[time-nuts] Z3805A cooling requirements?

[SAIDJACK at aol.com]

Hi Bob, et. al.,
 
this discussion prompted my interest to see how a 10811-600160 unit that we 
 have sitting here on the bench would react to airflow changes. I had a 
thin  layer of anti-static shipping material around the unit, and I opened up 
one  side of that layer and then pushed air into that side.
 
The unit is from a 53132A counter, and has the 53132A OCXO PCB with voltage 
 regulator attached to it, and is being fed from an external stable power  
supply.
 
Please see the frequency plot attached, I used an air gun to push air onto  
the unit for about 60 seconds or so. The unit had quite a large and 
immediate  change in frequency, up to about 3.5E-011 and then down to -4E-011 at 
which  point I captured the image (while it was still drifting negative). I 
did not  wait until it settled back (will take probably 30 minutes or more I 
would  guess), but it went quite a bit more negative after I captured the  
plot.
 
This airflow sensitivity is probably a combination of the circuitry on the  
support PCB, the OCXO, and the external power supply.
 
The stability before the unit was exposed to the airflow was quite good as  
can be seen in the left half of the plot, and I think this illustrates how  
adding a bang-bang type airflow to these oscillators can worsen ADEV 
performance  significantly, albeit we are talking 10's to 100's of parts per 
trillion here..  I would much rather see the unit perform as it did initially in 
my test setup  without the airflow though.
 
bye,
Said
 
 
 
 
 
 
In a message dated 12/16/2012 18:11:02 Pacific Standard Time, lists at rtty.us 
 writes:

Hi

… and what I'm trying to say also comes from the real  world...
If you start putting "stuff" on an OCXO, be careful about the case  
temperature and how the OCXO is spec'd. A few mm of dead air can make a good  
insulator. That can boost the case temp quite a bit.

Bob

On Dec  16, 2012, at 9:06 PM, Magnus Danielson <magnus at rubidium.dyndns.org> 
 wrote:

> On 12/17/2012 02:47 AM, Bob Camp wrote:
>>  Hi
>> 
>> The gotcha here is that an un-cooled piece of gear  will heat up and 
cool down as it's work load changes.  There is no "magic  bullet" that keeps 
the temperature constant with zero airflow in a normal  design. Yes, I'm old 
enough to remember oil cooled computers. Still no  constant temperature and 
you have turbulence.
> 
> I agree that  there is no "silver bullet", but my point was that 
sometimes you kill one  property when you apply a solution to another problem. I am 
very well aware of  heating problems and cooling my components, as this is 
part of my real world.  But rather than isolating the full box, I'm talking 
about the TCXO or OCXO.  Just putting a small wind-shield over it changes 
things a lot at  times.
> 
> Cheers,
> Magnus
>  

-------------- next part --------------
A non-text attachment was scrubbed...
Name: 10811_60160.pdf
Type: application/pdf
Size: 20874 bytes
Desc: not available
URL: <http://www.febo.com/pipermail/time-nuts/attachments/20121219/78991ca3/attachment-0001.pdf>