[time-nuts] What makes a Symmetricom SyncServer S200 lock to the GPS signal?

[Tom Verbeure]

My electronics flea market SyncServer S200 came with a Furuno GT-8031H GPS
module. This module had its GPS week rollover on 2022/09/18.
I've modded the S200 with BNC connectors so that it outputs 10MHz and 1PPS.

When I connect it to a GPS antenna, the S200 never gets into GPS lock mode:
the 10MHz output stays at 9,999,996Hz.

I purchased an IL-GPS-0030-B module on AliExpress. These are supposed to
have their rollover date in August 2024, though that's not what I'm seeing:
it reports a date of sometime 2004, exactly 1024 weeks earlier than today.
While I don't strongly care about the date, I care about using the S200 as
a GPSDO, though it would still be nice for the date to be correct.

With the IL-GPS module, the S200 goes into GPS lock mode, and the 10MHz
output frequency is pretty much identical to the one of my TM4313 GPSDO.
(It's a bit surprising that the 10MHz of the dirt cheap TM4313 is
*slightly* more stable than the S200, but maybe that's probably just a
matter of PID control loop parameter choices.)

What I'm trying to understand is why the GT-8031H isn't and the IL-GPS is
locking when both are reporting an incorrect date.

I've already done the following:

- soldered probe wires to the 10-pin M12M module connector
- confirmed that both module are sending out a 1PPS signal to the S200
motherboard
- recorded the 9600 baud RX and TX Motorola binary commands bewteen the
motherboard and the modules
- wrote a script to decode the recorded request/response traffic to and
from the module

There are some differences in what the module sends back to the S200, but
nothing major jumps out. It's just that after ~20 minutes or so, onces the
module reports 7 to 8 satellites, the S200 goes into lock and only then
some additional data is sent out by the ILS-GPS.

The ultimate goal is to come up with an interposer board that intercepts
the GT-8031H data and patches the date.

Any ideas?

Thanks!
Tom