| Re: [chrony-users] Multi-PPS+RTC Setup Questions |
[ Thread Index |
Date Index
| More chrony.tuxfamily.org/chrony-users Archives
]
- To: chrony-users@xxxxxxxxxxxxxxxxxxxx, Hal Murray <hmurray@xxxxxxxxxxxxxxx>
- Subject: Re: [chrony-users] Multi-PPS+RTC Setup Questions
- From: Matt Corallo <ntp-lists@xxxxxxxxxxxxxxx>
- Date: Tue, 29 Sep 2020 19:27:04 -0400
- Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=mattcorallo.com; s=1601420505; t=1601422025; bh=9ElqrPChqmUnVRV9xXeYzOKBqlCshMfg0TWT6z83esg=; h=To:References:From:Subject:Date:In-Reply-To:From; b=apuGx6KmgoUW8cn6K85E5m2RDGu5Mt/gw2PAyO2dW9kHBIio24pD5ZaNc/aV9csSR w8mSR8l6IGtrfns1V9xa4LJBcPNbclHuGpEgZHQujgenBBlri7QuOqkPOahetVcvTM GYexzorbg/UQmaQd3JNgclUDxCdx5kkx01mDm4WQTukbaZti2a/j2Hswk/dBPo7zT9 vz2SimlkcpT7/S7xyQrfhX92fsxnM/ol3WFY6oG7Bs7DFhfJt2p3W8NB4+O+c1o/+9 oB9klyMwAnSMuVlEJOBTyr4nWmEMAeKuv1680ozgta2WtjpNu5KtGicYW1W12KqYEE Of+5/CO7C0thA==
On 9/29/20 7:16 PM, Hal Murray wrote:
Just a note, at least in my case, the PPS outputs are not variable at 30us,
they are fixed at a 30us offset from each other,
Many GPS systems have an option to correct for the length of the antenna
cable. It's worth checking on that.
Both set to 50ns, sadly - and my cable is't a few km long :)
An oscilliscope is the right tool for that sort of difference.
Yea, I need to go pick one up for this, at this point, I think.
Another thing to investigate is the other edge of the clock. Some PPS pulses
are 100 or even 500 ms long. Many are short 10-50 usec. Triggering on the
wrong edge of a short pulse could get you numbers like like a 30 usec offset.
UBX lets you set it - currently (and default) 100ms (and both ppscheck, which polls DCD, seems to think that it is the
configured length).
On Linux, you can see that with command line hacking. The PPS timing is
available via
cat /sys/class/pps/pps0/assert
cat /sys/class/pps/pps0/clear
You will get something like:
1601420619.999999124#1112896
The number after the # is a sequence number. The number before is the time -
seconds and nanoseconds.
ntpd disables the edge it isn't using. I don't know what chrony does. You
may have to disable chrony and reboot the system. (or write the code to turn
it back on)
It appears chrony, if started after ldattach, does the same.
I thing gpsd sets up the PPS devices. If nobody does, you need something like:
ldattach 18 /dev/ttyS0 &
I think gpsd tries to dance around a possible extra/missing inversion by
assuming the pulse is narrow and using the edge that arrives first.
That's likely a rathole, but it doesn't take a lot of time/effort to explore.
Yep, I think at this point I've poked everything possible, all the way through to kernel patching, and no dice. Hence my
agreement that its time for a scope (and nagging u-blox, cause, like, wtf).
Thanks,
Matt
--
To unsubscribe email chrony-users-request@xxxxxxxxxxxxxxxxxxxx
with "unsubscribe" in the subject.
For help email chrony-users-request@xxxxxxxxxxxxxxxxxxxx
with "help" in the subject.
Trouble? Email listmaster@xxxxxxxxxxxxxxxxxxxx.