6683 PPS in and PPS out drift

[xl600]

I'm synchronizing my 6683 using PPS from an external source.  The incoming PPS has jitter.

I then generate my own PPS (continuous) using the 6683 (All done via the MAX test panel, no VIs).

Initially, the PPS signals stay near each other within the incoming jitter as expected.  But over the span of an hour or so, they drift apart from each other well beyond the synch offset.  How does this happen?  Is the continuous mode PPS not really a scheduled output other than the first rising edge?

Thanks,

XL600

[xl600]

As an experiment, I spun my own PPS using the NI-SYNC create future event VI as opposed to using the MAX PPS out or NI-SYNC time based clock (1Hz) started at a second boundary.  It doesn't look like there's any drift (So far).  So let me change the question:

Does the PPS out or NI-SYNC create_clock.vi use the adjusted and synchronized board clock or do they just use a assume time based only on the start time specified (Then free running after that)?  If free running, only the create future event would seem to be able to follow an external clock reference.

[Wild_man]

Per the user manual for the 6683 it adjusts it's internal timebase to an external time reference on an ongoing basis. For PPS (input) it does mention that the source must be stable or error may be introduced into the system.