# Thermal resistance of an identical but larger heatsink

Discussion in 'General Electronics Chat' started by Norfindel, Jun 23, 2010.

1. ### Norfindel Thread Starter Active Member

Mar 6, 2008
235
9
Hi,

I have the specifications of a heatsink, including the thermal resistance for a certain lenght, but it can be buyed in longer sizes (it's likely an extruded heatsink).

My question is: knowing the thermal resistance of the 75mm long heatsink, how to calculate the thermal resistance of a longer one?

This is the page of the heatsink manufacturer, by the way: http://www.disipadores.com/ingles/media_potencia.htm

Thanks.

2. ### mcgyvr AAC Fanatic!

Oct 15, 2009
4,789
974
inversely proportional to the square root of the length

Oct 15, 2009
4,789
974

Apr 5, 2008
15,799
2,386
5. ### Norfindel Thread Starter Active Member

Mar 6, 2008
235
9
Thanks. That table is very useful.
They say that it's not linear, because the heat source is in the center, and it's colder the farthest away from it, and it makes a lot of sense.

Probably that means that if using several, correctly distributed devices, the factor will get closer to 1 again, right? If i have a big heatsink with two transistors, it would work the same that if the two transistors where on separate heatsinks of half the length?

6. ### mcgyvr AAC Fanatic!

Oct 15, 2009
4,789
974
Should be pretty darn close assuming your larger heatsink was selected so that heat from one device is not effecting the other significantly.

7. ### timrobbins Active Member

Aug 29, 2009
318
16
I don't have empirical data to hand, or a good book reference, but you have to significantly bias the position of the devices to a point 'lower' in the heatsink if only relying on natural convection. This is easy to comprehend, as the air flowing over the upper section of the sink has increasing temperature, and the rate of change of temperature with distance in the sink body is less as you move 'higher'. This equates to an effective length of heatsink that is shorter.

For diodes and thyristors, the nominal design case temperature will be much higher than for more exotic devices such as Fets and igbts, in which case the heatsink body is significantly higher in temp and so the effective thermal resistance of the sink is lower. One needs to always check on the test conditions for the rated sink thermal resistance.

Ciao Tim