Help me calculate LED Junction temp

[marcopolo]

Hi I have all the figures but I don't know quite how to put it together to calculate a realistic junction temperature.

My stabilised aluminium heatsink (old CPU heatsink) temp: 45deg C (with a big fan pointed roughly at it)
LED board 7up cutter XP-G R2 at 1000mA (23W)
stuck directly to heatsink with Artic silver 5: 0.0045°C-in2/Watt (0.001 inch layer)

XP-G thermal resistance: 6degC/W

Help on calcs please!!

Marco.

 

[uplite]

Calcs? GIGO. The heatsink temperature is not so useful.

Mount a thermistor on the PCB, right up against one of the central XPG's, and measure the temperature there. Much closer to junction.

-Jeff

 

[jeffosborne]

Hey marcopolo, you said LED board 7-up, is that an aluminum board, or FR4 type PC board stuff with lots of through-hole vias? Also, a heatsink sized fan, bolted directly to the unit, will be more efficient at heat removal than one generally pointed at it. What's you application for this setup? Cheers, Jeff O.

 

[HarryN]

Hi I have all the figures but I don't know quite how to put it together to calculate a realistic junction temperature.

My stabilised aluminium heatsink (old CPU heatsink) temp: 45deg C (with a big fan pointed roughly at it)
LED board 7up cutter XP-G R2 at 1000mA (23W)
stuck directly to heatsink with Artic silver 5: 0.0045°C-in2/Watt (0.001 inch layer)

XP-G thermal resistance: 6degC/W

Help on calcs please!!

Marco.

For a reasonable estimate, you can guess on somethings and frame it within a range.

First, within the LED itself = 3.5 Vf x 1 amp = call it 4 watts

4 watts x 6 C / watt = 24 C within each LED above its connection.

From this, the absolute lowest it is is 24 + 45 = 70 C

Adding in a few more factors
- Another 10 for the MCPCB + interfaces
- Another 10 for measuring in non ideal place

Rough estimate = 95C

You only use the wattage per LED for the temp rise within one led, not the total power, unless - it is all in one package.