[time-nuts] Re: 10811 behavior - addendum

[John Miles]

> -----Original Message-----
> From: Skip Withrow via time-nuts [mailto:time-nuts at lists.febo.com]
> Sent: Monday, February 27, 2023 8:15 PM
> To: time-nuts at lists.febo.com
> Cc: Skip Withrow
> Subject: [time-nuts] 10811 behavior - addendum
> 
> I just noticed that the ripples in the phase plots are the slow cycle at
> about 1/10Hz.  Also, I'm pretty sure the power supplies are not the
problem
> as the frequency of the ripples between the two 10811 units are slightly
> different.
> Skip Withrow

Looks like a beatnote artifact.  They are tough to avoid once you get down
into the 12s.  Check the 10811's frequency offset from 10 MHz to see if it
coincides with the beatnote period; it probably will.  

Tips:

1) Hit 'n' and use the notch filter to identify and remove the tone.  This
can take some fiddling.

2) Make sure the connection integrity is rock-solid.  Even the smallest
pigtail in open air will couple to the primordial 10 MHz background
radiation we all have floating around.  Don't lift the coax grounds at
either end.  

3) Cables should be of good quality and no longer than necessary.  As a
proxy for cable leakage, measure the shield's DC resistance from one end to
the other.  If it isn't << 1 ohm, it's asking for trouble.  

4) Re-run the measurement after tweaking the 10811's frequency to bring it
in line with your reference.  Adjust the frequency in situ with your DMTD,
so the load is the same.

-- john