Common heat sink calculation

[Munyua_wa_Mugo]

I am focused in using the common heat sink for my bridge rectifier , boost diodes and the IGBT. Below is the calculation of each of the dissipation;
P_DIS_Bridge=18w

P_Dis_igbt=24w

P_boost_diode=6w

RJC_BRIDGE=0.6DEGREE/W
RJA_BRIDGE=1.63 DEGREE/W

RJC_IGBT=0.58 DEGREE/W
RJA_IGBT=1.66 DEGREE/W

RJC_Diode=0.85 DEGREE/W
RJA_Diode=6.66 DEGREE/W
TJ=90 DEGREE
TA=50 DEGREE
how do i do the common heatsink for all the three power components?

[paulvdh]

In my opinion heatsink calculations are always very inaccurate because of large unknowns, and therefore trying to do it very accurate is nonsense because the final result may have an inaccuracy of a factor of two or even more. (But I’m also no heatsink specialist.)

Some ballpark figures:
Your total dissipation is 18 + 24 + 6 = 48W
Your worst case part is the IGBT, which has the worst dissipation of 24W
Your IGBT junction will be 24*0.6 = 15 degrees hotter then the heatsink, make that about 20 (25?) because you also have thermal resistance between the case of the IGBT and the heatsink, but you have not given a thermal resistance for that.
This would be a heatsink temperature of 90-20 = 70 degrees.
… and a heatsink temperature difference of 70 - 50 = 20 degrees.
Thermal resistance for the heatsink would then be 20/48 = 0.4 degree/w which is a quite big heatsink, so you’re probably looking at active cooling with a fan.

A more accurate method is to calculate separate heatsinks for each of the parts and then add the heatsinks together at the end.

I also do not trust dissipation numbers of switching transistors. Small changes in switching performance can have relative big changes in the power that has to be dissipated.

A bit of weirdness: What do yo mean with:

@Munyua_wa_Mugo

RJA_IGBT=1.66 DEGREE/W

It does not look like a Junction-to-Ambient figure. Even for a big TO247 that is more like 30 degree/W (I looked at an IRFP240) But as you are using a heatsink, Junction-to-Ambient figures are not relevant.

The thermal resistance for the heat transfer from your parts to the heatsink is missing (and this also has a relatively big tolerance.
The biggest uncertainty is always the air movement around the heatsink. When there is no air movement at all, any heatsink is going to overheat after some time while with a bit of forced air movement (which does not have to be much) thermal performance of a heatsink is easily a factor of 3 better then what it’s datasheet says.