> On Sep 26, 2004, at 1:11 AM, Russell McMahon wrote:
>> - When a tantalum capacitor is connected to a "high energy" source
>> (power supply rail) and is subject to an over-voltage spike of
>> relatively low magnitude, the capacitor can (and often does) fail
>> short circuit and allow destructive dissipation in the capacitor.

> Perhaps the problem is no so much tantalum itself, but the trend
> to put as much capacitance as possible in a package by fine-tuning
> the voltage rating (dielectric thickness, presumably.)  Those 4V
> 470uF tantalums are very impressive, I'm sure, but perhaps it was
> a bad idea to expect them to work on your 3.3V power supply rail.
>
> Likewise, I wonder if the older tantalums (16V and higher?) had
> the same failure mode, or whether they were just inherently made
> with more leeway in their voltage spec...

It's a fundamental tantalum capacitor problem, at least as implenmented. It 
has applied for decades.
Wet Al has a fuzzy failure mode. Tantalum is sharp and brutal. I have seen 
and heard and smelt Al cap failures, but they are quite different from 
tantalum ones, and AFAIK the failure mode is almost always open circuit.

> If your (reasnoably overrated) tantalums are exploding on your
> power supply rail because of over-voltage spikes, perhaps you
> have more serious problems than the failure characteristics of
> the caps themselves...

Not ever spiking them will help heaps - but it's a fundamental problem.


        RM

_______________________________________________
http://www.piclist.com
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist