Hi Markus,

I know nothing about FEI but have some experience with other brands.

All Rb clocks wear out, because the lamp gas is lost to adsorption
or diffusion.  Life is usually 5 to 20 years, depending on design.

It should be possible to get a service manual, or at least a schematic,
from FEI.

I assume that you verified that both supply voltages are well within
tolerance.  From a cold start, measure the current drawn by the 15-18V
input.  It should be close to 2 A at start, and drop to about 500 mA
as it heats up.  If the current does not drop as expected, there is
something wrong with one of the ovens.  If your power supply current
limits at less than 2 A, the unit may eventually get running but it
may take much longer than 5 minutes.  There are at least two oven
circuits, because the lamp and the resonance cell operate at
different temperatures.  Some designs use the same oven for cell
and crystal; some use three ovens.

Most units have a "lamp monitor" output (proportional to photocurrent).
The 5650 does not have this on an external pin, but there is probably
a test point inside.  If there is no photocurrent, check the lamp
temperature and its drive voltage.  If they are both ok, the lamp is
probably bad.  If there is photocurrent but no periodic dips, check
the cell temperature and check that RF is present.  Also see if the
10 MHz signal is way off frequency, e.g. 10 ppm or more.  If these
are all ok, the cell could be bad, but that is not a common failure.
If there are periodic dips, there is a problem with the lock circuit.

If you conclude that your unit has a bad physics package, perhaps
you can find one very cheaply with bad electronics, and combine the
two to make one good clock.  The opposite should be even easier,
as there are lots of oscillators with bad lamps :)

73,

Stewart KK7KA

To: <Undisclosed-Recipient:>
Sent: Thursday, December 25, 2003 4:27 AM
Subject: LF: Rubidium Freq Standard


Hi ...

I got a Rubidium Freq Standard, 10MHz, manufactured by Frequency Electronics
Inc. Its model is FE-5650A, still being manufactured. Aparently it doesn´t
lock after 4 or 5 minutes ... warm-up etc , 10MHz signal is there but
unlocked.

Does anybody knows something abt this device ? The manufacturer website have
some tech docs explaining how it works ... there´s a DDS inside etc but I
guess the main rubidium cell is not working (a 6.8GHz FLL).

73
Marcus, PY3CRX/PY2PLL
GG66rf
http://py.qsl.br
Active from "DC" to 24GHz