[time-nuts] Oscilloquartz 3210 Cesium Standard

Björn bg at lysator.liu.se
Mon Sep 29 20:09:49 EDT 2014

Hi Chris,

Nice work! 

I can't help you with the faulty unit. There are others on the list with deeper insight in that OSA Cs core. Hope they have time to chime in.

But I do have a telco variant of the same Cs core running in the basement. Its running continously (ocxo and ion pump) but its not burning the last fumes of Cs in the tube. This is since I don't see access or money available for a replacement tube.

As you noted. With ocxo warm and cosy and the ion pump running it does not take long for Cs lock when needed.

A mode where you can use the unit as premium ocxo (bva :-) ) standard that can be calibrated against the cs tube periodically as needed would suit my situation.

However there are (at least) two mods needed. Because unless the cs tube burnes, there is no power to the

 1)  output buffer card. And thus no output signal.
 2) potentiometer adjusting the bva frequency. Thus without cs tube running the bva is off freq.

This design might have been a safe one in the original market.

But I think modding the unit to let it run mostly as a very stable crystal standard is better for hobbyist use.

Have others done these kind of mods to prolong the Cs tube life? 

Kind regards,


<div>-------- Originalmeddelande --------</div><div>Från: Chris <syseng.greenfield at btconnect.com> </div><div>Datum:2014-09-29  22:43  (GMT+01:00) </div><div>Till: time-nuts at febo.com </div><div>Rubrik: [time-nuts]  Oscilloquartz 3210 Cesium Standard </div><div>
</div>Another update on progress:
Now at the stage where two of the three 3210's are working. The fault 
that was blowing the line fuse turned out to be a shorted reservoir cap. 
Managed to find an Ebay seller in Greece with exactly the same voltage, 
capacity and outline, though different manufacturer. The caps have a 
stud mount at the base of the can, so not easy to find, but bought four 
at just over 3 ukp each and replaced both caps in the psu, leaving two 
for spares.

From cold, the spec is up to 90 minutes warmup time, but if the OCXO 
and pump have been running for some time, both units lock up within 4 
minutes. Have done some setup. For example, normalised the gains and 
offsets in the preamps as per the manual and both units show the 
expected preamp level of ~165mV and 2nd harmonic amplitude of 9-10 on 
the meter, so assume both tubes are in good condition.  None of the 
settings were very far out.

Took a set of measurements from one of the good units and the faulty for 
comparison, with similar results, Good unit as follows, but faulty unit 
more or less the same:

Ion pump = EHT ok, meter = 0
Electron multiplier EHT voltage = 1800 volts, (100Mohm hv probe on o/p 
Cesium oven = Not open circuit, voltage = 5.5 volts
Ioniser = not open circuit, 1 volt p-p square wave at ~26KHz
Synthesiser output = 12.6317715, follows variation in OCXO and in lock
Backplane test points, tpb, tpc = 8.2 volts p-p square wave, ~137 Hz.
Multiplier 12.6... MHz input = 2v p-p (Scope)
Multiplier 137Hz input = 2.25v p-p (Scope)
Multiplier 180MHz output = +26DBm (HP 3406A rf voltmeter + 20DB attenuator)
Microwave tap on waveguide = -14.5DBm, threaded attenuator works 
(HP432A, 478A)

The 3406 and 432 are quite old and the sensor head is even older, but 
should be reasonably accurate. The 478A head was coupled to the 
microwave tap using a ~3 inch length of rigid coax, so not much 
attenuation. Hopefully, these figures may help others trying to debug 
these standards, but if there's anything i've missed, please let me know.

For the faulty unit, even with all the levels as expected, there is 
still no signal, nor 2nd harmonic, even with the preamp gain turned up 
to max. Connected a Fluke electrometer in series with the preamp input 
socket, but the current is at least 100 times down on 1nA. Looks like 
the tube is completely lifed, or has an internal fault. /FX:  Have 
visions of a sad, faithful and slowly dying 3210 left for decades in a 
rack,  in the dark, powered up, but long forgotten after the accurate 
reference it provided ceased to be used anywhere in the organisation. A 
ship adrift for thousands of years, but all systems still at least 
partially working and waiting to be discovered :-).

So what else am I missing in this puzzle ?...



time-nuts mailing list -- time-nuts at febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

More information about the time-nuts mailing list