Author Topic: Calculate power in rise and fall of switch mosfet PWM  (Read 536 times)

0 Members and 1 Guest are viewing this topic.

Offline AllTheGearNoIdea

  • Supporter
  • ****
  • Posts: 269
  • Country: gb
    • AllTheGearNoIdea
Calculate power in rise and fall of switch mosfet PWM
« on: December 03, 2020, 02:46:19 pm »
Hi I am hoping someone can give me some advice on how to measure or estimate the power being dissipated  in a MOSFET being used to control speed a spindle motor or my mini CNC machine.   It’s a low side driver, and the motor is around 2 amps at 24 volts off load.   Im driving the mosfet with a pulse width modulated signal at 8khz the rise time and fall time is around 600us.   The bit I’m not sure about is how to go about calculating the dynamic power.   I don’t think I’m that interested in the power dissipation during the on time as RDs On is low for this mosfet at 0.04 ohms.  I have a good selection of test gear etc.

The background story is I have designed and built a spindle driver and for my own interest I would like to calculate the power dissipation within my mosfet. It runs total cold. I just need some help with the math please.  I did make a video on this but wouldn’t recommend it as an hour of your life is a precious thing LOL but it shows the circuit and explain what I am doing

https://youtu.be/3fFVMd66VIY

But thanks for any advice regard Chris



AllTheGearNoIdea Where Its All About The Gear
 

Offline fourtytwo42

  • Frequent Contributor
  • **
  • Posts: 936
  • Country: gb
  • Interested in all things green/ECO
Re: Calculate power in rise and fall of switch mosfet PWM
« Reply #1 on: December 04, 2020, 09:55:23 am »
Hi Chris

Firstly I assume your 600uS may in fact be 600nS as your total period is only 125uS. Switching losses depend a lot on the nature of the load, inductive vs resistive for instance, assuming the former (motor) there is normally little current at turn on except the reverse recovery of any diodes in the circuit. The biggest switching loss is usually turn off where the peak on time current is being interrupted rather slowly. It is possible to calculate a probable loss but if you are able to measure current on an oscilloscope so much the better to see what is actually going on, or make a model in LTspice.
BTW how did you arrive at the 600uS/nS figure, is that a measurement ?

Regards
42
 

Offline trobbins

  • Frequent Contributor
  • **
  • Posts: 348
  • Country: au
Re: Calculate power in rise and fall of switch mosfet PWM
« Reply #2 on: December 04, 2020, 11:01:29 am »
As 42 indicates, if you have the test gear, then I recommend you acquire the FET drain current and voltage waveforms, with sufficient resolution about the switching transitions.  One crude method is to then calculate Pd = Vd x Id for small sections of that waveform, and then sum them up over one cycle, then multiply by the frequency to determine a constant power dissipation.   That may be easier for you than setting up a simulation, and then aligning the simulation waveforms to your measured waveforms, and getting the software to do that integration.  If you post some sections of the waveform then the forum can assist.  Some waveform sections within a cycle will have a relatively constant dissipation, such as when the FET is fully on and fully off, and so are easy to calculate.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf