[GreenKeys] [External] Re: modified teleprinter

[Jones, Douglas W]

From: Harold Hallikainen via GreenKeys [greenkeys at mailman.qth.net] -- Sunday, April 3, 2022 8:23 PM

> On Sun, April 3, 2022 1:26 pm, s shumaker wrote:
> > Space Shuttle Teleprinter, described as "a modified teletype machine"
> > https://www.liveauctioneers.com/item/124634614_space-shuttle-teleprinter

> Interesting! I wonder what the printing technology is. The printer
> mechanism seems to be behind a metal cover. But, I can see a ribbon. 

The ribbon is about an inch wide and follows a diagonal path across the paper.  Furthermore, it looks like the platten is in front of the ribbon in front of the paper, with the hammers striking the paper from the rear.  I've seen this arrangement just once before, I have samples of material printed on it on the web here:
-- https://homepage.divms.uiowa.edu/~jones/plato/#printing

Look at the bullet headed "A mechanical impact screen printer."  That has sample output printed on the printer I remember.  Unfortunately, the print mechanism itself was an OEM unit, packaged by someone else.  My guess is that the shuttle printer used the same or a very similar OEM mechanism but I have no clue who packaged it or wrote the software (or firmware) to work it.  The print mechanism was a single horizontal line of print hammers that oscillated back and forth very quickly, we informally referred to it as a wobble bar.  Each hammer scanned one segment of the page, and carefully timed impulses to the print magnets would hammer the paper against the ribbon and flat metal platten.  It managed about 100 dots per inch resolution and was good for graphics or dot-matrix text.  Here's an extended document printed on that printer:
-- https://homepage.divms.uiowa.edu/~jones/plato/IBMofficeSystem6.pdf

It's just text (including custom fonts) up to page 14, but on pages 15 to 18, there is line graphics, with an isometric drawing on page 18.

Note that I am only speculating about the shuttle printer, but the basic appearance sure rings bells.  In any case, running such a printer requires a microprocessor and enough RAM to buffer the dot-matrix image of at least one line of text.  The PLATO one had 32K bytes of buffer so it could render a 512 by 512 pixel image into RAM before printing it.  It's easy to imagine NASA going for someting much simpler.

The key reason to like this particular printer design is that it had so few moving parts, just a stepper motor for paper advance, another for the ribbon, a simple DC motor to spin the crankshaft that moved the wobble bar, and a row of print hammers on the bar.  Everything else was software.  Very lightweight.

          Doug Jones
          jones at cs.uiowa.edu