Forrest Christian wrote:

> On 6/17/2011 1:33 AM, Xiaofan Chen wrote:
>> Most likely you do not want to do this in the MCU. What if the MCU
>> has some glitches?
> The general spirit of what you mention is part of my reservations.
> However, I am at least somewhat confident that I can detect and shut
> down an overcurrent within a few ms, as opposed to the second or
> seconds of a PTC.   Plus, I can then report that back upstream.

No matter what protection circuit you use, if you want to recognize and
report a short, you just have to spend an IO pin and connect it to the
(protected) output side of the FET. This can be done in any case,
usually with little effort.

>> There are all kind of protection circuitry but I am not so sure if
>> there are good enough for not.=20

I think it boils down to some form of resistor in the current path that
reaches a threshold voltage when the current is too high. How complex
this has to be depends on the specifics of the situation; sometimes it's
as simple as a resistor and a transistor, sometimes an opamp may be
needed to be able to detect threshold voltages in the milliohm range,
sometimes you can use the voltage across the FET as threshold (that is,
the source-drain path as resistor).=20

You have to determine the minimum conditions where a "short" begins for
you, and the maximum conditions of what kind of "short" you can (or want
to) support. Between the two, the protection circuit needs to kick in :)

In general, like Xiaofan, I tend to look for a "smart" MOSFET in such a
situation; the ones that have protection built in -- I'm not normally
designing high-volume devices -- and only go to raw circuits when the
cost calculation requires it or no suitable smart MOSFET is available.

Gerhard
--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.