[time-nuts] Advice Z3816A jumping to Holdover: 1PPS TI exceeds hold threshold, then Recovery: phase alignment for 3-4 hours.

[Mark C. Stephens]

I have a Z3816A and it periodically jumps into holdover with error message:
Holdover: 1PPS TI exceeds hold threshold
Then for the next 3-4 hours it gradually adjusts the phase alignment.
The error message is Recovery: phase alignment [TI +435.5 us]
(when I noticed it had changed, they figure was likely to have been much higher)
The TI then gradually decreases, currently it is sitting on:
Recovery: phase alignment [TI +129.9 us]

I am hoping that I won't have to power cycle the GPSDO to get it locked again.

I emailed the seller (Yixun HK) and they are trying to tell me power cycle is only fix.
They also claim the cause is due to weather?

I explained to Yixun I have a lot of smart clocks here and none have exhibited the same problem ever.
This is the 3rd defective Z3816A I have got off Yixun and it is costing me a small fortune to send these back.
The first was reporting the 12v Supply was out of tolerance, a quick jump into pForth confirmed the A, B and C 12v supplies were over 12.5v.
I ended up sending it back to Yixun.
The second I was able to fix myself, apparently someone had changed the OCXO but left the insulating-spacing washers out from the pins of the OCXO.
This was causing a short on the EFC.
That Z3816A unit and now appears to be running normally except the PU is terrible, I'll run Cal for 24 hours and see if it improves.

However, Now the 3rd unit is getting this TI exceeded error.
I think I have had enough of shipping things back to these guys - they won't replace anything, including parts, unless you ship back to them first.


I would just like to get some informed opinions, could the issue be with:

a)      Antenna

b)      GPS Module

c)       OCXO

d)      PLL circuit on main board.

Many thanks,
--marki