deskthority

Hold on to your hats for this one.

A couple months ago, in July, this arrived on my doorstep: This poor 3101 was found inside an abandoned building, on the floor, in several inches of water. As you can see, everything is heavily water damaged and rusted together. Not a single key would depress, as all the springs had rusted together over the years. For some odd reason, I decided it would be a fantastic idea to restore this fancy zinc brick.
Some more pics: The back of the keyboard. The zinc case is heavily oxidized and missing much of the powdercoat. These are the rivets holding the case together. I had to dig with a dental pick for several minutes to even get through the layer of crud covering them, and even after soaking with WD40 and using pliers, they were just too rusted in to unscrew manually. I eventually gave up and just drilled them out. The assembly, after I finally got it open. The steel plates are heavily corroded, and will need to be either replaced or completely de rusted. At least, I thought that was all it needed. The bottom case. The cable is completely rusted in place by the steel retainer. I eventually had to get it out using a dremel and pliers, as no amount of pulling would cause it to even budge. Here is the label, still intact on the side of the barrel plate. As far as I can tell, the date is June 2nd, 1975. The assembly, completely out of the case. The right side was the most heavily damaged, having almost all of the numpad stems rusted away. I did have all the caps for this thing, but almost all of them had rusted bits of stem in the mounts. Many will need to be replaced. These are the actual switches themselves. Shockingly, and to my great satisfaction, the switches themselves are pristine inside. Zero rust or oxidation at all on the beam springs or flyplates. No replacement parts needed for these. The condition of the PCB and controller. While not pictured here, it took me almost an hour of yanking to release the PCB from the controller. The controller is completely dead, with several broken capacitors and oxidized traces. It's the same story with the PCB. Several of the capacitive pads are missing, and the connector on the PCB is very oxidized. I doubt I could salvage this, so I will most likely use a new PCB in combination with a TH controller.

Alright, we're all disassembled now. It's time to start making some improvements.
First things first: We need to do something about the case. Unlike this unit's twin, much of the powdercoat is still left on the case. This will need to be removed at a later date, but for now let's just get all the other grime off. This is after I took the two halves outside and pressure washed them both for several minutes. A drastic improvement. We'll set that aside for now, as there really isn't much else I can do with it other than remove the powdercoat, which I am incapable of doing at this moment. Let's turn our attention to the assembly next. I would obviously need to remove the switches at some point. However, almost every single one of them was rusted into the plate, or just bound there with pure dirt. So, I decided to just throw the entire assembly in a bath of Evaporust. This would hopefully improve the condition of the plates, which were probably the most corroded part of this whole thing. I also hoped that this would help loosen up the keycaps a bit, as they were all very stuck on the stems.
This is it sitting in the bath: I let it sit for about 3-4 hours on each side, turning it over once to get the bottoms of the switches. This offered a drastic improvement in the condition of the assembly. After coming out of the bath, I was able to use pliers to remove the keycaps from the stems. About half of the stems broke off inside the caps, so I will need to remove those later.
Here it is after the keycaps were removed: As you can see, the springs are completely rusted together. I considered trying to remove them and restore them, however it would be much easier to order replacement springs off of AliExpress. These would be a bit heavier than normal springs, but they would have the benefit of actually functioning as springs.
The switches, even after soaking in evaporust for many hours, still were resistant to coming out of the barrel plate. I considered using pliers to try and yank them out from the bottom, but there was too high a risk of damaging the housings this way. Therefore, I had to do this the stupid way. I bashed the top of the switches with a hammer, which removed them from the plate very effectively. This damaged the remaining stems quite a bit, but as I plan on replacing all of them anyway it didn't matter that much. However this caused some pretty major problems as seen here: The barrel plate, especially on the damaged right side, broke along the edges of the barrels. I suspect this would have happened anyway, as there were already hairline cracks all over the top plate. This means that I cannot just outright de rust it and use it, as the break will need to be repaired.
Now let's move on to the switches: I placed all the switch parts into a plastic bin, and again soaked them in evaporust for several hours. This drastically improved the condition of the switches, removing much of the accumulated rust on the housings and stems. I had hoped this would allow me to simply wrench out the metal inserts with pliers, but they were still stuck in there completely.
Before and after: And this is where we are now. I am sorting out the case, which I plan on having sandblasted and powdercoated to the original color. The biggest problems I have right now are the stem inserts and the barrel plate. For the stems, I plan on doing something rather drastic. I plan on using hydrochloric acid to dissolve the stem inserts without harming the actual beam spring component themselves. While I have yet to test this out, as the acid has not arrived yet, I am confident that that will at least partially help the issue. I am also planning on laser cutting new stem inserts, or having them resin printed by the wonderful SneakyRobb. As for the barrel plate, I have searched long and far for a replacement internal assembly. My hope was that Orihalcon, who parts out beamers on eBay, would have a spare assembly he could send me. He has not gotten back to my message yet, but I have learned through a friend that he doesn't have any for the 3101 specifically, although other models could be compatible. I was also considering machining one, but the cost for that would be colossal, and I don't have any money to spare currently. On the matter of keycaps, I have some replacements coming from a Selectric that I can use for damaged keycaps. I will also try the hydrochloric acid on these too, as ABS has a strong resistance to hydrochloric acid and should be undamaged. As for now, all I have to do is wait for parts and try and find a solution for the barrel plate problem. Once I test out the acid, I will update the thread with the results. Any help would be appreciated.

