IBM 1130 Restoration at TNMOC in 2013

15 February 2013 Update from Peter Vaughan

• Work on Console printer by Leny still progressing.
• All other work on hold until printer is working.
• In the next few weeks work on designing a console printer emulator and printer tester will start with help from Olly and Leny. This will enable the 1130 to be run without a printer and also allow the printer to be tested stand-a-lone.

12 March 2013 Update from Peter Vaughan

• A new volunteer, John Robertson, has joined us. He was an IBM engineer who used to service IBM Selectrics and IBM 029 punches and has agreed to look at the two we have. He spent Saturday afternoon looking at the 1130 console printer and has made some progress in fixing / lubricating the print head mechanism. Still a lot more to do on it but a promising start.
• Leny is still progressing with the other console printer in Leeds.
• Preliminary work has started on the console printer I/F box. The first hurdle to get over is finding a suitable connector that plugs into the existing ribbon cable connectors on the printer. The connector is an IBM special 2 x 12 pin but uses 1/8" pin spacing (3.175mm) which is pretty much non-existent today, so Olly is trying to find something that will fit it.
• I have managed to source 3 disk packs from the US which should be shipped in the next day or so. I'm also trying to get 3 packs from Stuttgart museum (who have a working 1130) with pre-loaded DMS software and a means to copy the packs to card decks. This is still on-going.

22 March 2013 Update from Peter Vaughan

• Decided to connect up the console printer after work was done last week to see if it does anything. The line feed is still intermittent but I did manage to actually get it to print something for the first time using the console / keyboard test card deck. However due to it not using the correct print head it printed rubbish. I will need to draw up a conversion chart to map the 1130 golfball with the typewriter one we are using to see what it is actually printing.
• The 3 disk packs have arrived from the US. They will need cleaning and checking.

09 April 2013 Update from Peter Vaughan

• Console printer tried again with printer/keyboard test deck but unable to complete any tests due to test stating the printer was busy. This could be a printer fault or a logic fault in the 1130. Once the console printer is working better this will be investigated further.
• Tried again the following week with a test program written by KevinB entered via the console switches. This allows a single key code to be entered on the console switches which is then printed each time the program runs. Printing did occur (and the expected chars) but the quality of print was poor and was part black and part red. After printing a few chars the rotate tape jumped off it's reel so end of testing.
• John Robertson visited again and took the tilt/rotate mechanism apart, degreased and re-lubed it. The golfball head now moves much better with no sticking. When trying KevinB's program again the print quality was very good and with a lever adjustment the printing was all black. Rotate tape came off reel again, possibly when the case of the character changed so this will need further investigation. paper feed still not working properly.
• The 3 disk packs received from the US look in good condition with a few signs of the head touching the surface (which has been seen on all the packs I have inspected so may not be an issue as such). They will need a proper clean by hand before they can be used.
• The 029 punch is currently not working after John spent some time servicing it on his second visit. He spent his last visit on the console printer but he hopes to get the 29 working again on his next visit.
• The 1442 reader/punch is not working due to cards sticking in the cornering station. Adjustments will be needed to fix it.

11 May 2013 Update from Peter Vaughan

• 1442 card reader was mis-feeding cards at the cornering station a few weeks ago but the problem has now gone away - the higher temps may have something to do with it.
• Work on the Selectric console printer is still on-going by John, with times when it will print and times when it will not. Trying to run the console test card deck or any other test deck is failing due to the printer appearing to be busy as reported by the test programs. This could be a console printer fault or a logic fault. Work is continuing.

12 May 2013 Update from Peter Vaughan

• Having sourced a non-working Selecric typewriter off e-bay for spares and spent a little time looking at how it worked, I decided to look at why the space and tab functions on 1130 console printer were not working. Both functions were controlled by a single cam and after hand-cranking for a while and comparing it with a similar cam for the carriage return I noticed there was a spring missing. I managed to find and fit a replacement from Olly's spring box and with a bit of lubrication and cleaning both the space and tab functions now work reliably.
• What was unexpected is what happened when I switched on the 1130 and ran what I thought was the normal CPU test program still in core. After a few seconds the console printer started printing... the CPU test does not use the console printer so I was not sure what was happening... I then remembered I was trying out the 1132 printer test program a few weeks ago but kept getting the busy condition. I then restarted it and switched on the 1132 and the console printer started printing out lots of messages, and continued to do so... the console printer was... wait for it... actually working!!!! It would appear that the test programs used the tab function on the printer and because that was not working the printer never completed the operation and thus appeared to be busy.
• One small issue through... the golfball used on the 1130 is a special one but it's fixing tab is broken. For now a standard typewriter golfball is being used.. the result being what is actually being printed is complete rubbish. So this will slow things down somewhat but at least we have the chance to convert the text being seen now and start making progress on getting the rest of the system working.

