve it." - Lord Kelvin