[time-nuts] Temp control of LPRO

[Neville Michie]

Hi,
I have just commissioned a temperature control for my LPRO rubidium  
oscillator.
When I read the specs I noticed that at 18V supply and about 40*C the  
unit only requires about
8 watts. That is getting close to what my stand-alone power source  
can maintain.
Running at a higher temperature takes a little load off the ovens on  
the lamp and the detector,
but it does have a penalty in MTBF. At 40*C I thought the trade-off  
was worth it.
Now the local crystal oscillator is not temperature controlled and  
its control voltage drifts widely.
So I bolted the LPRO to a plate of aluminium about 1/4 inch thick. On  
the other side of the aluminium plate
I bolted heat sink fins about 3/4 long over the whole length.
The whole unit mounts inside a piece of about 3X4 aluminium  
extrusion. At one end in a little plenum
chamber I mounted a12V  1 watt 40mm brushless DC ball bearing fan, in  
the AL plate I drilled a
deep well in which a bead thermistor lives.
The whole thing is insulated in 1/2 inch plastic foam sheet.
When the fan running on 12V the temperature is held to 4 *C above  
ambient. At 8 volts this rises to about 7*C.
The fan is quite unobtrusive on 8 V.
The thermistor is in a bridge circuit and switches the fan in a PWM  
control mode with a 50 second cycle time.
The surface of the plate has less than 0.1*C variation over a day  
with no sign of the temperature cycling.
The oscillator control voltage now rises on startup and then becomes  
rock steady.
The next step is to get on with the building of the gear to measure  
how well the LPRO now performs
at constant temperature.
Cheers,
Neville Michie