> > so many inaccuracies... if you are graduated from a high engineer school we can
> take
> > this to pm if you want because seeing the reactions to my last post, I don't think
> > it's wise to post here even further details of the mechanics
>
> Yes, I graduated as an electrical engineer with first class honours, no I am not
> interested in arguing over private messages. I've seen plenty of semiconductor device
> failures in real life. The most spectacular was when an IGBT in a 15kW 3-phase
> inverter literally exploded in front of my face (collector-emitter punch-through),
> but most of them have been far more mundane. I know what you're saying - there
> shouldn't be enough energy supplied to break covalent bonds so the chemical
> structures shouldn't change. It's nice in theory, but the compounds used in
> semiconductors aren't perfectly stable, and they degrade naturally at room
> temperature. They degrade faster at higher temperatures, and the smaller the process
> the less degradation it takes to cause failure.
As you wish good Sir, just one final word, there are many parameters that needs to be considered such as thermal isolation of the chemicals, dissipation and more generally thermodynamics and that's pretty fugly, so don't want to post about this here. Even the energy levels on a sub atomic level should be considered for completeness, and to be entirely honest I no longer have the level to analyze this. But micro devices engineers are no fools and they know all the mechanics perfectly and all the risks are measured.
Anyway I think we basically say the same thing from the beginning but you're focusing on the epsilon probability of reactions happening at low energy levels due to some very rare spacial configurations of the electrons. When you deal with large quantities, it's always easy to point at the epsilon, but it doesn't make the statistical pattern any less true.
|
Index 
Flat 
Re: An interesting prototype has turned up
Reply
Re: An interesting prototype has turned up *edit*
