On Sat, 30 Mar 2019 13:28:35 +0000, Stephen Wolstenholme wrote:

On Sat, 30 Mar 2019 12:02:20 +0000 (UTC), Martin Gregorie
wrote:

On Sat, 30 Mar 2019 21:04:14 +1100, Lucifer wrote:

I had a 5 volt UPS from a Data General mini computer which had a 200
amp HRC fuse.

Unsurprising: old computers ate power.

The CPU of 1902S mainframe I was sysadmin for in 1972 was entirely solid
state (discrete transistors) and with 32 kWords (96kB equivalent) of
ferrite core memory. That CPU used 20KW. Add in 6 tape decks, two disk
drives, printer, cardreader and a comms multiplexor and the system
consumed 40 KW when fully up and running - typical for kit in that era.

I never worked on 1902S but I worked on System 4 and the largest 2900's.
The current demands were very high but generally low voltage so some
engineers would work on live power supplies. Mending the 3 phase mains
primary supply in the middle of the night could be dangerous. The
secondary output was much lower voltage but high current. We kept an
adjustable spanner that had been dropped onto the 5.6V DC supply. It was
to remind some engineers of the power involved.
The spanner had blown like a fuse. I can't remember the current limit
but it was huge!

No System 4 for me - only 1900, 2903 (the 2900 DFC in a different orange
box running microcode to emulate a 1900), and a fair bit of time at the
BBC writing systems for their 2900s.

Somebody working in the computer room managed to drop a short crowbar
into the BBC UPS when it had its covers off, but caught it before it did
any damage - and then had to stand there holding it while the pair of
2960s were shut down and the UPS turned off. Were you there, by any
chance?


--
Martin | martin at
Gregorie | gregorie dot org