[time-nuts] Possible replacement for FE5680A

[Jim Lux]

On 3/12/12 2:20 AM, Azelio Boriani wrote:
> Yes, build three and compare them but your comparator has to have a very
> low noise floor. I think that in addition to building new higher precision
> clocks you need to build new lower noise TI counters.
>

This is part of the thrill (or frustration) of working at the state of 
the art limit.

At work we build state of the art deep space transponders with very low 
added Allan deviation (4E-16 at 1000 sec).  Figuring out a way to prove 
that they work is often harder than designing and building the actual 
article under test, particularly when you want to demonstrate that it 
will do it with a slowly varying input frequency.