Don's observation is an astute one, since a linear supply doesn't generally
allow both positive and negative supplies to interact. Unless this is a
MAJOR ground loop thing, we're probably looking at a switcher.
Has anybody tried powering on the drives AFTER the IMSAI box is in
steady-state?
Dick
----- Original Message -----
From: Don Maslin <donm(a)cts.com>
To: <classiccmp(a)classiccmp.org>
Sent: Tuesday, April 25, 2000 4:44 PM
Subject: Re: Shugart 851 power problem
On Tue, 25 Apr 2000, Tony Duell wrote:
> >
> > Hi,
> >
> > I have a pair of 851s in a nice case with power supply.
>
> One detail you've omitted : What sort of power supply is it (linear or
> switcher)? The reason I ask is that with most linear supplies, a short
on
> one output will not turn off the other outputs,
while with most
switchers
a short on any
output will shut the entire supply down (there are
exceptions to _both_ of those 'rules' though...).
It seems like it must be a switcher, Tony, as the 851s do not use -5v
power.
- don
> > They are attached to a Cromemco 16FDC in an IMSAI box
> > with a Cromemco ZPU and an Expandoram 64K board. I'm
> > using the ROM enable on the 16FDC to drive the phantom
> > line. There is a reason for these details...
> >
> > Here is the problem:
> >
> > #1 Power up the drives by themselves:
> > At the drives:
> > 5 volt line reads 4.9
> > -5 volt line reads -4.9
> > 24 volt line reads 23.9
> >
> > #2 Power up the IMSAI box:
> > At the drives:
> > 5 volt line reads 1.2
> > -5 volt line reads -0.9
> > 24 volt line reads 2.6
> > and (needless to say) the drives don't work. The Cromemco
> > ROM monitor gives all sorts of errors trying to access them.
>
> This looks like a classic case of 'latch up'. You may know about a
> component called a thyristor/SCR, a 4 layer semiconductor device with 3
> connections (anode, cathode, gate) that once turned on by applying a
> current to the gate connection will then stay turned on until the
current
> through the anode connection drops below a
certain (low) value.
>
> Well, it turns out that it's hard to avoid making these when making
> integrated circuits. You don't want them, but they're there. Normally
> they do no harm, but sometimes if a pin on a chip is pulled past the
> supply rail they turn on and short out said supply rails.
>
> And you could be doing just that. Signals on the controller cable from
> the S100 box could be latching up chips in the drives. Thus shutting
down
> the drive's PSU.
>
> What I don't understand, though, is why correctly-powered drives do this
> when a powered-down controller is turned on. I could understand if
drives
connected to a
powered-up controller would refuse to turn on. Nor can I
understand why turning off the controller and turning it on again makes
any difference.
#3 Power off the IMSAI box:
(Same as #1)
#4 Power on the IMSAI box again:
At the drives:
5 volt line reads 5.0
-5 volt line reads -5.0
24 volt line reads 24.0
and the drives work fine.
What gives? Does this make sense to anybody out there?
-tony