[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [gpsd-users] ntpd time accuracy with gpsd and 1PPS

From: Greg Troxel
Subject: Re: [gpsd-users] ntpd time accuracy with gpsd and 1PPS
Date: Mon, 11 Nov 2013 20:41:00 -0500
User-agent: Gnus/5.130006 (Ma Gnus v0.6) Emacs/23.4 (berkeley-unix)

Yakov Shmulevich <address@hidden> writes:

> On Wed, Nov 6, 2013 at 2:27 PM, Greg Troxel <address@hidden> wrote:
> Whеre is the value 70ms? How did you calculate it?

From the 78, and I don't know why I said 70 now.   Perhaps I should have
said 78, or 78+9.

>> ntpd will steer the clock to the signal, so the SHM(1) line being -0.017
>> does not tell you about accuracy, just that ntpd is tracking the pps.
>> The line indicates an offset of -9.8 with delay of 18, so
>> either it's off, you have massive asymmetric delay, your PPS is off, or
>> some combination.

> What is bad in this situation? As I understand, in case of
> my server will be stratum-2 NTP server. So, the 18 ms
> delay is legitimately. But -9.8 ms offset is because it not
> synchronized with Also with PPS it take time for ntpd to
> synchronize on 17us, because it starts from ~30msec offset.

So 17 us offset is basically zero.   The point I am trying to make is
that a remote server being off, asymmetric delay, and PPS being off are
not really possible to distinguish without more measurements.

> I am working with GPS module through FT2232H (USB to dual serial) with 1PPS
> connected on DCD pin.
> Did you have some evaluation of pps delay for this kind of GPS receiver
> connection?

I have not measured it.  With USB, I'd estimate 700 us of delay.
Also, you should check how long your pulse is and which edge is being

Sorry, but there are no easy answers for your situation.

Attachment: pgpXcFnwdyJj2.pgp
Description: PGP signature

reply via email to

[Prev in Thread] Current Thread [Next in Thread]