FOUGHT FOR TWO HOURS TWEAKING SHIFT MAGNETS
I could sometimes get the magnets set up so that for a few times the machine would properly shift to upper case and hold the ball there until a lower case letter came along later but then it went back to refusing to trigger upshift.
I can feel a bit of resistance in the trigger mechanism, the roller arm that is pushed by the armature of the upper case, magnet. Depending on how still it is, the magnet can trip the upper case but not latch in to place, trip and latch, or fail to trip at all. The settings are very very finicky, which is not correct. There should be a reasonable range of adjustments that will work properly.
I spent some time lubricating and checking the trigger and shift clutch, hoping to find and eliminate the sporadic stiffness. I was not successful when I ran out of time on Friday. I had spent most of the day helping out at the US Space Force Museum in their Sands center just outside the gate of Cape Canaveral, so I didn't have much shop time that day.
CORRECTED THE SHIFT BETWEEN CASES
The next day I was able to coax the magnets into the magic orientation so that they would reliably fire off a shift when it was necessary to change between upper and lower case sides of the typeball. I had previously dialed in the ribbon color adjustment and thought that with this function now working, the 1053 was truly done, the last item for the restoration of the IBM 1130.
PEELED ANOTHER LAYER FROM THE ONION - NEW PROBLEM EVIDENT
Alas, as I was testing the shifting I realized that something odd was happening. The design of the controller logic will fire off a cycle for shifting automatically when the new character is on the other side form the prior printed character, then print the new character.
Indeed, when the prior character was lower case and the new one is upper case, I would see the typeball spin and the upper case character printed on the paper. However, when the prior character was on the upper case side and the new character is on the lower case side, I would see the upper case version of the new character printed and then the shift take place.
I studied the logic diagrams but everything related to shifting was totally symmetric, with no reason I could see that the print would take place before the shift. After some time, I did come to understand that the logic did fire off the shift, but would then trigger the print before the shift had moved much. A protective mechanical interlock on the typewriter holds the shift drum from moving while a character print cycle is underway, which is why it froze and then finished after the print even though it electronically was commanded first.
The cue for the controller logic to fire off the print cycle is that microswitches on the typewriter indicate that the shift has completed - or actually they show when it is busy and other actions must be held off. I got out my hand crank tool and checked the timing when the switches opened and closed during the shift cycle, but they were pretty reasonable. They matched my working 1053 as well.
It was then that I noticed that the shift mechanism was leaping forward during the UC to LC shift operation, which caused the microswitch to falsely report the shift operation complete. I did a short video of the shifts, first the LC to UC type and then the failing UC to LC.
I carefully studied the principles of operation for the 1053 as well as the working 1053 I own and the donor parts. I found that the shift clutch has a nylon brake shoe that slows the movement of the clutch from UC to LC, but that is not doing its job on this typewriter. The fix seems pretty straightforward, so when I get back to the shop tomorrow I expect to finish this (hopefully) last little issue off and move on to extensive diagnostics of the restored 1130.
 |
| Nylon brake shoe above shift clutch drum |
It appears that I just have to loosen the mounting screws and move the brake shoe to cause the shift cam to move at the correct slow speed.
No comments:
Post a Comment