[time-nuts] Re: Query about 5061B performance and tuning

[Magnus Danielson]

Jim,

So, what I see is a few things:

The "hook" to the left, I wonder how data was captured and processed. 
This have the tell-tale of a particular processing issue that we fought 
before, and I just want to ensure that you have not caught that issue. 
The issue was how overlapping frequency measurements was interleaved and 
then processed further as if they where instant measurements. This 
creates a false surpession. False in the context of "true" MDEV (or even 
true ADEV as we saw).

The slopes themselves look as they should. S/N and a few other factors 
control that. Tube running out may cause a raising S/N, but other 
trimmings may optimize S/N for the beam rate you do have.

The bump at 4E4 for 5061B6 is the tell-tale of the 86400 s period, that 
is that of the day. You have some dependence on the variations of the 
day. I would suspect thermal variation. If you can correlate phase or 
frequency deviations with that of environmental sensors, such as 
temperature, that could help you to estimate the size of the problem. 
That you have 4 other 5061B is great.

Hope it helps!

I don't think I have a healthy tube in my 5061 in order to help you.

Cheers,
Magnus

On 2024-05-22 02:09, AC0XU (Jim) via time-nuts wrote:
> Time Nuts-
>
> I have attached some MADEV estimates for several cesium beam oscillators in my home lab (I have previously posted references to my researh papers about ensemble ADEV measurements).
>
> Two of these oscillators, labelled 5061B1 and 5061B6 perform worse than the others. 5061B6, in particular, seems to have very poor lock.
>
> 1. 5061B1 had a failed tube in which I subbed a very old but working tube. Prior to the failure and substitution, it performance was as shown as "5061B".  All standard tests from the service manual are successful. The Beam I is rather low (15) and cannot be increased.
>
> 2.  5061B6 is unit which, according to all the standard readings, should be performing well. Beam I easily set to 25. Etc. However, it isn't.
>
> Any suggestions about what I should be looking to test/tweak on these two units?
>
> Thanks!
>
> Jim
>
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe send an email to time-nuts-leave at lists.febo.com