David;

A high side switch connects a positive supply to the + on a load.  The load is
thus ground referenced.
A low side switch connects a load to ground, so the load naturally connects the
second lead to a positive source.
When an N channel MOSFET is used as a low side switch, things are relatively
easy.  Drive the gate with 6 or more volts (usually) or 3 or more volts (for
logic level fets) and the fet turns on.  But then, you usually wire the car with
lights which are grounded, not common to +12V.  When driving a high side fet,
the simplest solution appears to be a P channel MOSFET.  All statements above
hold true, but oposite.  Pull the gate down with respect to the positive supply
a prescribed amount, and the fet is on.  But life is rarely that simple.  P
channel fets are more expensive and less populous.  Usually long lead times
too.  So you want to use an N channel fet instead.  BUT, you cannot get by
simply.  The source connects to the load, and the drain connects to the plus
supply.  Since you want to drive most of the positive supply into the load
(minus a few hundred millivolts), the gate, which is driven positive with
respect to the source, must be driven higher than the positive supply.  Say
12V+6V for a non-logic level fet.  Well, you normally do not have a voltage
available that is this value in the auto.  Enter high side switches.  They have
a boost pin, which uses a diode and capacitor to pump up a positive supply with
respect to the positive rail.  This is significantly less expensive than using a
DC/DC converter.

There is more, but this is the bulk of the pertinent facts.

Chris Eddy

"David E. Olson" wrote:

> I've been perusing some of the ways that I can drive higher current and
> voltage (automotive) things such as lights, and the multi-channel
> high-side/low-side switches (from Infineon) seem to fit the bill. I've
> fooled around with them and they do work for what I'm doing, but I'd like to
> add the theory to my practice.
>
> However, my confusion is to what's the general difference between the
> high-side and a low-side designations. It's obvious from the spec sheets
> that there's a bunch more current running through the high-side switches. Is
> there more? Same thing applies to other companies designation of a
> high-side/low-side driver (I'm assuming they're similar to Infineon's
> switches).
>
> Bummer that my Art of Electronics book doesn't have a reference to them... I
> guess I'm a dazed software guy in hardware hell...
>
> -DO