On Apr 25, 5:49*am, "Arfa Daily" wrote:
"Gareth Magennis" wrote in message

...





"James Sweet" wrote in message
news:Kn4Qj.6369$Mm1.4242@trndny07...

One thing to watch out for is the hysteresis in these devices. *The
device I pointed out will turn the fan on at 30 degrees, but it won't go
off again until the temperature drops to 20. *If your room is warmer
than that it will never turn off!

Gareth.

Could build something. I recall seeing some temperature controller
circuits that used a diode as the sensing element. Should be easy to
adjust the hysteresis by selecting component values.

Well this is the range of those devices:

http://uk.farnell.com/jsp/level5/mod...=en/212166.xml

Note that the opening and closing temperatures on some of them are the
wrong way around, but I reckon that the normally open switch that will
switch on the fan at 55 degrees C and turn it off at 35, part no. 1006853,
would be ideal. *55 degrees is not going to cook any electronics and it is
unlikely that room temperature is ever going to get to 35. *Case solved,
one component, no design work required, and extremely reliable.

Gareth.

If you do go down that route, I think that the sensor needs to be inside
whichever piece of your equipment generates the most heat (with the
attendant potential electrical safety issues which that may cause on a piece
of kit with a switch-mode power supply ...) because if the temperature is
going to reach 55 deg inside the glass-doored cabinet, then it's going to be
a whole bunch hotter than that inside the equipment, and trust me, that *is*
electronics 'cooking' temperature. Sort of electrolytics on toast if you
like ... *:-)

For me, your original power slave switch idea was the better option.

Arfa

Yes, I think I agree - when I first saw pictures of those thermal
switches, the first thought that occurred to me was: "Now where am I
going to bolt one of these onto my components..." A bit too invasive,
I think, at least for me, but it is a great idea otherwise.