LM224

[tresguey]

I have a question about temperature ranges on data sheets. I asked a question to an engineer friend of mine about a PWM circuit I am working on. I chose the LM224 by TI without noticing the operating range was on that chip. It will be in about a 200F environment and it is rated at 185F. My engineer friend says it ok because it will not be running nowhere near max capacity. That did not make sense to me. But then again I'm very green to electronics.

[Audioguru]

If you want every circuit you make to work properly and last for a long time then use an LM124 that is rated for a max temperature of 257 degrees F. Your customers will like you.
If you want to gamble and replace some of your products under warranty then use an LM224. Your customers will hate you.

[uncle_bob]

Hi

What will likely happen:

Things like leakage and offset voltage may be out of spec. Output current may not quite be as good as you would like.

All of that is a maybe. You are talking about VERY old chip designs. I would not be surprised at all to find identical performance between the various temperature range chips.

That leaves the package it's self. Back when these parts came out, plastic was not a wide temperature range material. High temperature parts came in metal or ceramic packages. All of that has changed a lot in 40 years.

So in this one case, these specific parts, doing what you are doing ... likely not a problem. It's mighty easy to stretch this logic WAY further than you should. High speed digital circuits do not do well at 125C when they are the 70C rated parts. Package is the same ... parts selection is different ....

Bob

[Seekonk]

A long long time ago  i worked in a repair shop while in college. Fixed a lot of car radios that used TO-220 transistors.  I ordered the lowest voltage of the series type.  With a voltage breakdown tester I sorted them by voltage.  Actually there were only two types, a few low voltage ones and well over 90% were more than the maximum rated voltage for the series. The tested higher voltage ones were used for higher powered amplifiers.   Every once in a while there was one with a squirrelly curve that went right into the bin.  Point is manufacturers don't make a good, better, best type of a die.  These qualities are sorted out when needed.  To fill orders of the common stuff everything is shipped out.  In less than a second they can sort out parameters that statistically indicate performance under certain operating conditions.  As said before, actual cases do matter.  If your circuit out performs typical published specs it will likely work well at extended temperatures. Chances are with you even if you don't test.  They don't try to make bad parts.