IBM Selectric with mechanical key linkage to print mechanism |
The 1130 makes use of its console printer by sending the Tilt and Rotate codes, the adapter causes these to pull on solenoids at the right time. A solenoid is an electromagnet which pulls in on a metal cylinder - the other end of the cylinder is attached to some lever in a mechanism. Current into the solenoid produces a mechanical movement. Each of the several tilt and rotate rods has its own solenoid, and the basic print cycle lever has a solenoid. The adapter activates the tile/rotate solenoids and then fires the print cycle.
If another key were pressed while the selectric was still doing the prior character, or a key was typed while some activity like a carriage return were occuring, the mechanism would jam. On the physical Selectric, mechanical means are used to lock out keystrokes until they are safe. Shifting the typewriter to upper case is going to cause a hard jam if the unit is already in upper case condition, thus these kinds of conditions must be avoided too. The way this gets handled on the 1130 is by inserting microswitches to report such conditions, and by having some microswitches activate to report where the mechanism currently is within the rotation of a power rod, such as the one that powers the basic print cycle. The adapter takes those indications from the switches plus uses a timer to delay the minimum time needed for the print cycle to complete, all to allow safe printing. Tabbing, backspacing, indexing down a line, shifting to upper/lower case, returning the carriage and other actions are handled by different power rods triggered by their own levers - yanked by appropriate keys on a typewriter and yanked by solenoids on the console printer. The 1130 adapter sends signals to solenoids installed for all these functions as well.
Microswitches providing feedback from E50 Selectric |
The mechanical keyboard is not used on the 1053 console printer, with the solenoids accomplishing the hooking and pulling of rods and levers that the key movement would accomplish on a plain typewriter. The 1053 printer corresponds to the IBM Selectric 2 - the typeball used on the 1130 computer is based on that version of the typewriter.
I own a Selectric 2 that I intended to modify by installing my own solenoids and microswitches. The solenoids are already purchased, but not the switches as the detailed plan for placement was still to be done.
Solenoids I intended to use to convert Selectric 2 into a console printer |
I found an IBM Electronic 50 on ebay - this is a much newer Selectric based system that has solenoids and switches already installed, since it has decoupled the keyboard from the printing mechanism.
Electronic 50 has keys separated from print mechanism |
Seolenoids and circuit board in Electronic 50 Selectric |
However, two issues remain that must be overcome. First, the typeball used on the E50 is incompatible with the Selectric 2, featuring more character positions to cite one difference. Thus a typeball from an 1130 won't work properly. As I hoped to find an APL typeball and run 1130 APL on my replica, this will be a barrier. Second, the microswitches and means of reporting the rotary position during operation of cycles, such as a basic print cycle, are different. Optical pulses are sampled by the circuit board in the E50, rather than microswitch contacts being activated. My interfacing my require some custom microswitches or an adapter circuit to convert to the kind of feedback signal expected by the 1130 console printer adapter logic, which believes it is connected to a 1053, not an E50. I may need to produce one or more custom typeballs or translate tilt/rotate codes on the fly in order to produce the intended printed characters from the E50 typeball.
The IBM solenoids operate at 48V, requiring me to design and build a circuit card taking 3.3V logic inputs and activating the solenoids with 48V at sufficient current to work properly. That is still in design, but I will build a prototype board shortly to begin testing the integration of the E50 into the 1130 system.
No comments:
Post a Comment