Hello Hans
Thanks for this very interesting article. Therefore we might assume, that the
PPS processing is quite accurate (even on a slow ALIX box) and delay is mainly
introduced by the internet connection. It's an asymmetrical cable link with 200
/ 20 MBit/s and the average round-trip-time to the NTP is about 9 ms.
Mainly because of my personal interest, I thought it would be interesting to do
some (simple) measurements and to learn more about the processing time of the
hardware I'm using. Maybe it's easier to take the box to my workplace and
compare it to the Meinberg NTP which we're running on the LAN.
Regards
Reto
-----Ursprüngliche Nachricht-----
Von: address@hidden [mailto:address@hidden Im Auftrag von hans mayer
Gesendet: Dienstag, 5. April 2016 22:51
An: address@hidden
Betreff: Re: [gpsd-users] Enable PPS Echo
Dear Reto,
How is your Internet connectivity ?
Is it maybe ADSL ?
If so look here: http://blog.mayer.tv/2016/03/16/ntp-and-adsl.html
Kind regards
Hans
Am 05.04.2016 18:17, schrieb Reto Zbinden ST:
Dear all
I'm not really sure if this is the right place for my question, but maybe
somebody may help me:
I'm running Debian with gpsd and ntpd on an older ALIX board as the time source
for my local network. This setup is working well for quite some time and the
clock is kept to within a few microseconds of the 1PPS reference.
But if the time is compared to some reliable public stratum 1 servers (e.g.
metas.ch), it seems that the local clock is always a few milliseconds behind
these servers. I'd like to do some more precise measurement to figure out the
latency of the 1PPS signal processing and to adjust the fudge time. According
to what I've read, the delay can be measured using an oscilloscope and the PPS
echo feature. But I couldn't figure out how the echo function can be activated.
Can anyone provide some guidance on how to enable the PPS echo function? Or
maybe there's another way to do some measurements? Many thanks for your help.
Regards
Reto