Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations cowski on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Junction vs. Case Temperatures

Status
Not open for further replies.

EngAP

Mechanical
Jun 6, 2007
52

Quick question from a mechanical engineer to electrical engineers:

How much hotter is the junction temperature compared to the case temperature for a chip?

Is there a range it is likely to fall within or is this a much too vague question?

Specifically, if the spec for maximum junction temperature is stated as 100degreesC what would you recommend the chip case temperature be limited to? Just looking for a ball-park figure for initial calcs.
 
Replies continue below

Recommended for you

Depends on the die-to-case thermal resistance, which varies ALL OVER THE PLACE. You need to look at the datasheets.

TTFN

FAQ731-376
 
and of course the power dissipated. The temperature difference between junction and case is the thermal resistance junction-to-case (degC/W) multiplied by the power in the chip (Watts).
 
Thanks - that was what i thought. Problem is that the manufacturer will no longer specify a thermal resistance (Rj-c) due some testing issue so I have no idea how hot the junction will be running, and the case is running close to the maximum junction temperature (within 5-10degC of max. junction spec.)

Been checking all over the net for datasheets and all can I find is the resistance from the die to the legs. The chip has a heat spreader plate integrated on its underside. Ideally I would want the resistances die to spreader plate, die to board via legs and die to case exposed to air to determine the % power exiting through the case and spreader plate where the temp is measured.
 


However, I know that the topside (air exposed) of the case is 10degC hotter than the spreader plate. if I know the power I could estimate the jucntion temperature using this temeperature relationship and assuming the resistance from junction to case is a multiple of the resistance from junction to spreader plate.

Would it be reasonable to assume that the resistance from junction to case is 10x that from junction to spreader plate or more like 100x?
 
If there is really no support available from your supply you should consider measuring junction temperature during operation

Possible for all devices containing at least on pn-junction accessible via the pins.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor