[time-nuts] Z380XA The saga of the aging 10811

[James Flynn]

Bob Benward <rbenward at ...> writes:

> 
> Continuing this discussion, I have included a PDF showing the past 
30days of
> EFC.  Amazingly, the drift has reversed direction!  Anyone have any 
insights
> into this behavior?  Each data point represents 10 seconds.
> 

I have found that three things can cause this behavior over a short term 
(days) which do not involve failure of the standard itself ( in no 
particular order ):

1) Power supply. Most standards are immune to this but, if your EFC 
circuit or phase detector involves active circuits, it can show up.

2) Grounds. Bad grounds or ground loops, especially those that share 
current with the oven heater can cause unexplained drift. However, these 
usually have a diurnal temperature signature unless your set up is very 
climate controlled. Screw terminal grounds or connections can "age" and 
change things. Best to ahve everything soldered.

3) Cheap components, especially resistors. I recently had this driving 
me nuts. Carbon comp. and the 2 cent metal film resistors have large 
temperature coefficients and are even sensitive to humidity. Go for the 
<10ppm / deg C metal film or SMD resistors. Some caps also can be tricky 
if the circuit involves large caps.  Leakage is a problem in high 
impedance circuits and can be unpredictable and non-constant.

Finally, there is literature that supports resonators can reverse their 
ageing slope. However, this is rare.

I hope this helps.  I have been there, scratching my head while the 
system seems to have a mind of its own.