German Model F AT Resurrection Journey

[Aruberon]

Good afternoon DT,

Its no secret that an IBM Model F AT these days are quite hard to find.
Ive been on and off looking for quite a while now so when two weeks ago, when a listing for an "Old IBM keyboard" shows up on a local auction site being an F AT, a German one to boot which is allegedly one of the more rare ones (source needed), it was gonna be mine.

50a5f0404b7cb41f18877e390ab4.jpg (332.53 KiB) Viewed 51179 times

72bb289f42559ccda27ef495fc91.jpg (287.36 KiB) Viewed 51179 times

Based on the initial pictures, not too bad, however IBMs dont tend to naturally exist in this country, and the ones that are labled as "Old IBM keyboard" tend to be quite sad, Ive rescued a couple of rusted out ones prior, but this one as it turns out will be quite a bit more sad than the rest.

photo_2022-11-08_14-38-54 (2).jpg (146.74 KiB) Viewed 51179 times

photo_2022-11-08_14-38-54.jpg (200.05 KiB) Viewed 51179 times

photo_2022-11-08_14-38-56.jpg (157.19 KiB) Viewed 51179 times

photo_2022-11-08_14-38-57.jpg (223.46 KiB) Viewed 51179 times

Inside we discover authentic backplate rust slime, some more rust and actually readable lables unlike the one sitting on the bottom case including the most important part, its birthdate: 13 December, 1985.

photo_2022-11-08_14-38-54 (3).jpg (265.06 KiB) Viewed 51179 times

photo_2022-11-08_14-38-55.jpg (205.88 KiB) Viewed 51179 times

photo_2022-11-08_14-38-54 (5).jpg (177.91 KiB) Viewed 51179 times

Controller board and locklight board are dirty, but look fairly ok, cable looks ok, both outside and various inside ones.
However what may or may not make or break this project awaits us sandwiched between the barrel and bottom plates.

photo_2022-11-08_14-38-53.jpg (201.27 KiB) Viewed 51179 times

A reasonable amount of violence later considering the state this thing is in, and we are in.
Inside of the backplate is fine, backside of the PCB is "fine", front of the PCB, lets not talk about it, could be a case of worse than it looks but its to be determined as of now.

photo_2022-11-08_14-38-52 (2).jpg (207.28 KiB) Viewed 51179 times

photo_2022-11-08_14-38-52.jpg (208.75 KiB) Viewed 51179 times

photo_2022-11-08_14-38-53 (2).jpg (224.99 KiB) Viewed 51179 times

photo_2022-11-08_14-38-53 (3).jpg (242.6 KiB) Viewed 51179 times

Back of the barrel plate has seen better days, but its to be expected, and the foam probably vacated quite a while ago.

photo_2022-11-08_14-38-52 (3).jpg (196.27 KiB) Viewed 51179 times

photo_2022-11-08_14-38-51.jpg (255.62 KiB) Viewed 51179 times

Flippers and barrels are atleast in ok shape, some rust on a few springs but thats easily reversible.

So thats where we stand for now, its gonna be a bit of a journey and depending on the PCB this project might end prematurely, but as its status is currently unknown itll be discovered later, updates to be posted as I go on.

[JP!]

Woof! That is a rough one. I'm all for tough projects but I think this one is a goner. As a consolation I am surprised that the springs seem to be in good condition. A simple soak in some Evaporust would clean these up. These German F AT's are rare as hens teeth in my experience and I am fortunate enough to own one. If it were me I would go through the trouble of importing one from the US and then transfer all of the German specific pieces over to the donor keyboard. This route will not be inexpensive but you will be left with a superior keyboard along with some extra parts which can then be resold to help offset your costs. You should end up with an extra case, case feet, cable, barrels, flippers and springs. These would all be good parts you could sell to help others with their projects.

