[time-nuts] 58503A alarm light on
csteinmetz at yandex.com
Sat Apr 4 20:21:32 EDT 2015
>I've seen the same sort of < 1 us holdover estimates out of mine as well.
>I for one do *not* believe that it's going to do 200 ns holdover in 24 hours
No, nor do I. But I do expect <1uS. Mine seems to stay within
<500nS over the first 24 hours with regularity, although I haven't
tested it very often.
>the objective of the prediction function was to decide if it could
>make 10 us over 24 hours
> * * *
>In this case, a >11 us estimate says "I'm not going to make it".
>That means the
>box is telling you it is not able to do what it should.
One expects a safety factor between the minimum spec and typical
operation. I'd be concerned if a 58503 was consistently
predicting >2uS (to me, >2uS means "I may still meet spec, but I'm
not as stable as I should be").
More information about the time-nuts