[ntp:questions] Assistance with PPS on Windows

Chris Albertson albertson.chris at gmail.com
Thu May 5 00:12:00 UTC 2011

On Wed, May 4, 2011 at 3:20 PM, Ken Link <klink at numberzero.org> wrote:

> Also, it appears that while the preferred peer continues to be synced
> and polled every 64 seconds, the PPS peer seems to "toggle" from being
> a PPS peer to being rejected nearly every poll (of 16 seconds). Is
> this because the polling rate is different between the PPS source and
> the other sources, or is my serial cable too jittery? The PPS source
> claims to have an offset of +/- 10 ms and a jitter of between 2-3 ms
> depending on the poll.

Maybe a slight misunderstanding about how ntpd works.  First, it never
"syncs" to a server.  Instead it adjusts the rate of the local clock
so as to match the rate of advancement of set of servers it considers
"true".   It may be that your PPS falls  in and out of that set.

Once you get to this level of refinement it is good to be exact.
Adjusting the clock's rate is much different from periodically syncing
the time.  Also using the aggregate rate from a set of references that
have all pasted a clock selection algorithm is different from
selecting one server.  To understand why the PPS may be used you ned
to look at NTP's clock selection algorithm

OK now we can think about why your PPS has poor performance (most of
us can get micro, not milli second level performance from a GPS
derived PPS.)  So how to figure it out.  I'd start by looking at the
GPS.   How good is the PPS signal?  You can't fix what you can't

Chris Albertson
Redondo Beach, California

More information about the questions mailing list