Relatively quiet until Friday 9/19 but a bit of progress on several fronts

1442 CARD READER/PUNCH RESTORATION

The National Museum of Computing in Milton Keynes, UK has a restored 1130 system and I have been the beneficiary of all the experience they amassed during the effort. Peter Vaughan has been sending me detailed descriptions of what they discovered, how they repaired or adjusted it, and other tips for a successful restoration. Yesterday, Peter send me a long email with advice about the 1442 after reading my posts.

His reader also had most of the rotating steel wheels semi-frozen due to congealed grease and used the same persistent gentle working of new lubricants that I did to get the turning. He pointed me at two that I may not have seen, underneath the punch unit out of sight.

My reader pressure roller, a metal roller that is attached to a rotating lever assembly under strong spring pressure holding the roller against a rough rimmed large wheel underneath the card path. The two wheels pinch the card between them, while the rough rim provides the leverage to keep the card moving forward. I discovered that this wheel, even when the old lubricant is replaced, grinds and rolls unevenly. I believe a bearing inside is broken.

Metal roller wheel on this lever arm, riding on rough wheel surface below, but bearings are shot

I can see some signs of wear on the roller wheel surface that shows this failure started while the reader was still in service back in the 1980s, as the wheel bearings progressively got worse. The wheel is not designed to be removed from the lever assembly, unlike most of the metal rollers which have a circlip holding them on an axle.

I expect I may have to grind the end of the axle to release this damaged wheel, find a compatible substitute from a source like Graingers, and invent a way to secure the wheel on the axle when I replace it.

The other area where I was seeing problems was the cornering station, a spot where the card enters while moving from right to left through the punch station, stops, and then is propelled from back to front in a ninety degree change of motion. The card sitting in the cornering station would not start moving forward into the stacker area, but when I partially opened the top cover to release some drag, it worked fine.

Cornering station where card arrives (from top in this picture), stops, then accelerates to stacker (right in pic)

Metal surface needs to be well polished and under surface of plastic must be clean and slick

Peter noted that these areas with a clear plastic top cover over a polished metal bottom plate are highly dependent on good polishing and cleanliness otherwise drag increases until the card motion is impaired. I have to use a metal polish to get the bottom plates very shiny and slick, as well as use my IPA or similar cleaner on the clear plastic top covers so their under-surfaces are also slick.

They came up with a very clever modification that helps with testing. Normally, the NPRO function will only operate when the card hopper is empty, but the other functions that move cards will stop on many check conditions that we are trying to debug and fix. By putting a switch in the hopper empty circuit, we can do an NPRO while cards are sitting in the hopper.

This is helpful because the NPRO will pick a new card on each cycle as it moves the existing cards in the mechanism one station forward in the same cycle. If it works right, I can empty a large pile of cards one by one through the machine at a decent speed by pushing NPRO while the 'TNMoC' switch is activated.

The very detailed and helpful email from Peter also detailed all the grease fittings, which I will double check to be sure I forced new grease into all of them. Without new grease, those will not operate correctly.

KEYPUNCH INTERFACE DESIGN

I met with one of the other designers and discussed some alternatives related to the new interface, Stan Paddock at CHM. Stan found that a standard push on lug he had in his supply bin was a perfect fit on the SMS socket pins for the keypunch. This complemented my discovery two days ago that the Molex male pin for the .062" series connectors was an excellent fit for the relay panel holes. Now we have sources for the hardware to make up the cables and connect to the keypunch logic gate.

Push on connector that works on SMS socket pin connections

Backside of the standard connector

RECEIVED CE CARTRIDGE FOR DISK DRIVE ALIGNMENT

My latest eBay purchase, a CE cartridge for use with the 1130 internal disk drive, arrived tonight and I did an initial inspection. I see some marks on the surface on the path traced at specific cylinders - one dark track on the top and two on the bottom of the platter, separated by about 3/8" of radius from the center of the platter. They appear only at those three locations and to a first quick look they seem constant all around the platter.

My CE Cartridge arrived

Platter surface, hard to see due to reflections

Another view, some dust that is easily cleaned off surface

Trying to capture the darker tracks

The other surface, same constraints but twin lines are visible near upper third of the lower white area

Top picture attempt with different conditions

When I get good surface picture, it isn't far enough back to show marks

Reflections and blurry too

Final view

Until I can look much more closely at the dark traces and see whether they are raised, grooved and importantly if there is a hard particle embedded somewhere around the track, I can't assess whether the pack will be okay to use for its alignment purpose. If these are further out than cylinder 105, I need never put the heads over them. For alignment purposes, cylinders 95, 100 and 105 are used, while all the rest are superfluous but available for temporary use and testing

I will use my 99% isopropyl alcohol and lint free Kim-wipes to clean the disk platter, then inspect as well as I can. I may have to open the cartridge and check it while apart, it depends on what I can observe and clean using just the opened flap.