[jsheradin]

Nice project! I've fixed up similarly rusty boards although not quite as bad as that. I can definitely recommend Evaporust for the metal bits. Even if the PCB has some dead traces it's still likely repairable with bodge wires. Depending on the extent of the damage you might need to change out the controller for a QMK based xwhatsit. The self adjust algorithms are able to handle far from ideal PCBs that don't play nice with stock controllers.

Looking forward to updates!

[Muirium]

I salute you for this undertaking. That board’s innards are a bill of goods, all right. Do Greenock proud!

[JP!]

Here are the innards of mine as a bit of eye bleach. It is currently in a box awaiting for me to continue with restoration. The ink marks on my keys washed off pretty easy so this board should look mint once it's finished. It seems these early UK produced boards were painted and suffer from the paint flaking off. I was able able to easily take about half of the paint off as it chipped off fairly easily while the rest of it does not want to come off. Sanding in this case seems overkill and I don't want to apply any unsafe chemicals to strip the remaining paint off so I need to research this topic a bit more before proceeding. At least the paint on the bottom half is solid so I will leave that as-is with just basic cleaning. I plan to have the top shell professionally repainted as well as the barrel plate. Unfortunately I could not preserve the neat looking white foam which seems to be a hallmark of these UK boards.

Model F AT

[del20nd]

Do you happen to have access to a local machine shop? I think in this case, given the rarity of the board and the condition of that barrel plate, I'd consider paying somebody to sand blast it. That or spend a heck of allot of time applying and reapplying naval jelly.

Rust converter probably isn't going to cut it here. It's gone all chunky...

[John Doe]

Also feel bit surprised at the springs are at a relative better condition compared to rest parts, that's a positive signal.

Worst scenario I can think of is to replace OG controller with Xwhatsit as new brain as you may find some capacitors dead, and Evaporust will save alot of time to clean plates, then ultrasonic or just hands clean caps, barrels and cases?

Good luck

[Aruberon]

Do you happen to have access to a local machine shop? I think in this case, given the rarity of the board and the condition of that barrel plate, I'd consider paying somebody to sand blast it. That or spend a heck of allot of time applying and reapplying naval jelly.

Rust converter probably isn't going to cut it here. It's gone all chunky...

Wouldnt sandblasting the barrel plate potentially risk the plastic stabilizer rivets?
Speaking of, they all appear to be intact and in quite good condition.

Considering how thick the rust is, depending on what type of plastic those rivets are made from, I can go the phosphoric acid route instead of evapo rust for the bottom plate and barrel plate.

Either way there will be a decent amount of prep with a metal bristle brush before the chemical bath.

[guidemetothelight]

What a shame ! You´ve got the holy grail Model F (at least for me). I hope you can restore it to its former glory

[Aruberon]

Yesterday post-disassembly ended with all the keycaps, springs, flippers, barrels, screws and nuts in a dishsoap bath overnight to loosen up some of the goo and greese.

This morning however I was greeted with this, and the picture is not out of focus, its just very murky.

photo_2022-11-09_11-56-50.jpg (102.51 KiB) Viewed 50894 times

Tasty to say the least, after the cleanup, all the springs were separated from their respective flippers, and as per initial inspection yesterday, most springs are in pretty good condition with just a tiny bit of rust at the bottom, they came off with little resistance, with the exception of four springs which snapped clean as soon any kind of twisting force was applied, not too bad, fairly easy to replace.

Post separation the springs, stabs, nuts and bolts together with the 4 flippers being currently merged with their broken spring went into a vinegar bath, they probably wont need anything stronger as they are overall very lightly corroded.

photo_2022-11-09_11-56-50 (2).jpg (46.2 KiB) Viewed 50894 times

All the plastic went back into another green soap bath, still a bit dirty and grimy, but no significant mark, discolouration or anything of the sort, they should clean up nicely like most old IBM plastic.

[fohat]

Wouldnt sandblasting the barrel plate potentially risk the plastic stabilizer rivets?

