From: Winfield Hill
Subject: Re: zero-power toggle circuit; was, how to master electronics
Date: 15 Nov 2002 03:00:02 -0800
Organization: Rowland Institute
References: <3DC517EB.firstname.lastname@example.org> <3DC57B4E.email@example.com> <firstname.lastname@example.org> <email@example.com> <3DD39D90.firstname.lastname@example.org>
X-Newsreader: Direct Read News 4.10
>Winfield Hill wrote:
>> High-side switching is dangerous, unless an effective form of
>> short-circuit protection is provided, e.g. shutoff, foldback
>> current limit, a quick fuse, etc. Low-side switching is used
>> in cars, etc., because it has the nice property that a wiring
>> short simply turns on the switched device. Recommended.
> I would have thought in wiring terms cars are mostly high side switching.
> With a -ve earth chassis used for a return most wires between switches
> and loads are +ve.
OK, I've been poring over the Chilton and Hayes chassis wiring manuals
for my Cougar, and yes you are correct. I was able to find only a few
instances of low-side switching, used by the engine-control module for
a few solenoids. All of the important switching was high side.
I'll withdraw my comment asserting a lo-side switching preference, but
I'll stick by my assertion that an effective shutoff protection scheme
is required for high-side switching. A fuse alone is likely inadequate,
because in the event of a complete short many FETs may or will become
damaged before the fuse blows. One can add a current-sensing resistor
and some circuitry responding to the voltage across it, etc., but a more
appealing approach is to sense the voltage across the FET's drain-source,
which should be very low, say under 0.25 volts, etc. and respond to an
excess voltage. It's nice if the protection circuit can respond to FET
heating energy, an integral of time * voltage squared, to determine the
Another issue is whether the FET switch should remain shutoff until the
power is cycled, or if the switch should continually retry the shorted
load after a short cooling period. The latter scheme may be necessary
for some critical applications, but it does put more strain on the FET.
> Low side switching is frowned on in industrial controls systems because
> of that property. It is no doubt made illegal by some bloody european
> safety regulation or the other. Having things turn on due to the most
> common type of wiring short is generally more dangerous than having
> them turn off although personally I don't think it is anything to get
> that worked up about.
I take your point, sounds good.