off-the-wall question (possibly OT)

From: Christian Fandt <cfandt_at_netsync.net>
Date: Fri Apr 2 20:31:04 1999

>
>>Has anybody here ever heard a relay machine running, like one
>>of the early Zuse machines or the Harvard Mark-1? It occurred

Don't know much about the Harvard machine but I can sort of imagine what a
Zuse would sound like from seeing models and info at the Deutsches Museum.

A somewhat similar comparable thing I recall really well is hearing an old
relay/vacuum tube (valve) numerical position controller operate. The
General Electric Century Mark II NC controls were a mid-1950's design. We
had two of them, each running a large Burgmaster NC drill in our machine
shop. This was from 1978 to 1983 and I was responsible for troubleshooting
those danged tempermental things. They were tired, old, raggy hand-me-downs
from a sister Bausch & Lomb division who once did most of the machine work
for B&L's Scientific Optical and Professional Products divisions.

Paper tape was read by a Friden reader which was not all that quiet. As
each digit was read off the tape, a corresponding 10-position stepper
switch would step to that digit. More noise. After the last digit of the
possible 12 was stepped there was some clatter from the dozens of
hermetically sealed relays in the relay logic area and the error voltage
was set for each axis. The steppers would then immediately reset to zero or
home position. That was a bunch of noise. The servo drive amplifiers would
move the table to position. Relatively quiet. The drilling, boring or
tapping function would be performed. Some noise depending upon spindle
speed, feed rate and material. Then the cycle would start once again for
the next location on the workpiece.

For a very complex part or group of parts in a holding fixture and if only
simple holes were being drilled, *and* if the NC control was working well,
each cycle could be only 5 to 10 seconds long. If the number of holes being
drilled was, say, 30 to 50 or so one could imagine the sweet
buzz-buzz-bizz-bizz-buzz, clatter, clickety-click-clack, bzzzzz sounds of
an electromechanical NC positioning system at work going on for significant
periods :)

So, this experience helps me imagine a little of what the old relay
computers may have sounded like at certain times doing certain operations.
I spent *many* hours troubleshooting these NC units while in that noisy
machine shop trying to hear certain relays faintly click at just the right
time while holding scope probes in both hands (looking at other relay
outputs or servo signals, etc.) and operating the 'Start' button with my
foot (yes, it all looked strange). Relay races because of faulty contacts
were a common problem. The design relied upon very close timing of relay
contact closures or openings. The relays could not be opened to check plus
they were getting extremely expensive by 1982 (approx $100 each.) Other
than a scope, using the right cuss words occasionally helped too.

Only one other person at that time knew those controls as well and that was
the GE Field Service Engineer for the Mark II's who took care of most of
the Eastern USA *just* on those controls. I still have the printsets
somewhere around I think.

In a quiet shop with the control cabinet doors wide open, even carrying on
a normal conversation next to the controls was sometimes hard depending
upon the hearing ability of the person I was talking with.

>>to me (please don't ridicule me too much for this*) that the
>>sound of the robot in the old "Lost in Space" TV series may
>>have been intended to sound like a relay machine. I've heard
>>descriptions of them sounding like a thousand women knitting,
>>which could sort-of describe that robot sound-effect as well...

The Mark II's sounded like only a hundred women knitting plus a small
telephone Central Office switch and a Teletype.

>>
>> Cheers,
>> Bill.

Regards, Chris
-- --
Christian Fandt, Electronic/Electrical Historian
Jamestown, NY USA cfandt_at_netsync.net
Member of Antique Wireless Association
        URL: http://www.ggw.org/awa
Received on Fri Apr 02 1999 - 20:31:04 BST

This archive was generated by hypermail 2.3.0 : Fri Oct 10 2014 - 23:31:38 BST