The stabilizer tabs can be removed with a small flat head screwdriver and replaced with a larger one.

Use a small screwdriver (smaller than the size of the hole) to push them gently but firmly out from behind and most of the plastic "skirt" or "flange" should survive.

Replace them with a large flat head screwdriver in the "mouth" of the tab and the shaft of the screwdriver parallel to the surface of the plate. That will take a strong push, and don't build up too much paint or the paint thickness will reduce the inside of the hole.

[Aruberon]

Wouldnt sandblasting the barrel plate potentially risk the plastic stabilizer rivets?

The stabilizer tabs can be removed with a small flat head screwdriver and replaced with a larger one.

Use a small screwdriver (smaller than the size of the hole) to push them gently but firmly out from behind and most of the plastic "skirt" or "flange" should survive.

Replace them with a large flat head screwdriver in the "mouth" of the tab and the shaft of the screwdriver parallel to the surface of the plate. That will take a strong push, and don't build up too much paint or the paint thickness will reduce the inside of the hole.

Oh wow, based on how the back of the rivets looked like, with overflowing plastic way past the slot it seemed like a permanent fixture.

Screwdriver was applied to the back however and they all popped out relatively easily despite the amount of rust surrounding them on both sides, it did shed the thin overflow on the bottom side but they still fit nice and tight in their original slot without.

[Aruberon]

Some relentless scrubbing has taken place on the bottom and barrel plate since yesterday, especially after fohats advice about the removable stab tabs and honestly, there is surprisingly decent metal underneath, most of the chunkier bits simply peeled off.

photo_2022-11-10_14-13-00.jpg (191.77 KiB) Viewed 50785 times

photo_2022-11-10_14-13-00 (2).jpg (136.24 KiB) Viewed 50785 times

There is still a few pretty bad spots, and I intend to scrub it a bit further before its chemical bath, but thankfully mainly due to IBMs dummy thick steel plates, these should clean up nicely.

Worst place of course being where the authentic rust goo resided prior, but being on the outside of the bottom plate, it shouldnt effect the plates despite some pretty bad pitting.

photo_2022-11-10_14-13-00 (3).jpg (154.63 KiB) Viewed 50785 times

Alas none of the lables survived, but thats alright considering the birthday is documented in this thread.

[JP!]

This is coming right along. A filler primer or glazing putty over the pitted areas should hide the imperfections.

[del20nd]

Wow. That looks great for just some elbow grease! That or I'm too jaded by my project car...

Seems like this project is well on its way though. I'm looking forward to more updates!

[Aruberon]

Its case inspection day and I suppose one of the biggest surprises yet, despite this keyboard potentially spending more time as a submarine than a keyboard throughout its life time, the case is pretty spotless.

What I initially thought was black scratchmarks on the bottom case came of super easy with some scrubbing, similar story for the top case, with its paint having next to no chips anywhere, bonus being the german lock light sticker having absolutely 0 damage.

photo_2022-11-11_12-27-14 (5).jpg (200.19 KiB) Viewed 50672 times

photo_2022-11-11_12-27-14 (3).jpg (112.09 KiB) Viewed 50672 times

The label on the bottom disintigrated into dust the second I touched it, but it was already faded beyond repair so it matters little.
More importantly, all four feet are present, with minimum damage and still glued to the case, I will leave them as they are.
The square IBM PC AT badge might also be junk, but there are some good reproductions of those floating about so worst case scenario itll be fine.


Inside was quite discoloured brown due to prolonged contact with rust, ive begun cleaning most of it up, there are still plenty spots left as seen on the pictures but overall, no damage, every clip on the front top and bottom are all intact

photo_2022-11-11_12-27-14 (4).jpg (196.75 KiB) Viewed 50672 times

photo_2022-11-11_12-27-14 (2).jpg (153.34 KiB) Viewed 50672 times

And lastly, the feet, some minor scratches and one of them is a bit discoloured, but the springs have 0 rust, the plastic 0 cracks and the ears are firmly attached still.

photo_2022-11-11_12-27-14.jpg (205.35 KiB) Viewed 50672 times

[del20nd]

That looks great! Especially lucky to still have those ears.

