I'm looking into a way to calculate the power dissipation of any arbitrary IC in circuit on SPICE. I eventually hope to create simple scripts that can list power dissipation of the device and potentially raise a flag if the device is over our de-rated maximums. I've been looking online and can't find any concrete methods to do so.

How do I model the power dissipation of an IC in SPICE?

  • \$\begingroup\$ Can't you just multiply the supply voltage by the supply current? What exactly do you mean by "raise a flag"? \$\endgroup\$ – Elliot Alderson Dec 19 '19 at 18:38
  • \$\begingroup\$ I meant as in alerting us that a part is going over rated value, that was just context nothing to do with the actual problem sorry. My issue with that is that the supply is going to be shared with other components, and the power generated by it isn't going to be consumed only by the IC in question \$\endgroup\$ – Jaywalk Dec 19 '19 at 18:41
  • \$\begingroup\$ You can always add a 0V source in series with the component of interest and use the current through that source. I feel like there are details you haven't told us. What kinds of models are you using for the ICs? Which SPICE? \$\endgroup\$ – Elliot Alderson Dec 19 '19 at 18:56
  • \$\begingroup\$ Spice3, for IC's I'm currently looking at op amps, but would like to model other parts as well eventually; any arbitrary IC. I'm sorry it sounds a little vague, I'm sort of in the research phase on this. Would looking at the currents going into the IC's multiplied by the supply voltage (minus any loading) be a good measurement of the power dissipation? I ask because adding 0V sources to all the circuits would make it look ridiculous. \$\endgroup\$ – Jaywalk Dec 19 '19 at 19:12
  • \$\begingroup\$ OK, I didn't know you were worried about "looks". \$\endgroup\$ – Elliot Alderson Dec 19 '19 at 19:20

The power can be estimated by summing all of the power flowing in and out of a part (because energy is conserved). This doesn't work well for some models that have voltage or current sources built into the model (that can magically dissipate and create energy).

Typically power can be calculated for an IC by looking at the current and voltage going in and out of all the pins (this is what LT spice does if you alt-click on a part, but it will add all the signals from an IC to calculate power dissipation automatically).

Keep in mind that even after the power has been calculated, there are further steps to finding the actual temperature of the part, like the thermal resistances and average dissipated power

  • \$\begingroup\$ I believe LTspice now has some sort of thermal modelling components in it, though I have no idea how to use them. \$\endgroup\$ – Hearth Dec 19 '19 at 19:24
  • \$\begingroup\$ Just used it yesterday, it's still the same old spice. One way I have found average power is using a b-source with a laplace low pass filter. \$\endgroup\$ – Voltage Spike Dec 19 '19 at 19:32
  • \$\begingroup\$ Huh, I could swear I saw some kind of heatsink modelling system last time I was using it. Perhaps I'm on a different version, or maybe it's some third-party package my university provided. \$\endgroup\$ – Hearth Dec 19 '19 at 19:34
  • \$\begingroup\$ If you do, its something that you have to add in to the model. There is a temperature sweep, but this is a global temperature sweep \$\endgroup\$ – Voltage Spike Dec 19 '19 at 19:36
  • \$\begingroup\$ @VoltageSpike Just a side note: try to avoid Laplace in .TRAN, they can be dreadful. Whenever you can, replace them with a proper RLC network for infinitely faster and more accurate results. \$\endgroup\$ – a concerned citizen Dec 20 '19 at 17:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.