18 May 2013 Update from Peter Vaughan

• Managed to fit original 1130 golfball head to the selectric so testing could continue in the short term.
• Ran console / keyboard tests and after an issue with the paper feed causing a paper-out condition it passed all the tests. There are a few character alignment issues that John needs to look at but its working enough to run other tests.
• Ran the SCA (Synchronous communications adapter) tests and they all passed as far as I could test without a system at the other end, so it looks like the Sync serial interface is working. This may give us an option to get programs into the system and out to punched cards.
• Ran the 1132 printer function tests but this failed early on with emitter errors. The error codes produced do not give much info or what the emitter codes are so further reading needed to understand the fault.
• John found the cause of the 029 punch not working... the contact used to detect when a card is engaged in the punch station was not working reliably. Once the contact switch was adjusted the punch started working.

25 May 2013 Update from Peter Vaughan

• Spent time checking the signals from the 1132 printer. The first to be checked was the CB signal generated by a photocell on the outer ring of the printer emitter disk (a disk on the main print drum shaft with holes at set intervals which are detected by shining a light through the disk and detecting it with photocells). The raw signal looked a bit weak but the amplified and cleaned signal was all over the place and not consistent with the number of holes on the emitter disk. After adjusting the lamp brightness things then looked much better. I then checked 8 bit signals generated from the 8 rings in the emitter disk. These 8 bits produce a code representing the next character on the print drum. These looked OK so the printer test program was run again but emitter errors still occurred. The emitter disk was very rusty so it was carefully cleaned with a wire brush and the test repeated and, the printer started printing. Another milestone achieved!
• Initial runs of the printer test program produced about 15 missing print columns and some of the positions were constantly printing. After repeating the printer test for an hour or so all but 1 column was printing characters, although some seem to be printing the wrong character consistently. So very good progress with further work needed to improve printing.

27 May 2013 Update from Peter Vaughan

• Continued with 1132 investigations. After running the printer for most of Sunday the final non-printing column started to print so we don't have any failed solenoids or driver logic faults which is very good news.
• A number of the print columns were still printing the wrong character and it turns out it is always the character just before the one that should be printed on the print drum.
• Monday was spent trying to find out if the wrong character print was caused by the 1130 itself or a logic fault in the 1132 or a solenoid activation problem or even a mechanical problem. Using a digital scope the bit signals coming from the 1130 were checked and these appeared to be correct so the problem was not caused by the 1130. The signals were checked within the 1132 and these also looked correct. I then took some video of the print clutch mechanism and watched it back in slow motion and this clearly showed the clutches for the bad columns were being activated slightly earlier than the other columns even though the same character was supposed to be printed in all columns. The signals that activated the solenoids were then checked and all solenoids were being activated at the same time. So if the solenoids were activating correctly the only conclusion was there was a mechanical cause for the wrong character being printed, something was causing the clutch to be activated slightly early. Checking and aligning the clutch and associated mechanics appears to be very complex requiring the removal of the print mechanism. As the printer os working enough that will have to wait for another day.
• So, apart from a dozen or so mis-printing columns towards the end of the line, all the printing logic in the 1130 and 1132 has been checked and is working so we now have a working 1132 line printer.
• The last remaining part of the system to get working is the disk drive, so that is the next task.
• At the end of Sunday the 1442 card reader started to mis-feed cards in the cornering station again so this will need to be looked at and resolved before loading any more diag cards.

02 June 2013 Update from Peter Vaughan

• After card mis-feeds on the 1442 card reader last week the cornering station was checked and thoroughly cleaned with degreaser. The card feed wheels to the stacker were also cleaned using platen cleaner and restorer which removed the shiny surface. After running several hundred cards through there were no signs of any cards sticking in the cornering station.
• John spent Saturday adjusting the Selectric console printer to try and improve the character printing alignment. He managed to remove the 1130 golfball print head without breaking it and fitted a standard typewriter golfball for the adjustments. During the adjustments John managed to earth one of the contacts with a spanner causing a small spark. Even though the console printer power switch was off, the 1130 was still powered on from when I was working on the 1442 and it looks like some power was still present in the console printer. When we then tried to run the console printer diagnostics the printer would not go ready. The earthing may have damaged some of the logic or power supply so this will need to be investigated.
• One thing I did notice when the console printer was not working was the keyboard did not work either. As this was completely separate from the printer this was strange and a possible clue to the problem. I checked the logic and PSU diagrams at home and discovered that the contact that was earthed had a 48V supply to it. I also discovered that if there was no 48V supply, both the solenoids in the console printer and the 1130 keyboard would not work, which is what was observed. I stopped by the Museum Sunday and checked all the fuses in the mid-pack PSU. The one for the 48V supply had blown. A suitable replacement was found and fitted, and after loading the console diag again the console printer ran and completed the console printer tests successfully. So it looks like the only thing that failed was the fuse, there was no logic damage.

