> > Not as elegant, but an LM35, comparator, and a MOSFET come
> to mind
>
> True. I could throw a PIC in there and PWM the fan. Remember the
> Microchip application note on measuring temperature based on how
> long it takes the WDT to time out?

Harold, I've used that method several times and it works fine, if
you want or need a controller with added smarts (for example
compensating for thermal inertia). Although WDT is specced fairly
coarsely in the d/s, in a particular chip it's quite predictable and as-
near-as-dammit linear. You can take the resolution of WDT
measurement down as fine as you like for the situation. For a fan
simple on-off might be sufficient for general temperature control
rather than PWM

Unless you're heart's set on finding a thermostat, an electronic
method may be the quicker option

I just noticed that my web pages are still down after the service
problems I had a couple of weeks ago. I'm trying to resolve that

Maybe by the time you get this, perhaps later today, the link below
will be working, and you can see the WDT vs temperature figures

http://home.clear.net.nz/pages/joecolquitt/0wdttemp.html

Although that was done with an F84A, all PICs I've tried are very
similar

-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist