Close

It's Gluebricant!

A project log for IBM Selectric III Restoration Project

I got myself a Selectric typewriter in a pretty bad shape... This is gonna be a rough ride!

keri-szafirKeri Szafir 12/18/2022 at 09:390 Comments

The simple and obvious problems I mentioned before are nothing compared to the fact that this typewriter has been standing for years with no proper maintenance and oiling... Same problem as I once had repairing a vintage record player - the grease became sticky and viscous, turning into glue. So, Gluebricant. Even taking the typing element and cassette off was a problem; it just needed so much force I was afraid to damage the machine's parts. Same with getting the print rod to turn. What is more, the keyboard is not working (keylevers not engaging the interposers), and I've got my suspicions about the cause... There's a clever interlock mechanism named compensator tube, which has slots the interposers go into after they're pressed; this tube is filled with ball bearings with only one interposer's thickness of space left. Whenever an interposer is in the tube, it pushes the ball bearings aside and goes between them, leaving no space for any other interposer, thus locking all else out besides the one for the key currently pressed. Clever, clever. There's an additional interposer for keyboard lock when the typewriter is in the off state (it's operated by the power switch mechanism), and my suspicion is that if these bearings were lubricated, and the grease or oil became sticky, the balls won't move and no interposer can go fully in.

Also, the typewriter has its carriage return mechanism stuck in the engaged position, pulling the carriage all the way to the left not disengaging after it's done. One of the levers is in the upper position and after I put some force on it to make it go down, the carriage is left in peace.

Anyway, it's mostly just taking deep dives into the service documentation, analyzing the mechanism schematics and operational principles, hypothesizing about causes of problems and testing these hypotheses. Makes me wonder if anyone in the hacker community had as complex a restoration as I'm having now.

Discussions