On 6/17/2011 1:33 AM, Xiaofan Chen wrote:
> What is the P-MOS you are using?
IRF9540NS in D2PAK.   It's oversized mainly to survive said short.  I'm=20
driving it with level shifting circuit using jellybean transistors and=20
resistors.
>> 1) PTC's don't seem to be the right tool for this job.   Yes, they seem
>> to work, but only 'mostly ok'.  I can go into more details if necessary.
> I trust you. But why not give more details so that the other people
> will learn from you as well.
It seems to mostly be a speed thing.  With a PTC, I find the following:

1) If I'm using a switching supply the supply will usually shut down=20
before the per-channel overload protection.  This is both good and=20
bad.   Although on balance, probably bad, since it also effectively=20
shuts down everything on that supply.

2) if I'm using a high-amperage supply with fuses protecting the entire=20
device - often the fuse will blow before the PTC will trip.   Obviously=20
I'd rather have something not need to be replaced in a short.

3) Assuming I use a higher-amperage fuse so it doesn't blow before the=20
PTC trips, then you run into issues where the item causing the short is=20
damaged.   Considering the most common device causing the trip is a wire=20
cutter or RJ45 crimper - the resulting bright flames shooting out of the=20
crimper isn't nice.
> 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. =20
However, I am at least somewhat confident that I can detect and shut=20
down an overcurrent within a few ms, as opposed to the second or seconds=20
of a PTC.   Plus, I can then report that back upstream.
> There are all kind of protection circuitry but I am not so sure if there
> are good enough for not. If you consider short circuit protection,
> then what about other protection -- over temperature, over voltage,
> reverse polarity, etc?
Reverse polarity on the input side is taken care of - another P-MOS=20
fet.   If someone plugs the output into something else which is sourcing=20
power, they deserve what they get.     Overtemp should be handled with=20
the overcurrent shutdown - after all, I'm less than 1 amp through a 16A=20
rated MOSFET which should be normally fully on  (0.117), or fully off=20
except for a few us once every second and thus should be dissipating=20
well under 1/10th of a watt under normal circumstances - at most 4 or=20
5*C rise.

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