Subject: Re: soekris box still keeps terrible time
To: None <netbsd-help@NetBSD.org>
From: David Lord <david@lordynet.org>
List: netbsd-help
Date: 11/16/2006 07:12:25
On 16 Nov 2006, at 0:39, James K. Lowden wrote:
> David Lord wrote:
> > I use ntpd to sync to internet time. A few of my pcs have clocks too
> > far out for ntpd to work effectively as the natural clock drift is at
> > a greater rate than ntpd can lock to. The fix is to adjust the value
> > of TIMER_FREQ in kernel configuration to offset the system clock
> > drift. I aim for within -10 ppm to 0 ppm. Of course temperature and
> > power supply voltage can cause fairly wide variation so near is good
> > enough vs >>500 ppm drifts of some pcs. Other pcs are within 30 ppm
> > without any tweaking.
> >
> > eg. from K6X400_0.32
> > ! # K6X400 is approx -60 ppm so aim for > +50ppm change
> > ! # default 1193182
> > ! # for +68
> > ! options TIMER_FREQ=1193250
>
> Thank you, David. This is very helpful, if still a bit mysterious.
> Here's what I'm seeing:
>
> cherry$ grep ntpd /var/log/messages |head
> Nov 15 11:09:29 cherry ntpd[29870]: time reset -2.326574 s
> Nov 15 11:09:29 cherry ntpd[29870]: frequency error -512 PPM exceeds
> tolerance 500 PPM
> Nov 15 11:14:53 cherry ntpd[29870]: frequency error -512 PPM exceeds
> tolerance 500 PPM
> Nov 15 11:34:10 cherry ntpd[29870]: time reset -3.754409 s
> Nov 15 11:34:10 cherry ntpd[29870]: frequency error -512 PPM exceeds
> tolerance 500 PPM
> Nov 15 11:39:33 cherry ntpd[29870]: frequency error -512 PPM exceeds
> tolerance 500 PPM
> Nov 15 11:54:35 cherry ntpd[29870]: time reset -2.699728 s
> Nov 15 11:54:35 cherry ntpd[29870]: frequency error -512 PPM exceeds
> tolerance 500 PPM
> Nov 15 11:59:57 cherry ntpd[29870]: frequency error -512 PPM exceeds
> tolerance 500 PPM
> Nov 15 12:16:59 cherry ntpd[29870]: time reset -3.915989 s
>
> cherry$ config -x /netbsd | grep TIMER
> options TIMER_FREQ=1189200
>
> I don't know what units TIMER_FREQ is in. Taking a naive guess, I add 512
> to it, giving me 1189712. Since you suggest overcompensating, perhaps
> 1189800?
>
> Is that the idea?
Exactly.
It usually takes me a few of shots at it but I normally trim down the
kernel in several steps anyway to add extra features and remove any
not required.
David