Newsgroups: sci.electronics
Path: utzoo!henry
From: henry@utzoo.uucp (Henry Spencer)
Subject: Re: DC power supply for IBM PC?
Message-ID: <1989May28.214142.26384@utzoo.uucp>
Organization: U of Toronto Zoology
References: <4682@drivax.UUCP>
Date: Sun, 28 May 89 21:41:42 GMT

In article <4682@drivax.UUCP> macleod@drivax.UUCP (MacLeod) writes:
>... Since the internal AT power supply takes
>120Vac to produce +5Vdc @20A, -5Vdc @.5A, +12Vdc @8A, and -12Vdc @.5A,
>shouldn't it be relatively easy to cook up a modification to the power 
>supply (or a replacement for it) which would run directly from 12Vdc power?

Sorry, not so.  For one thing, don't confuse the 12Vdc from your power
source with the power supply's +12V output.  Your 12V battery will be
putting out something probably within a couple of volts (either way!) of
12, and it may be fairly rough if there's anything else on the line.
The power supply's +12V output is within a quarter of a volt or so, and
is fairly pure DC.  Getting the latter from the former is not simple.
Getting 5 from 12 is not so bad, since there's some headroom to work
with there, but if you want a simple circuit it's going to be very
inefficient and fairly bulky (translation, it will need considerably better
cooling than your current supply and probably won't fit in the same housing).

Your supply is most probably a switching supply which has fairly specific
expectations about its input voltage and doesn't use any intermediate
low-voltage-unregulated-DC step in the course of producing regulated DC
outputs.  There isn't likely to be any modification that will work; you'd
probably need a total redesign.

NB I am not an expert in this and may have missed something subtle.
-- 
Van Allen, adj: pertaining to  |     Henry Spencer at U of Toronto Zoology
deadly hazards to spaceflight. | uunet!attcan!utzoo!henry henry@zoo.toronto.edu