Chuck said > Speaking of keyboards, were there any computer keyboards or typewriter > keyboards with interposer mechanisms such as used on IBM keypunches? I > recall that was one thing that had a very different "feel" from a > typewriter keyboard. It changed my keyboarding style.
There were some video terminals with mechanical Selectric keyboards, one for example being the RCA Spectra 70/752 from the mid/late 60s. For this RCA engineers substantially modified a Selectric 1 keyboard with many new bespoke mechanical parts to output 7-bit ASCII directly. The earlier RCA 6050 video data terminal and models 6051-1, 2, 3 Interrogator used a powered ASR33 keyboard, and the main console of the IBM Office System/6 workstation also used a powered Selectric keyboard. I've been working on a project for the last few years on and off to recreate this terminal's keyboard, starting with fixing a rusty seized Selectric II to operating condition then splitting the keyboard off from the power frame and powering the filter shaft with a motor in lieu of the normal print shaft gear train doing it. I machined out the RHS keyboard frame filter shaft (you can see this happening in this picture https://www.thingiverse.com/thing:4862890) to accommodate a new filter shaft cycle clutch partially made from IBM spring clutch parts (not a Selectric clutch but from some other IBM device as it has a urethane drive gear on it, just like the RCA one) and a new 3D printed clone of the RCA cycle cam/anti-backlash arrangement. The cycle clutch latch rod has been moved from the centre to the right hand frame edge as RCA did. Also made a new interposer from a scrap of sheetmetal and additional bail rod from my stash of Selectric bits for RCA's Bit 7 ASCII, and put spring returns on these as this was previously done by the Whiffletree mechanism in the power frame. So I'm about 2/3rds through it. Next is to finish the cycle detent and also recreate the key interposer ASCII microswitch rack with its 50-or-so millisecond-delay steel clamping plate (RCA called it the 'storage bar') that is driven off the filter shaft via two steel crank arms pivoting on a new axle rod RCA put across the lower rear of the keyboard. The plate momentarily holds the 7 clevis rods for enough time for the terminal's electronics to strobe the microswitches. Final power is to be an old 110v IBM AC synchronous motor from the 1960s I found in my dad's collection of bits that looks identical to the RCA motor, need to make the armature aluminium fan it had and make the standoffs to the keyboard frame. The pinion on it thankfully meshes with the urethane cycle clutch gear. In the meantime I power the keyboard with a LEGO M Power Functions motor with some 3D printed brackets and a printed LEGO-tooth-compatible gear with a boss that clamps onto the filter shaft just exterior to the new cycle clutch. Been taking lots of photos but I've not done a page on this yet until there's more exciting things happening with it (ie. actual electrical encoding) however it is fun to power it up and type with a 100% genuine-action Selectric keyboard. This being: Feeling the key drop, which pushes the key lever off its detent spring, sets up the bail pattern, then the filter shaft rotating to punch the key interposer's end and move the bails, and then reset the key lever on the upstroke. All with some of that satisfying Selectric 'clunk'. I say 'some of' because in building and operating this I've realised that a lot of the Selectric 'feel' people rave about actually comes from the print shaft, tilt/rotate tapes mech, shift cam, Whiffletree and typeball movement all operating in concert (along with its sound) in a fraction of a second, the keyboard mech is only part of it. Steve.
