[ntp:questions] Re: PPS and refclock timesource differ consistently by ca. 8.8 ms

Ulrich Windl Ulrich.Windl at RZ.Uni-Regensburg.DE
Tue Nov 25 09:55:08 UTC 2003


Michael Sierchio <kudzu at tenebras.com> writes:

> Would someone care to explain why my refclocks always look like

Basically _you_ are responsible to fudge PPS and serial time messages
into agreement (and towards the right value). You must read some
documentation about the clock, and you need some basic understanding
of the delays from your clock to your CPU's interrupt pin. Then just
"fudge". (in a nutshell)

Ulrich

> this?
> 
> oPPS(0)          .PPS.            0 l   64   64   37    0.000    0.134   0.057
> +ULINK_M325E(0)  .WWVB.           0 l   63   64  377    0.000   -8.680   0.049
>   LOCAL(0)        LOCAL(0)        10 l   58   64  377    0.000    0.000   0.001
> 
> Assume for the sake of argument that the other stratum 1 servers agree
> with the WWVB clock, and that my ntp.conf is
> 
> #
> server  127.127.22.0
> fudge   127.127.22.0 flag2 0 flag3 1
> #
> server  127.127.34.0 prefer
> fudge   127.127.34.0 flag1 0
> #
> server  127.127.1.0
> fudge   127.127.1.0 stratum 10
> #
> #########################################################################
> 
> plus public servers.
> 
> Attempting to fix the gap by fudging (time1) fails to align these.



More information about the questions mailing list