2
\$\begingroup\$

I decided to use IRFP460 in a two-switch forward converter. This Mosfet can dissipate 280 watts. Can I get 1000 Watt from my converter?

As a general rule how much power a Mosfet can transfer with respect to its maximum dissipation loss?


I concluded that the mosfet can dissipate a derated loss power at a specified temparature if thermal resistance from junction to ambient be a required value; eg 2 c/w for 20 watts at 100 deg. Celcuis of case temperature. Is it possible?

\$\endgroup\$
2
  • \$\begingroup\$ Duplicate, and that one is a much better formulated question too, imo. \$\endgroup\$ Nov 14 '15 at 6:28
  • 1
    \$\begingroup\$ I wouldn't look at it as a ratio. The dissipation of the mosfet depends on how fast it is switched, and how often it is switched, and what the voltage is at time of switching, and what the current is while on. Also, and this is very important, the power dissipation is specified when the case temperature is 25C! It is not possible to keep the case at 25C while it is dissipating 280 Watts unless you use some type of fluid cooling system. \$\endgroup\$
    – mkeith
    Nov 14 '15 at 8:00
1
\$\begingroup\$

The 280 watt figure is really optimistic .you would get this on an infinite heatsink that was held by some means at 25celcius.Now the fet would be running at Tj max of 150 celcius .This is all totally unreasonable .For good reliability 100 celcius is more acceptable and its reasonable that your heatsink could be at say 60 celcius .If the heatsink were to be at 25 celcius it would have to be infinitely large.So we only have a differential of 40 between the junction and the heatsink now.If we use some insulater like a mica washer or silpad etc then this adds to the total thermal resistance between the junction and the heatsink.Well 1 degree celcius per watt seems a reasonable total figure.Therefore 40 watts of dissipation is acceptable and your SMPS if its worth its salt will waste less than this per device .In fact a good SMPS that has 2 switches should waste 10watts per switch.

\$\endgroup\$
2
  • \$\begingroup\$ Good answer. I see the linear de-rating factor of 2.2 W/(deg C) there. It means I should to decrease the dissipation at higher case temperature than 25 degree centigrade. So for 50 Watt dissipation, the maximum temperature of the case is 102 deg. centigrade. true? \$\endgroup\$
    – Rasool
    Nov 14 '15 at 10:41
  • \$\begingroup\$ As a comparison, a typical cpu may dissipate 100 Watt of power and its case temperature be 70 deg. Celsius. \$\endgroup\$
    – Rasool
    Nov 15 '15 at 6:53

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.