[vcf-midatlantic] PDP-11/10 PSU Repair/Warning
systems.glitch at gmail.com
Fri Jun 7 11:39:05 EDT 2019
Apparently some of the other DEC power supplies include a fuse on the AC
input for the 5411086 board! I guess they discovered the problem in the
field, too. It is a little ironic that the board that generates power fail
blew up :) That's why I'd suggested the analog circuit that generates AC
LO/DC LO -- both of them had gone away on the backplane. I also find it
interesting that DC LO really has nothing to do with the output of the
various independent regulator modules. You'd think since DC LO is open
collector, each module would be able to pull it down and it'd only be
released when all modules were stabilized. That is of course one of the
issues with memory write-protect lockout on nonvolatile RAM on S-100: you
have to wait til all three supplies stabilize, and then factor in enough
time delay for each board to also have stabilized before releasing *PRESET
and/or memory write lockout. I guess that's why they built a signal in for
just that purpose on IEEE-696!
The current draw through the connector should be 2-3 A, the combined output
from the 5411086 board's +15V and +8V supplies isn't supposed to exceed 4A
and is fused at 5A. It's switchmode (albeit kinda weird early switchmode,
using a uA723) so current out should be less than current in, even
accounting for efficiency losses. A 1A load on the +15V supply didn't
register on the (granted, somewhat insensitive) power meter I was using on
the AC transformer primary side. I plan to include a 0-15A AC current meter
on the "actually in a box" version of my AC test supply, for finer
measurement of module consumption at given loads.
I've also seen various old computers and other machines with fuses
bypassed, or something that's not a fuse in the fuseholder (cut off pieces
of bolt or rod are common). Not a good choice! I did once see a military
computer power supply with "battle fuse" switches -- the switches bypassed
the power supply fuses, presumably so that you didn't lose computer power
in a critical situation from a fuse breaking due to concussion from a shell
hit or something.
On Fri, Jun 7, 2019 at 11:29 AM Herb Johnson via vcf-midatlantic <
vcf-midatlantic at lists.vcfed.org> wrote:
> This caught my eye, so I looked at the schematic. Ironically, the
> shorted full-wave rectifier is shown on the "power-fail" circuit
> drawing. But the rectifier supplies all the unregulated DC voltage.
> Unfortunately the fuse F1 is *after* the rectifier and not before. Thus
> adding a fuse before the rectifier is wise.
> Bypassing the now-destroyed edge-connector pins is also wise. Did you
> determine the normal AC current on those pins? I suspect a few amps,
> thus those pins are at risk anyway.
> Thanks for calling this repair out. It's a cautionary tale about vintage
> power supplies in general; look for AC-side fuses, install if not. I"m
> looking at an IMSAI; the prior owner bypassed the AC fuse holder! ;(
> Herbert R. Johnson, New Jersey in the USA
> http://www.retrotechnology.com OR .net
More information about the vcf-midatlantic