Bonding Wire Dissipation

[powerelec]

The IC I am designing has a power mos inside it. Half the power loss is in the MOSFET and half in the bonding wires to connect the MOSFET. When calculating the junction temperature of the chip based on the package Theta-JA and the ambient temperature, should I ignore the loss on the bonding wires and just consider the loss in the MOSFET itself? Or should I add the bonding wire loss too since the bonding wire is inside the package?

Thanks in advance!

 

[erikl]

You should add the bonding wires' loss too - both losses contribute to temperature rise of the package - don't they?
But you can use Theta-PA for the wires' loss of course.

 

[powerelec]

- - - Updated - - -

Thanks Erkil, I assumed that since bonding wires are good thermal conductors - the heat in the bonding wire would be easily conducted out of the chip through the wire and the pins. So that heating effect can be left out of the equations. Also what does
Theta-PA stand for? It does not seem to be a commonly specified parameter for most packages.

 

[erikl]

... what does Theta-PA stand for? It does not seem to be a commonly specified parameter for most packages.

Sure it is, at least for power devices. With "Theta-PA" Package-to-Ambient is meant, but more often it's called case-to-ambient, nominated as R_θCA, which is the difference of R_θJA and R_θJC , s. e.g. this data sheet: View attachment FDT434P-FET.pdf , p. 1 & Notes on p. 2 .