[time-nuts] FE-5680A heat sink

Bruce Griffiths bruce.griffiths at xtra.co.nz
Mon Jun 8 23:21:52 EDT 2009


Leigh L. Klotz, Jr WA5ZNU wrote:
> Chuck,
> This device is a pull from a larger system, probably a cell site. 
> It's clearly designed to be mounted on something, as it has lots and
> lots of holes around the edge.  I suspect there's some thermal
> management that's missing.  The FEI sheet gives "typical" data for the
> 0-50C range, though presumably that's ambient temperature.
> I don't want to cool the physics package per se, but I do want to at
> least approximate what kind of thermal solution ought to be supplied.
> Quite a few of these have been sold, and I've gotten good advice from
> others on this list about calibration.  I'd hoped that someone would
> have experience with the thermal management.
> So far I've seen
> http://www.dd1us.de/Downloads/precise%20reference%20frequency%20rev%200_4.pdf
> which shows (but doesn't describe) a heat sink on the bottom.

If the case is Mu metal or similar then the  modification shown in the
above article will likely destroy its magnetic shielding properties
(unless its annealed in a hydrogen atmosphere at 400C for several hours
after machining).
The unit is designed like competitive units to be bolted to a heatsink
or panel. The trick being to maintain the baseplate temperature within
the specified temperature range (-5C to +50C).
Since the power dissipation is low (11W @ 25C) you wont need an
extremely low thermal resistance heatsink unless your ambient
temperature is relatively high. If the ambient temperature is 40C then a
1C/W heatsink should maintain the base plate temperature below 60C.

Just mount the 5680A on your heatsink and monitor the steady state
heatsink temperature (with a thermocouple, RTD or other contact
thermometer - could even use photochromic temperature sensing strips)
both with the FE5680A upside down and the right way up.
The infrared thermometer is unreliable unless its reading is corrected
for the emissivity of the target in the 20um infrared region.

> Leigh.
>> I'm puzzled.  I admit that I don't have a lot of experience with
>> Rb standards, but I do have a bit of experience with the HP-5065A
>> Rb standard.
>> In the 5065A, the entire physics package is enclosed in an oven.
>> Assuming that your Rb is the same, and I believe it is, your plan
>> to force the physics package to run at room temperature is just going
>> to make the oven work harder in its never ending quest to maintain
>> stable temperature.  If you monitor the current draw of the 5680A,
>> you will probably see that it goes up when you put a fan on it.
>> Sometimes, you just have to let electronics run hot.
>> What did the manufacturer suggest?
>> -Chuck Harris
>> Leigh L. Klotz, Jr WA5ZNU wrote:
>>> I've attached an SMA connector to my FE-5680A and built an external
>>> linear power supply with a TO-3 7815.
>>> I ran the device today for about half an hour, and used an infrared
>>> sensing thermometer to measure the external case temperature.
>>> It got up to 48 C externally in the physics package area.
>>> I put a muffin fan on top and it brought it down to 38C but I don't
>>> think this is a good plan because I worry about the effects of the
>>> fan's magnetic field on the Rubidium system.
>>> I found in my junk box a finned Aluminum heat sink that's exactly
>>> the same size as the FE-5680A and plan to tap it around the edges
>>> for 4-40 hardware to attach to the many screw holes.  Even so, this
>>> heat sink will be on the bottom, so the FE-5680A will have to be
>>> operated upside down for this to help.
>>> Has anybody got good thermal management solution for this device? 
>>> This is the one currently selling on eBay in the 25x88x125mm chassis.
>>> Thanks,
>>> Leigh.
>>> _______________________________________________
>>> 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.
>> the instructions there.
> _______________________________________________
> 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