[time-nuts] Oncore VP problem

Tom Van Baak tvb at LeapSecond.com
Wed Jul 18 22:57:19 EDT 2007

> The other thing I've wondered about is how did HP correct the but in the early 
> sawtooth error back when SA was on?


Someone with a 'scope probe could tell for sure but my
understanding is that they didn't correct for sawtooth
in hardware or in software.

One reason may be that early PVT6 or VP's didn't have
a reliable sawtooth correction message. Dr TAC could
comment on this.

The other reason is that, as many of us have pointed
out, sawtooth correction is nice but not necessary for
a GPSDO. The longer the time constant the more true
this is.

So the HP SmartClocks had two things in their favor in
this respect. One is a very high-performance OCXO
allowing a longer time constant than other GPSDO.

The other is that they were designed when S/A was in
effect. Everyone designing GPSDO's in the 1990's needed
to have longer time constants -- to reduce the effects of
S/A. Consequently, sawtooth was less of an issue than it
is now.

So those are my three hunches. Anyone know for sure?


More information about the time-nuts mailing list