[time-nuts] HP 5370B low frequency modulation

Magnus Danielson magnus at rubidium.dyndns.org
Sun Jul 22 19:03:37 EDT 2007

Hi fellow time-nuts,

As I was going to make a casual measurement of the HP 5370B 10 MHz output, i.e.
that of its HP 10811-60111 OCXO, I came to see an unexpected modulation. It has
a slow modulation frequency modulation of about 5,2 mHz or there abouts. Quite
digital/squarewave-like I might add. Not actually squarewave, but rather some
form of PWM signal. Another interesting aspect was that it had a fairly steady
high, abrupt fall down a slight recovery, a jump upwards and then fall up to
the steady high level. The high frequency was the longer part of the time.

The cycle-time of these are low, for the 200 5 sec samples I took, I had about
10 cycles or so. So it is about two minutes cycles. The rate varied greatly.

I have not seen any specific details in the HP 5370B service manual about where
the HP 10811-60111 EFC is hooked up to anything, but I have noticed that more
or less everything is grounded on the connector. Now, if there is low-frequency
modulations of any large current there, it may kreep in. But it seems a bit
high for that.

I used both a Pendelum CNT-90 as well as a HP 5372A to measure this. I used
both the Cs and Rb as reference and it made no difference. The CNT-90 between
the two showed only noise, so this quick and dirty three-cornered hat certainly
pointed its crookied finger at the HP 5370B. The HP 5372A was using its own
well heated 10811-60111 as reference but that told the same story.

I might add that the 5370B was running in "cabed" condition, i.e. top and
bottom lids where off as well as the right side. The top lid needs to be off if
you want to have a chance of trimming it. The 5370B has been in bypass 9 days,
so I consider the 10811 fairly heated.

Has anyone else seen this? Tom?


More information about the time-nuts mailing list