# Do I really need 1 square inch per watt ???

Discussion in 'Electronic Basics' started by [email protected], Nov 13, 2005.

1. ### Guest

I found these really cheap (\$0.29) 10 ohm 20 watt power resistors in a
TO-220 case. They are a TEPRO TO-220 style thick film heat sink
devices. I was planning on combining these together in various
configurations to make a power resistor in the neighborhood of about
10-30 ohms 500-1000 watts. I plan on mounting a bunch of these on an
aluminum plate as a heat sink. Using the 1 in^2 per watt rule of thumb
I would need 500 in^2 - 1000 in^2 of surface area. Is this really true
or could I get by with a smaller plate and a cooling fan? For the 500
watts that would amount to about a 24" x 24" aluminum plate. What
would be the best heat sink solution? I really like the small size and
price of these TO-220 power resistors, but are they going to be too
much of a hassle to heat sink? Should I maybe go with a bunch of
\$0.69 15 watt square prism shaped power resistors? Any advice would be
greatly appreciated. Thanks

2. ### Peter BennettGuest

For that power, you do need a REALLY GOOD heatsink - like a large
aluminium extrusion (or two) with lots of fins, and a fan.

A portable electric heater is about 1000 watts - that's the amount of

--
Peter Bennett, VE7CEI
peterbb4 (at) interchange.ubc.ca
new newsgroup users info : http://vancouver-webpages.com/nnq
GPS and NMEA info: http://vancouver-webpages.com/peter

3. ### Doug McLarenGuest

| >Using the 1 in^2 per watt rule of thumb
| >I would need 500 in^2 - 1000 in^2 of surface area. Is this really true
| >or could I get by with a smaller plate and a cooling fan?
....
| For that power, you do need a REALLY GOOD heatsink - like a large
| aluminium extrusion (or two) with lots of fins, and a fan.

The new dual core Intel paxville Xeon cpus use nearly 200 watts per
chip at full power. All that heat goes through a square that's about
one inch square into an elaborate heat sink, with a big fan. If the
fan fails, the chip burns up in a few seconds.

(At least the older Athlons did. I've heard that Intel P4s are better
about throttling back on the power when they overheat, but even so, at
200 watts, I wonder if it could shut down enough and quickly enough to
save it's own skin.)

Just an example of how not to do things

4. ### Bob EldredGuest

If you want to waste that amount of power, you are probably doing it on a
short term intermittent basis for testing purposes and not continuously
unless you are making a heater. Is that correct?

For that, you can mount several of these resistors on a small aluminum plate
and immerse the plate with resistors into a coffee can full of water. This
will dissipate a KW easily for a while. As the water boils off, add more
water. I've done a 10KW load this way and it works great. Depending on the
voltage you may or may not have to insulate the leads and/or the can. For
high voltages, oil can be used instead of water but water works better. It
looks like your voltage is in the 100 to 200 volt range. I'd insulate the
leads with RTV silicone goop and isolate the can from conduction to anybody
or metal object. Be safe.
Bob

5. ### Jasen BettsGuest

so \$15-\$29 in parts plus heatsink, paste, bolts, interconnect, and 50-100
holes to measure mark and drill. any reason why you're not planning on using
a hank of nichrome wire and a fan.
You may be able to run the resistors at a higher temperature than
transistors can handle.
one with fins and a fan?
go with a heating element if you can find one with the right impedance,

Bye.
Jasen