Thanks for your comments NickS.
Really, this problem is really directed at fault analysis type people. People that have experience with electronic device operation outside of specification, not the reverse.
I really do appreciate everyones efforts.
Won't I don't appreciate is negative posts NickS. If you can't meet the contraints, go away. Keep your opinions to yourself if you can't help solve the problem. I don't think anyone appreciates posts like yours. You sound more like a politician than an engineer!
Thanks for your input Steve, as it did make me think of something else and I have a solution. Yes, it was hard... and I did lots of trials (with errors) but we got there in the end. Its amazing how the manufacturers of electronic devices under-rate components. Yes, the component will be degraded if over spec operation is induced, but most components don't die easily! They are all a little like fuses, where you have to run them at 400% to make them blow in a short time.
With that note, I would like to close this thread. Ciao
GreenTea, The problem is officially overly constrained.
-No more than 2.5V
-No more than 50mA
-No smaller than 1206 which is massive for surface mount
-No more expensive than a few cents
-Can't be a trace.
I'm sorry but there are only a few of component types that you can even buy for a few cents. A designer often is forced to compromise on one thing for another. Its not a lack of thought or trying, physics is just littered with balances and to be exceptional in one regard almost alway presents a new problem you are forced to account for.
Now you seem fixated on the idea that we are not trying hard enough, but I say something has to give. There are plenty of ways to guarantee what you want but not for the price, size and power you have declared. I know for a fact that carbon film resistors are used as electronic igniters because they burst into a violent flash of energy when massively overdriven. But that requires serious current like from a car battery.
Steve has made a great point on making parts marginal (which I think is the best you can hope for) but I am not sure you could even ruin a 1206 size diode with such low power. The only chance I think you have is to put down the smallest diameter wire you can find and hope it blows out with 50mA.
