2 Mar
2009
2 Mar
'09
10:15 a.m.
Some S100 chassis had multiple primary taps, CVT or a in the case of
Inetgrand {and a few others} boxes regulators.
The problem is the early machines had to have really robust transformers
as the +8V bus could easily be 25A using early boards. By 1980s with
64k ram cards (64k on one card) and multi serial IOs on one board a
typical system might take from 3-5 board at far lower current than the
earlier 4k ot 8k based memory systems. The side effect fo that was the
bus voltages being upregulated would float higher and regulators would
run really hot. Near the latter days typcail systems were system on a
card (z80, 64k or even 128k, 2 serial, FDC and printer port) making the
whole load typically under 3-4A at +8V and the need for more boards
unlikely.
Allison
Subject: Re: S-100 power supply voltage ranges
From: "Barry Watzman" <Watzman at neo.rr.com>
Date: Mon, 02 Mar 2009 10:42:55 -0500
To: <cctech at classiccmp.org>
Actually, the factory fix was a "bucking transformer", a whole second power
transformer installed out of phase in series with the primary power
transformer so that the voltages were subtractive. There is a factory ECN
on this topic. But the bridge regulator solution was common among the user
base (I myself have used it) as a lot simpler and cheaper, and almost as
effective.
------------------------------
Message: 3
Date: Sat, 28 Feb 2009 11:07:56 -0600
From: Jim Battle <frustum at pacbell.net>
Subject: Re: S-100 power supply voltage ranges
To: General Discussion: On-Topic and Off-Topic Posts
<cctalk at classiccmp.org>
Message-ID: <49A96F6C.6060602 at pacbell.net>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Chuck Guzis wrote:
All of the S100 boards I've seen use simple
linear regulators, so you
need some headroom--but not too much. A lot of S100 7805s and LM323-
5s were operated within an inch of their maximum current ratings (some
had bypass resistors installed) and generated a lot of heat. I
wouldn't run the supply rails any higher than I had to.
Cheers,
Chuck
The Sol-20 had an app note about this problem. Apparently they shipped them
for a while with transformers that were out of spec with the end result that
the unregulated power ran high. The recommended fix was to take a power
bridge, mount it to the case, and wire it in series with the positive
voltage run in order to get a ~1.4V drop.