[time-nuts] FE 5680A Warning - THE CLUES

WB6BNQ wb6bnq at cox.net
Sat Nov 12 22:22:36 UTC 2011


RESENT WITHOUT ATTACHMENTS


Fellow Timenuts,

Here are the clues that will answer some of your questions concerning
the “newer
version” FEI Rubidiums and also it’s “Kissing” cousins from Symmetricom;
the newer
series XPRO and SA22c.  It appears both companies have came with up with
same idea
at the same time or copied each other to a large extent.

In the old days the “Atomic” loop was in the analog domain.  The loop
output signal
was a fixed frequency and any “other” frequency was derived from that
fixed
frequency outside of the loop.

These newer devices have incorporated a different paradigm using a
digital control
process for the actual “Atomic” loop.  Thus the output frequency is
truly part of
the loop rather then derived from the loop.  The AD9832 provides but a
fraction of
the control loop error signal allowing for adjustments to be made in
steps of parts
in 10 to the minus 13 (10E-13).  The “C-field” adjustment is done with
either an A/D
or some digital potentiometer, but may also have some possible external
access.

The end result is that the output frequency is factory selected by
changing the
frequency of the main crystal oscillator.  This makes it very hard to
change a
Rubidium with an “odd-ball” frequency because of internal firmware.

All of this is gleamed from two (albeit cryptic) documents from
Symmetricom.  One
document is on their XPRO unit and the other is on the SA22c unit.  In
the XPRO
document carefully read pages 15 through 19 (PDF pg 21-25).  Then
carefully read the
SA22c document pages 20 and 21 (PDF same).

Keep in mind that the info is written so as to not give away so-called
trade
secrets.  So you have to do some critical thinking and reading between
the lines to
get the whole picture.  The XPRO document block diagram on page 16
(PDF-pg 22) is
the primary key to all of the above.

I am including the two aforementioned documents as attachments. 
Hopefully they will
get through.  If not I will ask Didier (KO4BB) to host them on his site.

Bill....WB6BNQ



More information about the time-nuts mailing list