[time-nuts] PPS jitter - Oncore VP w/NTP
Poul-Henning Kamp
phk at phk.freebsd.dk
Wed Jun 22 14:39:08 EDT 2005
In message <42B9AEC8.7090603 at pacific.net>, Brooke Clarke writes:
>Hi Jim:
>
>I was using an Oncore VP and trying to set a Cesium standard. As part
>of the measurement the counter is set to average readings and often the
>sigma (jitter) was very high (hundreds of ns). This turned out to be
>caused by multipath. The fix is to change the elevation mask to a
>higher angle. Note when in position fixed (timing) mode only one sat is
>needed. The VP has a little over 50 ns of sawtooth so that's about as
>good as it will get.
You won't get anywhere close to that with a interrupt timed serial
port DCD signal on a normal PC.
>> % ntpdc -c kerninfo
>> pll offset: 1.106e-06 s
>> pll frequency: 80.865 ppm
This is not indicative of good hardware.
What does
sysctl kern.timecounter
say ?
>> maximum error: 0.006806 s
>> estimated error: 3e-06 s
This looks very typical for a interrupt signal.
The jitter warning thing is very likely if you load on the
machine which might disable interrupts for any amount of time.
--
Poul-Henning Kamp | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG | TCP/IP since RFC 956
FreeBSD committer | BSD since 4.3-tahoe
Never attribute to malice what can adequately be explained by incompetence.
More information about the time-nuts
mailing list