ADJUSTED COLOR TENSION FOR RED/BLACK RIBBON COLOR
A pair of solenoids on the left side of the typewriter (when viewed from the front) pivot a lever up and down to apply more or less tension to a thin plastic ribbon. The ribbon runs over pulleys with both ends hooked to the carrier. This allows the carrier to move from leftmost to rightmost column position with the ribbon sliding over the various rollers.
The lever ends up pulling on a spring loaded mechanism inside the carrier that adjusts the lift height of the typewriter ink ribbon during a character print cycle. When properly adjusted the mechanism either positions the upper half or the lower half of the ink ribbon in front of the character being printed.
The adjustment is the rightmost roller (when viewed from the front of the typewriter) which can be moved left or right to add or remove tension to the narrow plastic ribbon. With the machine under motor power, I typed characters while adjusting the roller. I set the left side lever to the black and then to the red position, watching the resulting typed color.
RAN CONSOLE PRINTER DIAGNOSTIC BUT EXPERIENCING HANGS
Now that the ribbon color worked well and the tab interlock switch was functioning properly, I expected to have a successful run of the diagnostics. I loaded the memory image of the diagnostic into core and started it up. It began typing and tabbing well but then came to a halt with a wait code 30FA that indicates that the typewriter never cleared a busy condition.
Using a continuity tester, I found the microswitch that was causing the busy condition. That switch is activated during a shift operation, moving the type ball between upper and lower case hemispheres. Shifting the ball significantly changes the tension on the rotate band; if a print cycle were to take place at the same time, it would simultaneously change the tension. The result might be a snapped metal band.
What I discovered when looking at the typewriter was that it was partway through a shift operation. While the diagnostic was typing characters, it shifted the ball to get to a next character but the shift mechanism didn't finish its operation.
There is a pawl that will block the mechanism from rotating when the typewriter is in the midst of a print cycle. The pawl was jammed into the teeth holding the shift cam from turning, but we were not in a print cycle.
ADJUSTING THE SHIFT LOCKOUT MECHANISM
The shift mechanism is a circular assembly that stops in two spots 180 degrees apart. Solenoids trip the clutch allowing the shift mechanism to make the half turn to the other position. A ramp rising out of the plane of the shift assembly will pivot an arm with a roller that the metal rotate band moves around. The high side of the shift mechanism moves the roller out on the order of an inch. That added tension rotates the typeball to its other side. When the shift mechanism rotates to the side where the ramp is low, the roller moves back to release tension letting the ball twist back.
A pawl will lower to block the teeth of the shift mechanism from turning, which will interlock the shift while a print cycle is underway. This is driven by a cam on the end of a shaft that turns during a print cycle. The pawl is held out of the way when the cam is at a high point but will rotate down to block the teeth on its low points, mid print cycle.
The cam has two setscrews so that its position on the print cycle shaft can be adjusted. When the print cycle is not active, the cam roller should be riding on a high spot with the result that the interlock pawl is raised .030 to .060 inches above the highest tooth on the mechanism.
The cam was in the wrong orientation so that the interlock pawl was just a bit too close to the teeth so that it can jam the shift mechanism from turning even though the print cycle has completed. I got out the allen wrenches and turned the cam to ensure it wouldn't jam up the shift mechanism.
Next up - reload the console printer diagnostics and make sure that the typewriter performs everything properly.
I never cease to be amazed at the sheer mechanical genius embodied in a Selectric!!!
ReplyDeleteIBM was originally much more of a mechanical than an electronics company. Its earliest products were scales, clocks, punched card equipment and similar works of art, crafted with the precision and genius of Swiss watch makers. Some electrical parts were involved but the mechanical engineers were the true heart in the early days.
DeleteIt was only when IBM began moving into computers that the proportions of electrical/electronic versus mechanical engineering shifted. Still, in products from the 1950s, 1960s, and 1970s, a surprising amount of mechanical functionality was embedded in their computer products.
Two factors combined to cause them to choose mechanical solutions when electronic ones existed. First, the cost of even single transistors were non-trivial in the early days, so that mechanical parts were cheaper. Second, they had so much talent and experience with mechanical designs while electronic designers were in relatively shorter supply.
Thus, IBM will use microswitches or coils to sense positions on rotating cams rather than using an electronic timing circuit. The Selectric typewriters were very involved mechanical devices with little more than a motor plus an on/off switch. Over time, the introduced more electronics including micoprocessors into successor typewriters.