[time-nuts] Newbie with questions, SR620 ADEV

[Tom Van Baak]

> Daer Tom,
> 
> Thanks very much for your reply. I not only see very good answer, 
> by reading your mails(and other replies from )I also got some 
> idea of how you reasoning the answer. Thats most valuable for me.
> 
> Another thing, I remembered you said once that all stabilities are
> relative. When I saw your chart of ADEV for some of your best
> frequency source at: http://www.leapsecond.com/museum/manyadev.gif
> what is your reference? especially when you measure the two 
> H-Masers in long time interval?
> 
> Cheers,
> 
> Lymex, BG2VO

Those plots were made from raw data so strictly
speaking they are all relative. But the frequency
reference in most cases was a maser so I didn't
"subtract the background".

The plots of maser vs. maser were also relative
so theoretically the performance of each maser
might be a little better at certain tau than either
of the two plots show. For example if two masers
have a stability of 9e-14 for some tau you might
be able to assume the actual stability of each
one at that tau is 6e-14 (since 9 / sqrt(2) = 6).

The right way to do it is with a "3 cornered hat".
If you have three clean pair-wise runs of three
similar frequency standards then you can use
a 3 cornered hat to isolate the stability of each
of the three standards. The three lines of code
that do this are at the end of:

http://www.leapsecond.com/tools/3hat.c

/tvb