That value of 80 amps is a calculated value. That means that it will not happen in real life, only in a calculated situation. The term for that technique is "Specsmanship", which involves creating part specifications that only apply under conditions that can never be achieved in the real world.

I agree the 75 A is inflated and think the Infineon link in post #2 makes that clear. My point was that your "millisecond" survival at 40 A claim is also unreasonable.

 

Here is the problem. I did not say anything like what you are objecting to. I said case to ambient was 70C/W.
You asked me to give a link, which I did. Do you know what “case to ambient” means, apparently you do not, since you go on about massive heat sinks and forced air cooling.

I did not say that it could mot work with a proper heatsink, which you also seem to think I said.

Bob

When performing a thermal design, the purpose of a heat sink is to reduce the "case to ambient" thermal resistance. When you refer to a TO-220 package directly to air, it is only one instance of case-to-ambient. In the case of a power semiconductor, it is an irrelevant case. There will always be a heat sink of some type. I am very aware of what case-to-ambient means. There can be some confusion because "non-power" components will have a case-to-ambient specification on the data sheet which means the case directly to air. However, this specification for a power semiconductor is irrelevant. This thread is about a power semiconductor. So, heat sinks are not only relevant, they are essential. My posts are relevant because they are part of understanding how a TO-220 package can support a MOSFET controlling 80 amps. Isn't that the whole point? As a reminder, the title is "How the heck can a TO-220 case MOSFET carry 80A?".

To be fair, I read the Wikipedia article. It is all about heat sinking and throws in the 70 C/W figure as a special case.

 

Hi,

Junction to ambient is just that, the junction in the package to the ambient with no heatsink. It is typically high like 60C/watt.
That is for when you want to think about how much the package can do without a heatsink.

Case to ambient:
For the bare case, a rule of thumb is 60C per square inch of surface area per watt, given relatively small surface areas like 1 square inch. So if you have junction to case you can then add case to ambient if there is no heatsink.

Case to heat sink is the effect of the thermal grease and the distance between the metal bottom of the package and the bottom of the heatsink. It is a tiny distance but still something that could be important.

Junction to case is due to the difference in temperature between the die and the metal of the package.

Heatsink to ambient depends on the heatsink and any forced cooling.

Of course these are all taken as thermal resistances in series and we have two cases:
Junction to case, case to ambient (with no heatsink).
Junction to case, case to heat sink, heatsink to ambient (with a heatsink).

If someone wants to get or make a measurement of the lead size (width and thickness) i can try to calculate the
approximate maximum temperature rise of the lead.

Oh one last thing...
Many of the ratings are based on a 25C case temperature. Once the case temperature rises (and it will) the power handling capability of the package goes down. That in turn means lower allowable current.
What that in turn means is that the case temperature would be part of the max current calculation for a given design knowing all the thermal resistances.
So in other words if we were to use P=i^2*Rds we would really have to incorporate the temperature rise and derating factor(s) and thermal resistances in a formula type expression or do a couple iterations.

 

Some additional help -

https://www.infineon.com/dgdl/smdpack.pdf?fileId=db3a304330f6860601311905ea1d4599

http://www.ti.com/lit/an/spra953c/spra953c.pdf

https://www.analog.com/media/en/technical-documentation/application-notes/AN-1604.pdf

Calculators -

https://www.heatsinkcalculator.com/free-resources/heat-sink-thermal-resistance-calculator.html

https://thermtest.com/thermal-resources/thermal-resistance-in-series-calculator#calculator

Simulator -

https://www.allaboutcircuits.com/news/an-introduction-to-the-texas-instruments-webench/



Regards, Dana.

 

The one time that we had a test device designed by a consultant the heat sink selected was just barely adequate and the result was that the pass transistor was running just a few degrees below the max allowable temperature. Our customer for that device was a very demanding one that demanded 100% reliability. My fix was to change to a much larger heat sink, which lowered the temperature a lot. It was simply not worth running the transistor that hot, with the related shortened time until failure, to risk not changing the heat sink.
An adequate safety margin is usually the cheapest insurance against failure.

 

"Case to ambient" is effectively case to the heat sink. It probably does not include the junction to case resistance.

No, case to ambient is case to the air without any heatsink.

Bob

 

The one time that we had a test device designed by a consultant the heat sink selected was just barely adequate and the result was that the pass transistor was running just a few degrees below the max allowable temperature. Our customer for that device was a very demanding one that demanded 100% reliability. My fix was to change to a much larger heat sink, which lowered the temperature a lot. It was simply not worth running the transistor that hot, with the related shortened time until failure, to risk not changing the heat sink.
An adequate safety margin is usually the cheapest insurance against failure.

Good comment. For readers not familiar with effects of temperature, here is a link to typical info on failure rate vs temperature.
http://www.interfacebus.com/semiconductor-wirebond-failure-rate-with-temperature.html

If I was the consultant I would have designed in a massive heat sink and charged more money to change it after the customer complained. I make more money AND I'm off the hook for any failure. (Flag: Sarcasm)

 

Good comment. For readers not familiar with effects of temperature, here is a link to typical info on failure rate vs temperature.
http://www.interfacebus.com/semiconductor-wirebond-failure-rate-with-temperature.html

If I was the consultant I would have designed in a massive heat sink and charged more money to change it after the customer complained. I make more money AND I'm off the hook for any failure. (Flag: Sarcasm)

The tester de3vice was built by our technicians after the consultant had departed. It was one of those situations where he had been hired for another project and still had 2 days left on his contract when that one was finished. So he was given this tester because I was busy. Nothing was wrong with his design, except that the safety margin was too small by my standards. Of course, if anything failed I would be the one to create the fix. Fortunately the bosses accepted my thinking of "Do it right the first time so that we do not have to do it again." But in 5 years with no failures they realized that it made a lot of sense.

 

So why did this thread pop up??? Especially given that my comment i the lat one appearing, and given that I had not made any additional comments. When email notifications appear in my mailbox I do not check to ee how old the thread is.

 

So why did this thread pop up??? Especially given that my comment i the lat one appearing, and given that I had not made any additional comments. When email notifications appear in my mailbox I do not check to ee how old the thread is.

You mean pop up in your email? Maybe you should post something about this in the site suggestions area.
Yeah i see your previous post was the previous last post in the thread on Jan 26 of 2020, well over a year ago (ha ha). Maybe the email system was working VERY slow last year and this year