So, you want to predict component temperatures do you? Part VII
This series, despite being somewhat lengthy, is by no means a complete overview of the various methods, options and approaches to predicting component temperatures. Here is some stuff I didn’t cover…
My bias in this series is towards the prediction of component temperature via simulation using a 3D CFD approach adopted in our (as ever) market leading FloTHERM and FloTHERM.PCB products. 3D CFD is not bound by many assumptions, it’s quite computationally intensive, providing a simulation result in the order of minutes to hours. Spreadsheet and network modelling methods produce results in the order of instantly / a few seconds. They can do this as they reduce the entire system under consideration to a highly simplified abstracted definition, often in a thermal resistor network format. In such a model a thermal resistance representation of a component is linked via other manually defined resistances to other parts of the system and eventually to an ambient/inlet/outlet. I say ‘they do it’, the point is the user creates the abstracted representation and, if they are expert enough in converting or representing their 3D electronics system to an interconnected 1D type network, can get an acceptably accurate simulation from such a model. Me, I’d prefer to keep things 3D, not have to so fully rely on my ability to abstract, wait minutes (or go home to sleep) and come back to results that I have a greater confidence in.
I did not cover Psi_jt and Psi_jb as thermal metrics. They are somewhat akin to Theta_jc and Theta_jb described already. Unlike the experimental or mathematical derivation of Theta_jc and Theta_jb that forces all of the heat either up or down respectively and calculated as:
Theta_jx = Tj – Tx / Power
thus making those Thetas true thermal resistances.
Psi_jx is defined in the same way (replace Theta_jx with Psi_jx) but measured in an actual operating environment where all the heat is not forced up or down, the P in the above equation is the total operating power and not (necessarily) the power going from junction to top. The theory being that this is more akin to the actual operating environment. It is this rather subtle distinction that makes these Psis metrics not resistances. As with all such modelling approaches, if the component environment used when these metrics were derived is similar to the actual operating environment of the component then the use of these metrics for prediction will yield accurate results. There are many more but this guide from Intersil and this one for TI explain things more fully and much more eloquently!
Finally it’s well worth mentioning a fascinating alternative approach for the derivation of thermal resistor network representations of packages. Micred was acquired by Flomerics a few years back. The flagship Micred technology product is called T3ster (pronounced trister). It is an experimental measurement device plus associated processing software that will power a package up (or down), accurately record the resulting Tj vs. time response as the final wave of the heat leaves the package (or blasts out of the die if powered up) then process that measured information to determine the various resistances AND capacitances that make up the package, that the heat experienced inferred from the various gradients and plateaus on the Tj vs. time curve. This information can then be used to propose a 1D ‘ladder’ network with a series of resistances and associated nodal capacitances that could be input into a spreadsheet, conduction network or a full 3D CFD simulation model. Since being introduced to this capability I have been in a constant state of impressedness (?) of both the beauty of the technology+theory as well as the value such an approach provides. I strongly advise you check out the intro video on this page to learn more!
Thus ends this series!
4th Jan 2010, Ross-on-Wye
- Facebook Live Event on Tuesday July 11th – Frontloading CFD: How and Why?
- SEMI-THERM 33 – ‘A History of Commercial CFD’ Short Course
- Talking CFD Podcast – Democratization, Appification and Strategy
- A Novel Approach to Reducing Heatsink Mass Whilst Preserving Thermal Performance, using FloTHERM
- Response Surface and Sequential Optimisation of a Heatsink Using FloTHERM. Part 6 – Response Surface Models
- Response Surface and Sequential Optimisation of a Heatsink Using FloTHERM. Part 5 – Sequential Optimisation and Compound Cost Functions
- Response Surface and Sequential Optimisation of a Heatsink Using FloTHERM. Part 4 – Response Surface Inspection
- Response Surface and Sequential Optimisation of a Heatsink Using FloTHERM. Part 3 – Cost Function Response Surfaces
- Happy Lunar New Year! – FloEFD Investigates Sky Lantern Aerodynamics
- Response Surface and Sequential Optimisation of a Heatsink Using FloTHERM. Part 2 – Design of (Computational) Experiments