22 June 2013 Update from Peter Vaughan

• John spent last Saturday removing a friction feed mechanism from a donor Selectric I typewriter and fitting it to the console printer. This went almost perfectly apart from one bar used to move the friction feed mechanism away from the platen to feed the paper in. The bar from the donor Selectric was too short. I had also acquired a larger 15 inch Selectric III off ebay and it's friction bar is much longer, possibly too long, but one end is slightly different so will need modification.
• Work has started on the disk drive. Stewart had done most of the maintenance and lubrication last year and had managed to load a disk and move the heads over the disk successfully for a short period. I removed the Plenum air filter to check its airflow and it was fine. Having replaced a seal It was refitted the following week. The filter is missing on the main air intake to the blower. Some suitable filter foam has been ordered off ebay from which a filter will be made.
• The disk packs we have were dismantled and the surfaces checked and cleaned. Of the 6 packs we have, only 2 I would class as very good, although all have signs of heads touching the surface. 2 of the packs are likely unusable due to bad surface damage and oxide loss with 2 others having noticeable oxide surface damage but may still be usable for testing.
• Once the inlet filter is fitted and the Selectric work completed a disk load will be attempted using a hopefully good pack.

30 June 2013 Update from Peter Vaughan

• Console Printer: John attempted to alter and fit the friction feed bar from the 15inch Selectric but the end part cracked. He spoke to John Pether and they concluded the end part from the original bar could be removed and welded to the 15inch bar. John Pether will do the welding during the next 2 weeks.
• 2310 disk drive: A filter was made for the air intake and fitted. The disk was powered on but no air-flow was detected from the blower through the Plenum. The rear filter was removed but this did not change the air flow detected. It looks like the blower is not operating so this will need to be checked.

15 July 2013 Update from Peter Vaughan

• 2310 disk drive: Blower motor checked with meter. Starter and main coils appear to be OK but their is no continuity to the common neutral connection. Motor was removed and taken apart and insulation around the wiring inside the motor and where it enters has gone brittle leaving the wires exposed. The temp sensor on the neutral wire was also open circuit which is why the motor did not start. As the type and temp of the thermostat was unknown, an alternative was found. This was fitted and new wiring fitted between motor windings and existing cable. Also there were some plastic caps over the bearings which have disintegrated at one end but motor still seems to move OK by hand so will need to see if it will cause a problem. Motor now needs to be checked on the 1130.
• Console Printer: John is continuing with the retro-fit of the friction feed mechanism. This has now need fitted successfully and adjusted. There is now a new problem with line-feeding which needs to be resolved.

21 July 2013 Update from Peter Vaughan

• 2310 disk drive: Blower motor was connected outside the main drive and when power on the 1130 was applied it started up and ran normally. Was left running for 1 hour to check it did not get too hot or cut-out and all OK. After re-fitting into drive it was tested again and passed the airflow test, which also proved the Plenum filter was good. Repeated with the inlet filter fitted and airflow still OK. Once the 1130 console printer is working again we will be ready to try the disk diagnostics with one of the disk packs.

06 August 2013 Update from Peter Vaughan

• Console printer: John identified and mostly fixed the line feed issue - as well as changing a component he also needed to swap some springs. The line feed alignment is not quite right, as is some of the character alignments but the console is working enough to be used again. There is still an issue with the front bar that holds the paper to the platen. Neither of the donor Selectric bars fitted so as a temp measure the bar on the SUSE Selectric was 'borrowed' until a modification can be done on one of the donor bars.

17 August 2013 Update from Peter Vaughan

• Console printer: Line feeding needs to be looked at as it is not working reliably, in some cases a double feed is occurring and printing is not always in a straight line. Also the paper has a tendency to move to the right over time so printing slowly gets closer to the left edge. This may be related to the line feed issue.
• 2310 disk drive: Another milestone has been reached... After running the disk init test (0308) and the disk function test (0309) the disk drive has passed all its tests. 2 disk packs have been tested (one found in D-block last year and one from US) with no read, write or seek errors reported. Next activity is to run the DIMAL card deck to produce a bootable pack containing all the test card decks.

30 September 2013 Update from Peter Vaughan

• 2310 disk drive: Successfully created a DIMAL disk pack with a few diagnostics loaded from cards, and I also punched a fast start card for the pack. Successfully booted from DIMAL disk pack using fast start card and also ran the Console printer and CPU test diagnostic (03A1) successfully.
• Console printer: line feed still an issue and printing is occasionally unreadable so adjustments needed.

11 October 2013 Update from Peter Vaughan

• Console printer: John spent some time adjusting the line feed mechanism and character printing alignment but it is proving difficult to adjust due to wear. Currently it performs 2 line feeds rather than one.
• DMS disk packs: Have requested 2 disk packs from Stuttgart Museum containing the DMS (Disk Management System) software but still waiting to hear back.

08 November 2013 Update from Peter Vaughan

• Console printer: John has adjusted the print head as best he can so it can be used again.
• Console printer: Further tests have shown letter alignment issues still exist. John will try to make further adjustments.
• Still waiting for DMS disk pack from Stuttgart Museum.
• We have also been offered a disk pack from the guys who run ibm1130.org

15 November 2013 Update from Peter Vaughan

• Console printer: John has adjusted the printer again and initial testing with the typewriter ball suggests things are looking much better with no obvious character alignment issues. Will try again with the proper 1130 ball to check things are still OK.
• 1132 printer: Found a core store dump to printer program on card. This ran OK and printed the core store on the line printer OK.