# Heat Sink Calcs

Posted by Anonymous User
 Anonymous User Heat Sink Calcs July 03, 2007 02:17AM
The other night I started wondering about how well the current L298 + heat sink combo would do with non-standard motors. I just did a bit of reading on how to do heat sink calculations and figured I'd post the results. If anyone sees any errors here, please let me know. I'm new to this stuff.

The 298's datasheet says the maximum junction operating temperature is 130C, so we need to ensure this is never exceeded. Heat sinks have a specified thermal resistance given in degrees C per watt; lower is better. The currently specified heat sink is the Aavid 566010B03400G, which has a thermal resistance of 11.5. In addition, you have to account for the thermal resistance between the junction and the outside of the L298's case where it connects to the heat sink. This value is given in the L298's datasheet as 3 degC/W. The total resistance is 14.5 degrees C per watt, so if the junction is dissipating 1 watt, the junction will be 14.5 degrees warmer than room temperature.

If we assume we're in a warm room at 30 degrees C, we can allow the the junction to get 100 degrees warmer before it fails, which means it can be dissipating about 6.9 watts. It's not so easy to work backwards to figure out how many amps this equates to. If you look at Figure 1 in the 298 docs, it looks like 1.9A is a pretty good guess. At that current, the chip is dropping about 1.9V on the high side and 1.7V on the low side, for a total 3.6V drop which would be 6.8 watts.

All this assumes that you are running in full-stepping mode where both windings are never energized at the same time (I think this is always true right now). If you changed the firmware around to do half-stepping or high-torque where both windings could be energized at once, it looks like you'd be limited to about 1.2A per winding. These are absolute maximums, I'm sure you'd want to derate these numbers significantly.

I found this interesting: If I'm reading this right, the default RepRap motor takes around 0.85A if you have the torque at 100%. At that current, the L298 drops about 2.45V for 2.1W. Even with no heat sink the junction temperature will sit at around 73 degrees C above ambient, so around 100C, still safely below its max.

This was mostly an academic exercise because obviously the chip/heat sink are working fine for everyone right now... but I feel a lot better knowing that there's a healthy safety margin built in.
 Re: Heat Sink Calcs July 03, 2007 08:40AM Admin Registered: 17 years ago Posts: 1,915
"The 298's datasheet says the maximum junction operating temperature is 130C"

I can assure you that it runs very happily at over 200C. I've got the seared fingerprint to prove it.
 Re: Heat Sink Calcs July 06, 2007 06:30PM Registered: 16 years ago Posts: 23
Nice calculations emf

Edited 1 time(s). Last edit at 07/06/2007 06:36PM by savecore.
Sorry, only registered users may post in this forum.