# Thermal Resistance of PSU's heat sinks

I have a non functional PC power supply unit, and I would like to keep its heat sinks, but they have no model or part number in them. Is there a way to estimate, calculate or measure their thermal resitance? The PSU is a Sentey BCP 450-OC, here is a picture of the heat sinks

• A look in heatsink catalogs will give you a reasonable idea. – Russell McMahon Feb 15 '20 at 1:04

## 3 Answers

Measuring the effective heat resistance should be pretty straightforward: you'd affix load resistors to where you'd later put the components that should stay cool. You work in a controlled temperature environment (i.e. indoors at 20°C).

You put a measurable amount of power through the resistors (e.g. by using a lab voltage supply and measuring the voltage over and current through the resistors). You measure the temperature and wait until you've reached a stable state.

The difference of resistor temperature to ambient divided by the power is your thermal resistance.

But: honestly, I don't think this is worth it. Both the shape and the fact they're fixed right in front of a fan means that they are to be used with forced cooling, i.e. with a fan in a specially shaped enclosure. You'd have to measure in the enclosure and with the fans you're planning to use.

Is there a way to estimate, calculate or measure their thermal resistance?

Yes. I am assuming that you want to know the heatsink to ambient thermal resistance.

The calculations are complicated, somewhat easier for forced air (fan), than natural convection. I wouldn't trust any calculation that I would perform without verifying it with a test. So, ...

It is easiest to test. This heatsink is designed for forced air, it won't be very good without a fan.

Attach something to the heatsink that you can be made to dissipate a known amount of power. A chassis mount resistor will be a good choice. Measure the temperature rise of the heatsink, subtract the ambient temperature, divide by the power.

The fluid aerodynamics of forced air velocity overwing surface determines the thermal resistance coefficient in °C/Watt for case temp. rise.

The air velocity over the surface, not the fan exit, significantly lowers the resistance by some constant, k (< 1) times the Rca, Resistance case-ambient. This is more of an aerodynamic design requirement. One could come up with a more accurate answer, but based on my exposure to heatsink design and testing I would estimate 8 °C/W and with that airflow 1~2 °C/W.