Is the AT case painted like some of the other F's or molded like the M?

[Aruberon]

That looks great! Especially lucky to still have those ears.

Is the AT case painted like some of the other F's or molded like the M?

Good question actually, its been a while since I even bothered to look at my XTs close up, but comparing the two, it definately looks like a mold rather than paint, just retaining the same texture.

Do all most/all ATs come molded rather than painted, and if so does it use the less shatter prone M plastic vs the F XT plastic?
I suppose the latter would be nicer, Ive heard horror stories regarding 3178s for example and their bottom case dying in transit, mine somehow made it unscuffed.

[Muirium]

My all American AT (Property State of Texas Department of Human Resources) is M-style plastic through and through. Smooth texture, unlike any XT that’s passed through my hands. I like it, but yes it’s different.

Speaking of shipping snafus: my plastic AT beat the shit out of a Model M2 while in airline luggage. They’re plenty tough enough to look after themselves.

[del20nd]

My all American AT (Property State of Texas Department of Human Resources) is M-style plastic through and through. Smooth texture, unlike any XT that’s passed through my hands. I like it, but yes it’s different.

Speaking of shipping snafus: my plastic AT beat the shit out of a Model M2 while in airline luggage. They’re plenty tough enough to look after themselves.

To be fair, I don't think too many keyboards would lose in a physical battle against the M2. Even some of the better built rubber dome boards have a shot.

[JP!]

I believe all of the US made AT's that have passed through my hands 5+ have had the textured plastic cases while my UK made early production date QWERTZ AT has textured paint. With the paint removed the plastic case is completely smooth.

[inozenz]

I believe all of the US made AT's that have passed through my hands 5+ have had the textured plastic cases while my UK made early production date QWERTZ AT has textured paint. With the paint removed the plastic case is completely smooth.

I can confirm this.

[Aruberon]

Interesting, considering mine is a UK made qwertz, it appears to be a textured plastic case.

IMG_20221111_160140159_HDR.jpg (479.44 KiB) Viewed 50489 times

This transition from smooth plastic to a texture in that specific curve exists on both dividers between f row and numpad, both sides both up and down.

It neither looks nor feels like paint, I guess it's a production date thing rather than a production location case.

[Aruberon]

I was gone over the weekend, and yet to procure either evapo rust or phosphoric acid, but I wanted to start the process before I left, so with nothing but basic household vinegar at home and a 52l plastic cointained used for prior projects (the bottom fits a bottom and top plate more or less perfectly) I tossed them in as they were in the last pictures, added 5l of vinegar and went away for 2 days, when I came home they both got a 5 minute scrub and a wash and this is the results:

photo_2022-11-14_11-21-11.jpg (136.05 KiB) Viewed 50371 times

photo_2022-11-14_11-21-11 (2).jpg (128 KiB) Viewed 50371 times

That right there is about 95% of all the rust gone, with all the crusty bits just peeling off, the pitting as well is shallow enough so putty probably wont be needed, but an extra layer or two of primer during painting, which will make reassembly harder but Ive fought that battle on my very first XT restoration, its not pretty but it works.

Both plates have darkened significantly but thats to be expected, I do enjoy the outside of the bottom plate trying to resist with its unscrubbed heat treated steel, but itll go dark soon as well.

These will go back in their bath for another day or two and then they should both be done.

[JP!]

