Xref: utzoo comp.periphs:2312 misc.wanted:7012 sci.electronics:8722
Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!gem.mps.ohio-state.edu!usc!apple!sun-barr!decwrl!shelby!neon!kaufman
From: kaufman@Neon.Stanford.EDU (Marc T. Kaufman)
Newsgroups: comp.periphs,misc.wanted,sci.electronics
Subject: Re: UPS Information Needed
Message-ID: <1989Nov22.163150.26582@Neon.Stanford.EDU>
Date: 22 Nov 89 16:31:50 GMT
References: <1989Nov17.140437.13421@null.uucp> <15396@vlsisj.VLSI.COM>
Sender: USENET News System <news@Neon.Stanford.EDU>
Reply-To: kaufman@Neon.Stanford.EDU (Marc T. Kaufman)
Organization: Computer Science Department, Stanford University
Lines: 17

In article <15396@vlsisj.VLSI.COM> davidc@vlsisj.UUCP (David Chapman) writes:

>The best way to get zero switchover time is to buy a UPS that supplies power
>from a DC-AC inverter running off of a battery.  The battery is then charged
>by an AC-DC power supply.  This is the way our UPS runs; it has about 10 VAX
>minicomputers and 150 workstations hanging off the back.  Of course, it has
>a room full of batteries too.  :-)

Best Power (address previously posted by someone) uses a Ferro-Resonant
isolation transformer with a third winding for the inverter.  There is enough
stored energy in the transformer that the inverter can remain off until the
main power quits, then turn on to provide backup power.  A couple of side
benefits are that the battery derived power is sine-wave, and the mains
power is cleaned up a lot by the transformer.  They cost about 150%-200% of
a TrippLite, however.

Marc Kaufman (kaufman@Neon.stanford.edu)