With realism of the experience for a person using the 1130 replica as a key objective, it was important to recreate as much of the system as reasonable. That is one of the reasons to build a physical panel with the blinking lights at full size, as well as the rationale for using a Selectric typewriter mechanism for the console output. This entry describes the process of acquiring and/or building the appropriate parts of the machine.
Scrap parts for some mainframe systems are still plentiful in the used markets. IBM tape drive units are particularly active on Ebay at this time, where the operator panels include the same kind of switches and lights as are used on the 1130 - plastic rectangles in red, green, and blue for the pushbutton switches and rectangular plastic blocks in white, yellow, red and green for the lights. Some of the buttons are usable as purchased, for example the "Start" and "Reset" buttons, but others will need the wording altered to match the 1130.
On the console pedestal, alongside the blinking lights, are a rotary mode switch on the right and a red emergency pull switch on the left side. I found a suitable rotary electrical switch to actuate the mode switch, allowing selection of 'single step', 'single storage cycle', 'single instruction', 'normal', 'interrupt run', 'load' and 'display' modes. I am using a scrap part knob from a different IBM mainframe to make the mode switch appear as accurate as possible. The emergency pull switch is a problem, even though it will not be operational, because it is so iconic for IBM mainframes. These are apparently highly prized souvenirs and extremely difficult to find. At this point, I expect I will be building a replica to install on this 1130.
I have described the challenges of converting a Selectric typewriter mechanism to server as the operator console, but a brief discussion of some additional factors is warranted. I used Craigslist to pick up a Selectric II, bought solenoids from an electronics supply house, and then bought an Electronic 50 as that is potentially easier to convert. However, the mechanism is housed differently, since there is no typewriter keyboard on the 1053 console printer. I will modify or build a housing that matches the 1130, place it around the converted typing mechanism, and integrate the tabs/carriage indicator. I will need to add a rocker for tab set/release, as well as button switches for Tab, Space and Return. Finally, the front of the printer also features 16 white tipped, silver handled toggle switches called the Console Entry Switches. I bought large toggle switches with extended handles, but will need to convert the ends with white caps to match. Most likely, I will grind down the ends to a stub cylinder, then glue on a white plastic cap that I make with a 3D printer.
The keyboard is a mechanical device with the unique heavy feel of the 029 Keypunch whose mechanism it leverages and the unique key layout and key shapes. I am currently trying to buy a very similar keyboard on Ebay, which is photoelectric not mechanical but otherwise identical in appearance and layout because it was built as part of a keypunch system. The initial price was just too high for the realism it provides ($350 plus shipping), especially because I can't feel the action of the keyboard in advance and don't know if I can implement the interlocking and need for a reset key to release a 'jammed' situation. I sent emails to the seller offering a lesser amount, but the buyer has not deigned to reply, although it was relisted a few times and reduced to (a still too high price of) $299. After the last auction elapsed, I emailed one final offer of $200, probably way too much but I am still receiving deafening silence from the seller. Sometimes people have firm ideas of what they believe their goods are worth, in spite of continued market evidence to the contrary, and they become offended by any suggestion of a lower price, not even bothering to counteroffer.
The fallback is to create a keyboard using mechanical switches, some foam or other material to increase the keystroke resistance, and some method yet unknown to create the lockout (jam) condition. Another alternative is to separate the keyboard portion of the electronic 50 typewriter, which is already fitted with microswitches, and make use of some form of that as a mechanical keyboard. This would certainly provide the interlock and need for a solenoid to release 'jams', but requires custom keycaps and may be difficult to map exactly to the 1130 keyboard layout. More research will be needed.
The tabletop will be wood with white laminate, similar to the actual 1130 table, but not full height off the floor. Instead, it will look as if someone had used a chain saw to cut off the top of an 1130, leaving just six inches or so below the lowest edge of the tabletop. The metal portions will be copied, as these are all painted and can be replicated pretty well - pebble finish except for the plate atop the keyboard which is smooth - and in authentic colors.
I have access to real 1130 systems at the nearby Computer History Museum and also at a nearby private collector's home, allowing me to measure more accurately, test coloring and finishes, and test out the keyboard action. The CHM machine can't be touched, although I might be able to arrange permission using gloves, but the collector's machine is going to be fully accessible. I would like to check out the intensity and behavior of the lights on the pedestal console display, and to record the noises of the 1130, but neither of these machines are powered. There are two powered machines I might get access to for the light investigation and recording - one nearby at a second private collector's home and the other near London in the UK at the National Museum of Computing.
I am producing peripheral replicas as possible - paper tape, plotter, card and printer devices are all possibilities. Initially the plotter is handled by a clever unit built by Richard Stofer that converts the plotter instuctions of the 1130 adapter into HP-GL commands over TCP/IP, allowing it to drive any HP-GL compatible plotter or printer. If I buy a plotter that could be forced into a replica housing, I might be able to produce the 1130-like plotter, but at worst it will be used to link to a HP-GL compatible laser printer.
I have bought a paper tape reader mechanism on Ebay (it was how I came to meet one of the private collectors in the area) which I will fit into a replica housing to look like the 1130 peripheral. I bought the punch block for a teletype - the dies and fitted holes that are driven to punch holes in paper tape - which I only need to fit to solenoids and install with a sprocketed driving mechanism in order to build the paper tape punch. If this pans out, I will also need to buy some blank paper tape stock on ebay.
Punched Card equipment is in short supply, but I may be able to create some form of punched card input/output. There are occasional readers or punches appearing on ebay - right now there is a duplicator (both reading and punching plus the card transport capability) but the offered price at $495 is a bit high since there is no guarantee it is in working condition. I have a mechanical keypunch (Wright 2600) and a box of 2000 blank cards, both bought on ebay, thus I could move forward on the reader/punch if I choose.
Line printers are difficult to build as a replica, mostly due to the lack of continuous form paper supplies. and the fact that chain or wheel based printers have long been replaced by other mechanisms like laser printers, but the right basic mechanism might show up somewhere and trigger a subproject.
Disk IO on the 1130 uses the 2310 mechanism and the 2316 disk cartridge. Used cartridges are found on ebay periodically, but not the drive mechanism. However, several other makers built similar products, including DEC (RK-05 disk), Pertec, and CDC, any one of which could be pressed into service. The unique 'grunting' sound of the actuator might need to be reproduced, since the mechanisms for seeking and positioning are different on the compatible drives, but many options exist. First, I could put flash memory inside a disk cartridge along with zigbee or similar wireless, allowing the cartridge to be placed into a mock disk mechanism and accessed - playing some simulated disk sounds as it operates. I have bought an IBM 8" floppy drive mechanism as it seems to use a mechanism most directly derived from the 1130's disk, thus it might provide the right grunting sound. I could use a floppy itself as the disk cartridge, or just use that part of the mechanism for authentic sounds - more investigation and design work is needed.
Other peripherals such as the 2250 graphic workstation, optical mark reader and alternative readers and printers, are not in scope at this time.
No comments:
Post a Comment