🚀 go-pugleaf

Another story of the PDP-11/20 at the University of Illinois in 1973:

Sometimes the machine would stop working unexpectedly.  When this happened,
one of the common fixes was to give the CPU a swift kick from below before
trying to reboot.

The CPU was mounted in a standard relay rack, and there was open space below
it, so it wasn't hard to give it a good kick.  Those who didn't know what
was going on were usually quite puzzled.

This worked because the backplane was on top, with the CPU cards hanging
vertically from their card-edge connectors.  The bottom panel of the machine
had a foam backing that was supposed to keep the cards seated, but with the
vibration of the fans and whatnot, the cards tended to slowly sag out of
their sockets.  The kick re-seated all the cards quite effectively.

The design with the backplane on top was actually rather elegant, and if the
cards had been retained by something stronger than friction and foam, I'd
have liked it.  The box slid out of the rack on glides, then could be tilted
up for access to the cards or tilted down for access to the backplane wiring
(all wire-wrap, if I remember correctly).

					Doug Jones
					jo...@herky.cs.uiowa.edu