gpsd-users
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: GPSD PPS Jitter


From: Gary E. Miller
Subject: Re: GPSD PPS Jitter
Date: Mon, 20 Nov 2023 16:25:22 -0800

Yo George!

On Mon, 20 Nov 2023 17:11:20 -0700
George Sexton via <gpsd-users@nongnu.org> wrote:

> I'm running into a problem using GPSD. What's happening is the PPS 
> offset varies a great deal. Sometimes 7us, sometimes 250us. I got the 
> values by using gpsmon, and reading the difference between the 
> nano-seconds values on PPS outputs.

Uh, no, that is not how you measure PPS jitter.  For startes, don't use
gpsmon, it is a deprecated tool for devs, nit users.

> I've also tried using Chrony 
> configured with SHM refclock, and see the same kind of issue. Some 
> consecutive values of the clock_nsec field from GPS mon are:

The way you measure PPS jitter is with ntpd and ntpviz.  Over at least
12 hours.

> Configuration
> 
> GPS Unit: Garmin GPS-16X HVS

Ancient, 20 years old!  And never worked very well.

The spec on the 16 is +/- 16 micro seconds.  SO you are not so far
from the best you can get from that thatin.

> OS: OpenSUSE Leap 15.5 Kernel Version 5.14.21

Pretty old.

> GPSD Version: 3.25

Good.

> Serial Port: MosChip Semicondiuctor Technology Ltd. PCIe 9912
> Multi-I/O Controller 16550

Good.

You say nothing of your CPU, RAM, etc,  Is this dedicated, or also
runs a desktop?  Many variables.

> The garmin unit is connected to a custom serial port interface board 
> that's in turn connected to the 16550 serial port.

Throw that way.

> The Garmin/Serial port adapter were used very successfully with NTPD
> on an older computer and I would see variance of around 3-10us.

About as good as possible from the 16 (+/- 1 us).

> I tried adding nohz=off intel_idle.max_cstate=0 to my kernel boot 
> command line, and it didn't seem to help.

Depends on the hardware, and load.

> I'm really not sure where to go next on this. I've searched the web
> and haven't run into anything that looks helpful.  Can anyone point
> me in the right direction?

Get a better receiver!

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

Attachment: pgpE17EBY83JW.pgp
Description: OpenPGP digital signature


reply via email to

[Prev in Thread] Current Thread [Next in Thread]