[NLRS] Simple temp controller for crystal oscillator

[Dr. Gerald N. Johnson]

Nice circuit. I like the TL431C and its variations. Most PC switching 
supply designers like it also, its very often used to sample the 5 volt 
output voltage and to drive an optical isolator to control the switching 
IC on the primary side. If you want a zener with a theoretical sharp 
knee below 6 or 8 volts it IS the only way to achieve that. I have used 
it as the primary regulator in a couple power supply designs (one 
published in CSVHF, 2010, IIRC) to control power FETs. I have also used 
it in the regulator for the steam turbine on the Boone and Scenic Valley 
RR Chinese locomotive, regulating a shunt field 50 volt generator down 
to 32 volts. Its essentially an op amp and a 2.5 volt reference that 
will run from 2.5 to 30+ volts and if the input is greater than 2.5, its 
output pin is turned on and drags down to whatever voltage would pull 
the input down to 2.5 volts. Its data sheet says minimum output current 
should be 1 ma for a sharp knee but G(M)3SEK has used it in a regulator 
for 4CX250B screens and reported it works fine at 100 microamps. Still 
has a sharp knee. When used as a zener there is a range of bypass 
capacitors that will make it oscillate. When I use it with power FETs, I 
have found a 910 pf capacitor from output to input slows it to a speed 
the power FETs can keep up with so it doesn't overshoot.

I've been using foam around crystals for 49 years. I did that to the 
calibrator crystal in my 75S-3B that I bought new in the spring of 1964. 
I used white foam and I glued on aluminum foil as a reflector of radiant 
heat on the side of the foam next to the nearest tube. I achieved good 
enough stability that I made a FMT in the 70s with less than 1 ppm error 
on 40 meters using my micrometer slide rule eyeball to interpolate the 1 
KHz marks on the dial scale.

Dow blue board hunks around the crystal in my 10 GHz transverter helped 
it a lot. I could tell the difference with just random variations from 
air currents caused by me moving around the bench with the crystal 
exposed while listening to a crystal harmonic in the 1GHz region. Its 
not yet perfect, this circuit would probably do better, but that rig is 
not going to get attacked when its working as good as it did last year.

73, Jerry, K0CQ

On 3/7/2013 5:38 PM, Doug Reed wrote:
>
>
> <http://pensacolasnapper.blogspot.com/2011/03/temperature-controllers-for-qrss-part-2.html>
>
> This link has a fairly simple circuit for temperature control that can
> be added to any crystal or crystal oscillator. It certainly isn't as
> accurate as the DB6NT temp control but it is probably a bit more
> accurate than using a simple PTC resistor for temp control. It is a
> simple 4 part circuit and it probably will do a pretty good job. You
> might want to save a copy of the article for future reference.
>
> One of the other blogs on the same page mentions some tests he ran
> using a crystal can oscillator and putting it in a small foam block.
> The foam, even without a heater, stabilized the oscillator quite a
> bit. This would have possible application for making a cheap
> weak-signal source for the microwave bands.
>
> 73, Doug Reed, N0NAS.
>