Byron A Jeff wrote...

>But the net effect is that one of the regulators will always be maxed =
out
>because it runs warmer than the others.

Yes, one of the regulators will end up sourcing most of the load
current; but it will do so because it has the higher of the two output
voltage setpoints, not specifically due to its being warmer.

However, you could also have some interesting thermal dynamics going on
with two regulators in parallel like this.  What would happen, for
instance, if you have two regulators connected thus, whose output
voltage setpoints were very close at room temperature but they have
slightly different reference voltage temperature coefficients and/or
error amplifier gain tempcos?  The regulators might share current
differently at different load levels or temperatures; or they could even
swap current back and forth, heating and cooling alternately as some
kind of funky "thermal oscillator"; or, if both regulators had rather
strong negative output voltage tempcos they could settle into an
arrangement where they actually would share the load more or less
evenly.  Not likely, but possible.

The bottom line, though, is what Russell McMahon said: provided the
voltage outputs of the two regulators aren't vastly different and they
are both current-limited, there shouldn't be any huge problems.

The only "should never" I'm aware of is, don't ever just connect two
regulators in this fashion and casually assume they'll somehow
automagically share the load equally; they won't. =20

Also, there are a few possible complexities that must be taken into
account:

One possible "gotcha" involves low-dropout ("LDO") regulators; these
tend to be rather picky in their input/output filter capacitor
requirements, with respect to both capacitance and ESR.  With the wrong
capacitors these things can become unstable and end up oscillating.
Putting two of them in parallel could end up making the input/output
filtering requirements even more complex.

A second possible complication involves the particular overcurrent
protection mechanism that ends up limiting the current in the regulator
that initially supplies the bulk of the load current.  If its output
current is limited by the internal current-limit circuitry, all is
probably well and the second regulator will simply step in and start
supplying the remainder of the load current.  But if the first regulator
ends up tripping its overtemperature protection BEFORE it goes into
current limit (for instance, because the package size is small and
there's a large input-output voltage differential), funny things can
happen.  Some regulators may implement this overtemperature protection
by simply cranking back on the internal current limiter when the die
temperature goes too high; but others may operate by actually shutting
down the regulator entirely and waiting for it to cool off, whereupon it
turns on again.  Obviously such on/off/on/off behaviour wouldn't be too
helpful in this situation, so a careful study of the datasheet is in
order.

>If you have an example of where regulator outputs are tied diectly =
togther
>and they equally share the load, I'd sure like to see it because that's =
a=20
>part I would use.

Like I said, both here and in my earlier reply to the OP, this business
of equal load sharing simply isn't too likely to happen. =20

=46rankly, I think it's a whole lot easier--and cleaner--just to get a
heftier voltage regulator if you need more current. =20

Probably be cheaper, too.

Dave

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu