Path: utzoo!attcan!uunet!samsung!sdd.hp.com!hp-pcd!hplsla!tomb
From: tomb@hplsla.HP.COM (Tom Bruhns)
Newsgroups: sci.electronics
Subject: Re: Power Supply question
Message-ID: <5170115@hplsla.HP.COM>
Date: 28 Feb 91 17:43:41 GMT
References: <22940@hydra.gatech.EDU>
Organization: HP Lake Stevens, WA
Lines: 26

RE: Wiring regulators in parallel

Because each regulator would have a slightly different voltage it was
trying to regulate to, this is generally not such a hot (or maybe make
that too hot) idea.  If you look in a National Semiconductor linear
applications book or voltage regulator book, you will find some
suggested ways to increase the current output.  One way, for example,
is to put a resistance in series with the input to the regulator IC.
Assuming a positive regulator, you then connect a PNP transistor
emitter to power source, base to resistor-regulator junction, and
collector to regulator output.  Select the resistor so that the
regulator is drawing, say, 3/4 of it's rated current to provide a
drop across the resistor that will get the PNP turned on.  This is a
pretty simple circuit, and deserves a bit more consideration for
protection of the parts, but this should give you the idea of what's
going on.

Simplicity has its price:  for every volt you drop in the regulator,
at 10 amps you will dissipate 10 watts.  Are you prepared to get rid of
50 watts?  You better be!  An alternative is a switching regulator...
more complicated, but you are getting into the power range where it's
worth considering.  

BTW, if you use a linear regulator IC, you may well want to go with a
fairly high-current one, like an LM350K (3 amp), LM338K (5 amp), or
LM396K (10 amp) (but at $31 in DigiKey's catalog).