10/11 UPDATE:
So, the needles came finally. Yesterday, I applied the hydrochloric acid to the stem insert of a switch, and to the broken off bit of metal inside one of the keycaps. While it didn't go exactly as I'd hoped, the results were very promising: As you can see, both a part of the stem and some of the broken off stem in the keycap have been damaged or eaten away by the HCI. The issue it that it was not nearly enough to get rid of them completely, or damage them enough to completely remove them. Tonight, I am planning on adding more HCI to each component and seeing if this accelerates the process. I also ran into a bit of a snag with the stem insert. Originally, I picked HCI because ABS plastic had a very high resistance to a 10% concentration, which is what I'm using here. I did this off of the assumption that both the stem itself and the keycap were made out of ABS. However, it has recently come to my attention that beamspring sliders might be made of POM instead of ABS. POM, while still possessing some resistance to HCI, does not have nearly the level of resistance as ABS. A spec sheet I found online indicates warping effects if left in contact. However, I am still not 100% sure about the composition of the stem. Can somebody confirm this, and possibly recommend me a different, more POM safe acid?

Great job. Hopefully this goes as well as your XT restoration did.

kelvinhall05 wrote: 07 Oct 2020, 19:01 Great job. Hopefully this goes as well as your XT restoration did.

wow, Jeck good work on the Memespring, eventually we will be able to find a plate or have one fabricated I have told you you just need to try and get a shop that can bend aluminum or steel to make the plate bends, and then you could drill it out. aluminum would be better on cost and is much easier to drill with a power drill yourself.

kelvinhall05 wrote: 07 Oct 2020, 19:01 Great job. Hopefully this goes as well as your XT restoration did.

lol yes

Excited to see how this turns out!

Before:

After:

JP! wrote: 07 Oct 2020, 20:04 Before:


before.jpg

After:


after.jpg

Slightly confused, isn't the second picture a reproduction?

Another keyboard that went through hell
Congrats on the progress! Can't wait to see the final result.

Things have been recovered from the Titanic looking better than this keyboard. Good luck with the restoration!

Very nice, should reach out to rustoleum for a sponsorship

i think the results will be interesting, we have faith on you!

Don't forget to adjust solution if you try electrolysis. Tap water usually will not conduct enough depending on your locale.

This looks challenging and like great fun. Can't wait to see more. As the slider and the beamspring are intact, the hardest parts to recreate imho are actually in good shape. So good buy!

