Inverter Overheat

The temperature of the inverter housing doesn't matter much because it is not in good thermal contact with the power transistors, which is what actually matters. The case can be bearably hot and yet the guts could be dying.

 

Hear, hear! I did the same Vertex, and then, The dealer put the pink stuff in on the inverter recall. Oh well, it was free.

 

Short version: Folks, don't worry about your inverters being damaged by high temperatures. It's just not going to happen.

Long version: Stress testing of semiconductors often takes place at 200C and double the rated voltages for 1000s of hours, and failures are quite rare. In the Prius, you don't really have to worry about a wear-out mechanism occurring in the drive transistors (and these are the ones that would get hot). Additionally, the thermal sensor in the inverter shuts things down LONG before there is significant wear-out damage to the drive transistors.

The catastrophic failure mode would be for the bond wires to blow. We don't have to worry about these for two reasons:
1) Overheating transistors of the type used in the Prius (field effect transistors) don't increase the current through the bond wires, they decrease the current.
2) [and this is the most important one:] If you DID lose transistor bond wires, the inverter would totally die - you wouldn't be questioning whether it was bad, you'd KNOW it's bad.

It's pretty hard to smoke an inverter, and they're pretty much immune to latent damage due to overheating.

 

Toyota used IGBT power transistors in their inverters. Do they behave in the same way as FETs when it comes to heat? I ask, because the first generation 2006 Highlander Hybrid inverters are prone to failure, due to a burned power transistor. And as you say, once that happens, shutdown is immediate.

 

It would also be interesting to know whether Toyota's IGBTs are stress tested under 200C, double voltage, for thousands of hours.