[NLRS] DB6NT PTC crystal heater

[Doug Reed]

At 02:43 AM 2/12/2003 -0600, you wrote:
>Can you recommend a specific PTC for this task? I have an LO
>crystal in a 2m transverter here that I would like keep a
>bit warm. The drift isn't excessive but I figure a PTC will
>greatly reduce it easily enough.
>Thanks!  >Phil

The first article I saw about doing this is quite old now but it seems to
be a common tactic. On the other hand, the possible parts to use come down
to just a few choices. 
Digikey# KC001P-ND - 50 ohms 30 degrees C
Digikey# KC002P-ND - 50 ohms 40 degrees C
Digikey# KC003P-ND - 50 ohms 50 degrees C

All three parts are listed as 25 volt parts but for our needs that is more
of a maximum voltage suggestion and running it at 12 volts will simply take
longer to heat. The 50 ohm PTC will draw 250ma as it heats the crystal to
turn-over temp. At turn-over temp the resistance will rise and change just
enough to maintain the temperature. 

You need to choose the turn-over temperature. KC001 is a warm summer day
while KC003 is the standard temp for a crystal oven. The common wisdom says
if you try to operate a crystal that was cut for room temp at high temp,
you'll make the drift worse. Ditto if it was cut for high temp but runs at
low temp. I haven't personally verified that problem. 

For the most part and for what you're trying to do, the KC001P-ND is
probably an adequate choice. It is slightly above room temp, near the temp
of a warm day, and well above freezing. As long as you don't mind the
possibility of a little drift on hot summer days when the equipment is
sitting in the sun, KC001P is for you. 

KC002P is better if you need a little more high temp range and it was
probably the original value suggested in the article. If the equipment is
out in the sun or mounted in an attic or furnace room, this is probably the
best option. 

The KC003P should be used if the crystal was cut for an oven. You can of
course try anything. It probably will not damage the crystal but you'll
need to test it to see if the frequency remains within range and the drift
is tolerable. 

The original article suggested removing one lead from the PTC and soldering
it direct to the can of the crystal and adding a ground wire from the can.
This would provide the best heat transfer. Just be sure you don't heat the
can so much it comes loose from the crystal base or blow out the solder
plug where they sealed the can in inert atmosphere. 

An alternative is to use a good glob of heat sink compound for the thermal
joint. I suppose you could also try making a big solder blob on one side of
the PTC to give a flat surface for heat transfer to the crystal. (Buy more
than one PTC, just in case.<g>) If you do this you'll also need a
mechanical clip to hold the PTC to the can. Just bend a metal clip to fit
the can and PTC and be sure to add an insulating pad between the clip and
the positive lead of the PTC. The negative lead of the PTC should be
against the crystal but be sure the ground wire from the PTC goes to the
board. Don't rely on ground thru the crystal. This is the technique they
intended for the DB6NT. EF Johnson, Yaesu, and Wilson all used the metal
clip and PTC approach in some of their radios. 

And of course I'd suggest trying to add some sort of insulation around the
PTC-crystal combination. It will take a lot longer to stabilize if the PTC
is trying to heat the whole unit, not just the PTC. I expect the thermal
break is why the DB6NT has just a small ground trace to solder to the
crystal can rather than soldering to the full ground plane right next to
the crystal. 

That should be all you need except to find a few more parts to make the
Digikey minimum order. I believe Mouser(?) doesn't have a minimum order but
you'll have to get the mfgr part number from the Digikey web page to look
it up on the Mouser web page. On the Digikey web page, just do a search for
KC002 PTC or KC00 PTC if you want a few more to look at. Price is $2.54
each Qty 1.......

Hope this was what you wanted. 

73, Doug Reed, N0NAS.