Quick update, tested out the hydrochloric acid. Was promising, but not enough.

A new barrel plate could be sorted out with an afternoon of CAD and a fabricating shop.

Jesseg wrote: 12 Oct 2020, 07:47 A new barrel plate could be sorted out with an afternoon of CAD and a fabricating shop.

Funny you mention, the wonderful SneakyRobb actually made me a model of a 3101 barrel plate in CAD. But from what I've heard, milling a one off like that out of metal would probably cost more than the whole beamspring. So while that is an option, I would prefer not to go down that route.

headphone_jack wrote: 13 Oct 2020, 02:31

Jesseg wrote: 12 Oct 2020, 07:47 A new barrel plate could be sorted out with an afternoon of CAD and a fabricating shop.

Funny you mention, the wonderful SneakyRobb actually made me a model of a 3101 barrel plate in CAD. But from what I've heard, milling a one off like that out of metal would probably cost more than the whole beamspring. So while that is an option, I would prefer not to go down that route.

I don't think it would be too expensive, it would just be cut out of plate/sheet and then bent, surely you wouldn't reuse what's left of the original one?

That's an archeologic project there. I think the plastic part are the only thing that can be salvaged. At least the keycaps. But all that was metal there is beyond restoration.

Great job!

The original plate is beyond saving, but the original case is in the "so bad it's good" range... that mangled powder coating on zinc(?) looks killer.

Rayndalf wrote: 19 Oct 2020, 11:13 The original plate is beyond saving, but the original case is in the "so bad it's good" range... that mangled powder coating on zinc(?) looks killer.

I was actually kind of fond of the original color. Sadly most of it was dirt and washed off when I cleaned it. It looks much worse now, sadly.

Is mounting the plate too far gone though, or is it better after sanding?

Weezer wrote: 20 Oct 2020, 17:43 Is mounting the plate too far gone though, or is it better after sanding?

theres literally large parts where there is metal missing, it can not be reused

Alum is one of the best kept secrets among mechanics, it will dissolve a steel bolt wrung off deep inside an aluminum engine head. I don't know will it attack the plastic keycaps but it is one to consider.

If you want to machine the switch plate you could use a small hss bit to drill the pilot hole and then a greenlee knockout punch to do the hole itself. since all the switches are the same size you'll only need one punch.

If you drill the pilot holes first. then get the plate bent, you can knockout the holes afterwards. that will prevent stretching of the large switch holes. I'd make up a circular washer to drop inside the hole in order to act as a guide to drill the plate underneath.

Weezer wrote: 20 Oct 2020, 17:43 Is mounting the plate too far gone though, or is it better after sanding?

I could probably make it useable, but the only thing that removes the rust is wire wheel brushing with a drill. Even then, it's so pockmarked and pitted that it would probably break with any serious de rusting techniques. I might see if I can get it to hold the switches for the acid bath I plan on giving them, but beyond that I have no plans to use it on the final product.

Enjoy seeing the process of turning trash to treasure, update bit fast, man.

Update: After a long period of being completely stonewalled by these sliders, I finally am making further progress. The biggest leap forward is that a fellow enthusiast is willing to mill me brand new internal plates, vaporblast and paint the case, and try and get the PCB cleaned and functional. Coincidentally, he also happens to be in the process of creating beamspring reproductions, with brand new sliders readily available. He is going to test out various chemicals on the metal stem to try and get it to dissolve properly. If all else fails, I will probably just send him the rest of the switches to put new sliders on, using the beams from the old ones. It's looking like this project is getting pretty close to completion!

It could always be worse... someone is going to rebuild that enigma machine they found in a U boat wreck
https://phys.org/news/2020-12-divers-na ... chine.html

Nice. So SneakyRobb helped you out? He's a mad lad.

Weezer wrote: 07 Dec 2020, 20:21 Nice. So SneakyRobb helped you out? He's a mad lad.

Nope, Sneaky doesn't have milling capabilities. Different guy!