operating temperature range

Martin Hodge

I've been asked recently to list the operating temperature range of the Propeller ASC+ and I don't know how this is typically done. Does a sample of the board need to go though some certified testing procedure, or is it sufficient to use the rating for the least robust silicon component on the board?

LoopyByteloose

Well, silicon generally melts at around 150 degrees C. But you could just use the rated operating temperatures of the installed Propeller, the EEPROM, the FDTI chip, the MCP3208, or the voltage regulators --- whichever combination produces the narrowest temperature range would be the right answer.

Yes, the least robust componet at high temperature AND the least robust at low temperature would be needed.

You might want to clarify that the Propeller itself is tolerant of -55 degrees centigrade to 125 degrees centigrade, and offer wider temperature range support by special order.

JLocke

Did you mean to say the solder on the silicon melts at around 150 degrees C? I used to be a Field Engineer for a semiconductor furnace company, and I can tell you that at 1200 degrees C, you were more worried about some of the quartzware starting to sag. Never saw the silicon melt; but I've seen on television where they are pulling the ingots of silicon, but I don't recall the temperatures.

Mike Green

You do need to test one or more sample boards in addition to using the smallest operating temperature range of the components on the board (including passives).

LoopyByteloose

It is interesting to find that there is a market for sub-zero operation that is likely much higher demand than +100 centigrade. And there are a lot of silicon devices only rated down to 0 degrees C. Not sure if this a North American/North European/North Asian mix of cultures that readily adopt leading edge technology or what.

You could start with your freezer, and then try a cooler box with dry ice (freezes at -109.3 degress F or about -78.5 degrees C.). Liquid CO2 might even be ideal as it is supposed to be at -55 degrees C at 30bar. Not sure how to deal with the liquid being open to the atmosphere though... maybe blasting with a fire extinguisher.

You can get a bottle of CO2 at any Coca Cola, Pepsi, or beer distributor. I use to use it to drive nail guns and a variety of air tools without the racket or electrical hookup that a compressor requires.

I got the 150 degree limit for silicon off of power BJT thermal limits. I guess it is the junction that fails at 150 degrees.

http://scifun.chem.wisc.edu/chemweek/pdf/CarbonDioxide.pdf

http://www.interfacebus.com/Transistor_Derating_Guide.html

RDL2004

Liquid CO₂ is pretty hard to do at home because of the over 75 Psi requirement. I think if you just popped open a bottle, it would just vaporize instantly. I remember experimenting with it as a non-VOC paint solvent back in the 90s.

Genetix

You could specify very conservative limits such as 0 to 40 or 50 C, otherwise you would need to do testing which is best done with an environmental chamber.
If you need to this testing then use at least 3 samples though 5 or 10 would probably be better. For best results use product that was built at different times.

LoopyByteloose

You would need a pressure vessel of some sort. Dare I suggest you modify a pressure cooker to test your circuit board and then transfer liquid CO2 into it? It could be done, but it is a kind of Gyro Gearloose approach.

Martin Hodge

Specifying consertive limits based on the temperature ratings of all the parts is what I've done. 0 - 75c is what I've come up with.

Thanks for the input. I may pursue professional(ish) testing to expand that rating at a later date.