Heat Sink Calculation

[BootstrapCap]

Hello All,

I have a question regarding calculations on heat sink. It is pretty straight foward just like explainned everywhere (i.e **broken link removed** )

First you need to get the Maximum thermal resistance, which it is gotten by diving the max temperature increase by the power.

Rja < (Tmax - Tamb ) / P(W) (1)

And in case of MOSFET we get that:

Rsa = Rja - Rjc - Rcs (2)

where R´s are the thermal resistance for sa: Sink to ambient; ja: junction to ambient; jc: junction to case ; and cs: case to sink.
let´s assume that equation 1 gave us Rja ≦ 10 K/W and we know that a mosfet gives us Rja≈60K/W without any sink which is greater. Inoder to get a new Rja, it is possble to add a sink by using equation 2. Asusming Rjc = 0.5 K/W and Rcs = 1 K/W, we have:

Rsa = 10 - 1 - 0.5 = 8.5K/W (3)
An that is the thermal resistance of the sink.

Alright, with all that said. What happens if my Rja max was 0.2K/W which is avalue lower that Rjc which I cannot do anything about? The temperature will keep increasing until it reaches the final temperature which is higher than 150C. Thta does happen and predicted.


To overcome this, I see 3 options:
1. New part with lower Rjc
2. Allow part to grow hotter.
3. Dissipate less power.

Here is the other catch, by attaching metal blocks onto the tab of the mosfet ( therfore, reducing thermal resistivity ), it only reaches 115C. How can a heat sink, that cannot be calculate, exist?

Any insight guys?

 

[chuckey]

" it only reaches 115C.", this is probabley too low a measurement, also with such a small heat sink your measuring probe could introduce significant extra heat dissipation area.
Frank

 

[BootstrapCap]

" it only reaches 115C.", this is probabley too low a measurement, also with such a small heat sink your measuring probe could introduce significant extra heat dissipation area.
Frank

The point is how can the temperature stabelize? being Rjc higher than the needed Rja for a safer final temerature.

we are using a thermal camera, btw.