Fixing the Roasted Test Load 

more likely dipped the test load in liquid nitrogen

The devices are actually measured in extreme conditions by the manufacturer. This is done using a water cooled plate and an excellent thermal contact. Case temperature is monitored and the actual junction temperature is measured through junction forward voltage. This is how the thermal resistance is specified.

@@prt1527 Likely never was tested like that. It's a cheap Chinese product. Most likely it's just a reference to something from the original design mosfet's datasheet.

You could use a refridgeration loop to keep the case well below freezing. But that would only be for a whole bank of transistors, nobody is gonna freeze a single transistor unless it's some weird application where it needs to be frozen anyway (ex, it's running in the circuits inside of an industrial freezer).

​​@@bobert4522 I was talking about the ST replacement part and how power semiconductor datasheet values are determined. Not about the application.

can you provide link to that model you replaced with?

no its the max specs at 25c but also affects inrush, peaks, and non repetitive surge currants we need these specs for sizing and headroom we need all the specs to make a proper choice

@@davidknightaudio934 it maybe the spec at 25°C but you try and achieve those currents and maintain the FET at 25°C. it's done to make them stand out as i said. every engineer knows that you have to work with the smallest heatsink that works for your specified performance requirements. but you take the max current figures with a pinch of salt. it might prove me wrong if we look at the RDSon, max current, max junction temperature, and thermal resistance from junction to case at 25°C . would be great to see how hot the junction will get if we pull the max rated current.

@@TheEmbeddedHobbyist ur not getting what i am saying thats ok i am trying to explain that those figures are very helpful for things like headroom if theres a surge and telling you how it will handle power on inrush and so many other factors we need all the specs the entire data sheet and the rest of the specs are more important than the absolute maximun ratings ofcource

Particularly the pulsed drain spec and the tj @ 25c power rating

@@davidknightaudio934 All I'm saying is that they use the max ratings in large bold font on the front page, where the non engineers get excited over the part. Where as we understand that these max values all fall over when you start to look and understand the derating curves etc. So in affect they are just eye candy. I once used a 1ohm 1kw resistor which was mounted on a very large with very high powered fan cooled heatsink. As it was needed to disipate over 3kw's with all the loads mounted on it. the resistor blew up at about a 1/4 of its rating, reading the small print it needed a polished mounting surface not just machined to achieve the low thermal resistance it needed!

At least some of them are recovered from other scrap devices.

​​@@Mark1024MAKt's not good idea, some of the easyly broke when desoldering

He just bought or desoldered them from some old electronics during the old times and kept them to this day.

@@abeliever7301 - it depends on how good you are at desoldering 🤣

@@Mark1024MAK no it's depends on which device to use for desoldering.(Hot air gun/normal soldering iron) . It's not cost effective when it's come to Resistors, capacitor,transistors

Why remove the fan? It would make the test load bigger. A fan is such a small load and it only turns on when the heatsink gets hot.

Not that Tesla ;)

Mosfet in linear region do not work well with the heat.

I agree, it's kind of interesting. I would have guessed a mosfet first too. I made an electronic test load and used a mosfet as well and I'm trying to think if I even considered using a bipolar part.