contrary to popular belief the temperature limits on the spec sheet don't have a lot to do with the chip limit. in fact for accelerated life testing the manufacturers frequently run them at higher temperatures.

we've used many CMOS processors, starting with the RCA1802 at temperature to 225 Deg. C. for the last 20+ years. what happens is the leakage goes up due to the increase in conductivity of he Si so the device disipation increses. however, for a static device you can slow the clock cycle down and limit the disipation somewhat. there are several other considerations. for example the older MIL spec devices use a ceramic pack that was good for a few thermal cycles but died because the ceramic cracked and they lost hermaticity (is that a word?). the lower temperature plastic pack devices didn't suffer from this problem but had a higher mortality and shorter average life.

the manufacturers don't want people using the devices beyond spec because they don't want to deal with the increased failure rate or shorter life. if you, like me, don't care as long as the device works a couple of hundred hours and then you throw it out. then i'd adivse using a standard commerical CMOS device without a thermal cutout. if these devices survive three to five thermal cycles of about two hours they will generally last for several hundred hours at temperature.

remember though if your are involved in a real time application and you use a slower clock the timers won't work the same.

hdc3