Thermal Resistance between Chip to Gold Tin Solder

[TeddyLu]

I'm currently doing a heat sink calculation design and having a little confused moment with the area of the chip with a AuSn (Gold Tin Solder) pre-cut that's attached to the bottom of the chip.

I've tried looking around online for AuSn's thermal resistance value, but there is none besides a thermal conductivity. I know that the thermal resistance and thermal conductivity is related as one is based on the area and the other based on the thickness respectively.

My question is would it even be necessary to calculate the thermal resistance and put it into my heat sink calculations as the thickness of the material is so small that it's negligible (about 1mm thick) in the overall aspect of the heat sink design?

 

[Rive]

Depends on the relevance/task. Sometimes it's enough to prove that it is really negligible for the actual application.

... but for example I'm on a task where we are fighting over 100um thickness against 75um (Standard IMS vs. AlOx)

Thickness alone does not say much.

 

[TeddyLu]

I see..My application is just a simple design of a heat sink to keep the chip in recommended operating temperatures. Doesn't have to go into anything advance.

 

[Windadct]

Well I SHOULD know how to do this off the top of my head, and if you really want to work in a field where this is a relevant issue, then maybe you should also?

So - thanks Google:

So a 1 Sq cm die and 1mm thick ( seems to be very thick to me..) and AuSN k = 57 W/ m*C ( I prefer to reference Kelvin in these - but that is for another discussion) ...

= 0.001 / (0.01)^2 * 57 = 1/57

A 1 Sq CM Die will have a Rth of 1/57 = 0.0175 C/W... so... 100W -- will have deltaT of 1.75 C rise across the junction... BUT you say "attached"- is the solder actually soldered to BOTH the die and the heatsink? If so then this should be a good number, if not, you have to consider the junctions chip->solder Solder->HS

 

[TeddyLu]

Hi Windadct,

Thank you for your insight, the chip is actually soldered to a small copper moly copper plate. I was having a hard time trying to distinguish if the stated thermal resistance (θjc) of the chip from the data sheet actually meant the thermal resistance from the top of the chip to the bottom where it's in contact with another surface or in contact with a heat sink specifically.

Could you clarify upon that part?

 

[Windadct]

j = Junction, the source of the heat, typically a theoretical average- in the middle of the die... the "top" is not the best way to think about it ( if you put a sensor there is will be cooler than the Max Tj )

c = case

The typical model looks like a resistor voltage divider, with the Junction being the starting point and Ambient ( or coolant) being some form of an unlimited sink.

Tj @ "x" Watts -- the source
| Rjc k/w
Tc
| Rcs
Tsink ( surface)
| Tsa
T ambient ( coolant temp)

The Wikipedia also shows this..

Also - a description at PowerGuru.org ( larger devices )