[LinuxPPS] Offset and jitter of GPS/DCF PPS signals

Thorsten Mühlfelder thenktor at gmx.de
Thu Jan 22 16:33:30 CET 2009


Hi,

Linux 2.6.28 on the DCF server
Linux 2.6.28rc6 on the GPS server
both with LinuxPPS patch.

ntp 4.2.4p5 was built with the glibc nano patch.

All running on a Slackware based distribution.

Thorsten

PS:
The PCI cards:
Meinberg GPS170PCI v1.10
Meinberg PCI511 v1.00 (dcf)


On Thu, 22 Jan 2009 16:25:41 +0100
Heiko Gerstung <heiko.gerstung at meinberg.de> wrote:

> Thorsten,
> 
> which OS?
> 
> Regards,
>   Heiko
> 
> 
> Thorsten Mühlfelder schrieb:
> > Hi everyone
> >
> > I just wanted to ask what you guys think about my graphs. I'm using 2 servers with Soekris AMD Elan (133MHz) boards: Number one has a Meinberg PCI DCF77 card, number two has a Meinberg PCI GPS card. Both Meinberg cards have a PPS output, too. The servers get their timecode by DCF/GPS and are synchronized by the PPS signal of their own cards.
> >
> > Here you can see how the ntpd on the DCF server sees the PPS signal of the DCF card:
> > http://img-up.net/img/dcf-ppsURM2775K.png
> >
> > Here you can see how the ntpd on the GPS server sees the PPS signal of the GPS card:
> > http://img-up.net/img/gps-ppsoNL9T8F4.png
> >
> > Do you think these values are OK?
> > I think the DCF PPS can't get any better because of the technical specification of DCF77.
> > But I don't know about the GPS server? Is it OK or should it be better? Also I wonder about the constant 15 us jitter value.
> >
> > Greetings
> > Thorsten Mühlfelder 
> >   



More information about the LinuxPPS mailing list