Hello,

Calculated Power dissipated from a LM317 voltage regulator (9V output, 0.5A load current) = 4.5W
Ambient temperature, 30 degrees
Max. junction temp. (from data sheet) 125 degrees

125 - 30 = 95
Junction-to-ambient thermal resistance = 95/4.5W = 21 degrees/Watt

With a Heatsink added, specified thermal resistance of 3.3 degrees/Watt:

3.3 x 4.5W = 15 degrees.
30 + 15 = 45 degrees = heatsink temperature = correct?

 

No. You left out consideration of Vin in your calculation. There may be other errors. I stopped looking at the first occurence.

http://www.reuk.co.uk/LM317T-Heatsinking.htm

http://www.eevblog.com/2010/08/15/eevblog-105-electronics-thermal-heatsink-design-tutorial/

 

The power dissipated by the regulator is the voltage across it times the current through it. So you want (Vin-Vout)*Iout.

Assuming that that were to still come out to 4.5W so that we can go forward.

Is the 3.3 K/W (kelvins/watt, which is the proper unit since degrees celsius is not an absolute scale, but they are the same thing and degC/W is more commonaly used, so don't worry about it) the specified thermal resistance from the junction to ambiant with the heatsink in place, for just the thermal resistance of the heat sink?

If so, then the calculations are correct except that the 45C at the end is not the heatsink temperature, it is the junction temperature.

Also, your calculation of 21C/W should be noted that it is not the junction-to-ambient thermal resistance, but rather the maximum permissible value of it.

I would recommend carrying your units throughout your calculations.

 

Thanks.

My next question was going to be how can I calculate the case to junction temp. now with the heatsink used? It looks like I inadvertently answered it: 45 C.

 

No, 45C is the junction temperature (assuming 4.5W dissipated in the device, which is not going to be the case here unless the input voltage is 18V). The case-to-junction temperature is the difference between the temperature of the junction and the temperature of the case. To find that, you need either the junction-to-case thermal resistance (which is generally assumed to be the same as the junction-to-ambient thermal resistance without a heatsink for many packages) or the case-to-ambient thermal resistance for the heat sink.

Some IC's (including many TO-220 packaged parts) have a very low junction-to-pad thermal resistance where the pad is an exposed thermal pad specifically designed for heatsinking. If you're using a heatsink that mates well to that pad, then you can really keep junction temperatures in check with a lot less heatsinking effort.