LGP-30 Memory Drum

Chuck Guzis cclist at sydex.com
Thu Jan 5 00:51:22 CST 2017


The biggest single plated-media drum that I ever saw was a Univac
FASTRAND II dual-drum unit (counter-rotating) that used movable heads in
an interesting mechanical setup using levers, linkages and solenoids to
decode an 8421-type binary position to a physical position of the heads.
 Said unit was hooked to a Univac 1108.  Univac used drums well into the
1970s, IIRC.  I think the FASTRAND II positioning mechanism was
described in a 1960s FJCC report.

Said FASTRAND was equipped with microphones (known as a "ping" detector)
to detect when heads hit the plated surface.  I suspect that an
oxide-coated drum would have turned the coating to brown dust in a short
time.https://www.cs.auckland.ac.nz/historydisplays/FourthFloor/ReportsAndViewer/Reports/MagneticDrum.pdf

The one I witnessed in operation was installed on the second floor of a
rather elderly building bordered by a busy street.  One of the problems
dogging the installation is that the heads would "ping" every time a big
truck passed by the building.
-----------------------
Having bored everyone with an old story, I did a little patent prowling...

http://www.google.com.gh/patents/US2820688 describes the manufacture of
a drum unit.  Iron oxide in a shellac carrier spread on a drum surface,
then machined to the desired profile appears to be described.

http://www.google.com.gh/patents/US2771595 describes the basic idea of
using a magnetic drum to store digital information.

https://www.cs.auckland.ac.nz/historydisplays/FourthFloor/ReportsAndViewer/Reports/MagneticDrum.pdf
 is an interesting document from ERA describing the magnetic drum

One of the issues addressed by several patents is the stability of the
bearings--apparently, slight irregularity in bearing construction (or
wear) can result in disaster.  Something to bear in mind.

If one were re-working an old drum, the traditional way of getting a
smooth even surface is to plate with electroless nickel, then machine
and polish the surface to the desired finish.   This works for any
substrate, as electroless nickel is an auto-catalytic process not
dependent on the electrical properties of the substrate.

FWIW,
Chuck






More information about the cctalk mailing list