One beef I have with acid is that it indiscriminately attacks the good metal but probably not really an issue here. One thing to be mindful of is that once you pull this out of the acid bath is the metal could develop flash rust as it dries (factors such as humidity can exaggerate this) so you may need some sort of rust inhibitor. My first attempt to remove rust using citric acid removed the rust but after a rinse and a quick dry my plate eventually turned rustier than when I had started. Sorry to shill Evaporust once more but a quick 30 minute soak followed by a final wash/rinse cycle will ensure the metal does not start to rust again before you have a chance to do primer/paint.

[ifohancroft]

Loving the project so far!

[Aruberon]

So Ive cleaned off the PCB and Ive hit a bit of a roadblock, perhaps someone here has experience troubleshooting an XT/AT PCB?

There is life there, which is good, problem is just plugging it in unsquished and ungrounded causes it to freak out as these boards tend to do, making it impossible to test individual pads and to confirm if the controller itself is damaged, and if initial testing means reassembling the entire board every time I wanna test each trace, thats gonna be a lot of fun and time to invest.

There is definately gonna have to be some form of trace repair as a decent chunk of them are damaged, to be expected, how would you go about that on a board like this, sitting tight inbetween a steel bottom plate without padding and with plastic barrels blocking the top part?

I suppose via extreme sandwiching violence it would be possible to add a very thin layer of foam at the bottom as well incase wires would have to be run underneath, does that also mean it would in theory be possible to handwire the entire board incase the traces are too cooked?

Whatever does end up happening to the PCB im committed to restoring this thing cosmetically, and if the PCB is too far gone I will probably go the route of blasphemy as JP suggested and do a full gut transplant, which will be a very pricy transplant and will most likely take a while to source, but thats long term.

[Aruberon]

One beef I have with acid is that it indiscriminately attacks the good metal but probably not really an issue here. One thing to be mindful of is that once you pull this out of the acid bath is the metal could develop flash rust as it dries (factors such as humidity can exaggerate this) so you may need some sort of rust inhibitor. My first attempt to remove rust using citric acid removed the rust but after a rinse and a quick dry my plate eventually turned rustier than when I had started. Sorry to shill Evaporust once more but a quick 30 minute soak followed by a final wash/rinse cycle will ensure the metal does not start to rust again before you have a chance to do primer/paint.

Ive had surprisingly good results using an assortment of acids to derust my various projects, usually I dont run into the issue of flash rust as I do a final rinse with something basic, could be just water with a decent bit of soap or baking powder to neutralise the surface, it some flash rush does appear, a quick wipedown with some sandpaper will take care of that.

I would love to try some evapo rust but the problem is sourcing it in this country, Ive found one online shop carrying it so far and its hilariously overpriced compared to the US offerings, and considering id need about 5 l to do a full soak.

What I am considering is getting a smaller bottle of about 500 ml as a final rinse with once the acid bath is done for finishing touches.

[JP!]

the route of blasphemy as JP suggested

I have no idea on the PCB repair so perhaps someone else can better advise. In most cases I doubt this part is truly unique except in my case where my early revision uses a different PCB color. Using a vintage replacement is not as impure as using a modern reproduction and you could perhaps still use the original controller if you wanted. Speaking of originality, the backplate without the holes would not have been painted but I believe plated in some manner to prevent corrosion, perhaps using the good chemicals that they aren't allowed to use anymore. You could look into having it zinc plated and dipped into a yellow chromate solution to get that bling yellow/bronze iridescent finish.

Yeah I am thinking now that I might not have neutralized things properly when I did my final rinse/wash. I think Evaporust has a gel product which is good for items which would be difficult to fully soak but in either case might not be worth the expense to source in your country. If you really want to take derusting to the next level you could always build an electrolysis tank. I guess in the end perhaps a bit of surface rust isn't the end of the world. You could also use a self-etching primer which contains an acid which etches to the metal to provide a strong bond in preparation for a final top coat.

[fohat]

a board like this, sitting tight in between a steel bottom plate without padding

The Model Fs that I have disassembled had a clear mylar sheet with "keyhole" shaped slots matching the holes that "slide fit" the PCB into the curved metal plate.