Yet another beamspring refurb (5251)

[AJM]

Here I have a few pictures of my 5251 restorations from a few months ago. It's not a proper thorough restorations report, because - being my first beamspring keyboard - I wanted to concentrate on not making too many mistakes instead of making sure everything is recorded properly. But maybe it's still of a certain interest.

The purchase
I got it on ebay. That keyboard had already been "sold" shorty before for about $1450, but that must have been somehow cancelled, because it appeared again with the same pictures from the same seller. That made me a bit wary, but in the end I bid anyhow and got it for $1250 (which with shipping and import to Germany became $1600).
That's a lot of money. But after seeing people paying sometimes over $600 just for a new generic 60 % aluminium case from China (which isn't really that much different from the ones of dozens of group buys that came before), a beamspring keyboard with all its history and the engineering behind it doesn't look that much overpriced to me.
Especially because I got really lucky concerning its ...

Condition

5251-01.jpg (1.01 MiB) Viewed 14034 times

That's the condition, I got it in. Pretty nice, although the case was quite a bit dirtier than it looks in this picture.
There was a little chip in the plastic part of the case - down on one side, but it was glued back, so nothing's missing. All 4 feet were also there and in decent condition.
Only missing part was the original top row <8> keycap, but it came with a replacement that looks nearly perfect, apart from the legends that have yellowed in contrast to the originals.
The main reason, why I still had to take it apart completely, was that the switches were very very very loose. When I pressed one key, all the surrounding key were bowing down towards it, which looked quite disturbing.

The work

Cleaning of the plastic case was done in the usual way: At first water with dish soap and a sponge cloth, for the tougher dirt a microfibre cloth and 70 % alcohol.
A lot harder was the metal base of the case, which had red streaks on it, as if the paint of another metal part had rubbed off on it. Nearly nothing could remove that properly. With metal polish I managed to remove most of it in the end.

Now a few pictures of the freshly opened case. In general I would say "Yuck!" to such a sight, but here it was more a deep satisfaction to find it in untampered state – like an archaeologist finding an ancient tomb unopened and intact.

5251-02.jpg (1.08 MiB) Viewed 14034 times

5251-03.jpg (1.41 MiB) Viewed 14034 times

5251-04.jpg (711.05 KiB) Viewed 14034 times

The contamination shield ....

5251-05.jpg (1.74 MiB) Viewed 14034 times

.... having done a brilliant job:

5251-06.jpg (1.19 MiB) Viewed 14034 times

The inside – looking fine, even the visible part of the foam looks surprisingly ok, but under the switches it was of course all compressed to a thickness of nearly zero.

5251-07.jpg (1.11 MiB) Viewed 14034 times

But this completely compressed foam was truly diabolic stuff. Removing that was the most disgusting and annoying cleaning work I've done so far. It was like tar that stuck to everything - sullying the previously perfect outside of that part.

After the cleaning – with new seals, make from thick rubbery sticky tape. (I later cut them back to only a thin strip along the outside edge to make it compress more.)

5251-08.jpg (412.07 KiB) Viewed 14034 times

To find new foam in the thickness I need is always a challenge. Going by the amount of looseness of the switches I reckoned I would need something thicker/stiffer and in the end used a piece that I found in the workshop of my late father. One side is sticky and the other one has "pimples". No idea what it's called or what it's for.

5251-09.jpg (928.48 KiB) Viewed 14034 times

5251-10.jpg (1.14 MiB) Viewed 14034 times

On the sticky side I replaced the protective film (which obviously didn't stick very well) with clear film (meant for protecting the paintjob of a car). This side should be the one, where the switches would be pressed against.
The idea behind it:
- A bit more pressure against the switches. Through the stiffness of the clear film a bit more foam has to be compressed compared to just the small edges of the switches sinking into the foam. (As I found out later – probably not necessary)
- Instead of sinking into the foam, the edges of the switches can slide on the clear film, so the keys/switches can later be rotated slightly, should they look crooked. (I think that worked quite well.)

5251-11.jpg (861.98 KiB) Viewed 14034 times

But as it turned out, the foam was too thick/stiff. After assembling, the switches of the middle row were often very scratchy. That was because the two little plastic hooks of the switches, that hold them (slightly) in the upper plate – together with the pressure of the back plate weren't strong enough to press the switch – against the foam – completely out of the upper plate (the picture, which I made later, shows a correctly positioned switch with "free" hooks, though).

5251-12.jpg (49.45 KiB) Viewed 14034 times

If not completely pushed through, the little hooks were bent inwards and pressed against the spring, which resulted in an awful scratchy key feel (some keys were even completely stuck). The switches of the upper and lower rows were ok, because they are close the screws that hold all together. But as sturdy as the beamspring switch assembly looks, it still bends and I couldn't get enough pressure on the switches/foam in the middle.
If I do it again, I might actually think about a bolt mod – trilling at least two holes for bolts (below the return switch and between A and CTRL).
But this time I decided to make the foam thinner. Finding a new thinner foam and punching all the holes again didn't seem appealing so I decided to cut horizontally through the foam with a carpet knife. The result wasn't pretty, but it worked surprisingly better than it sounds.

Luckily that did the trick and finally everything came together nicely.

5251-13.jpg (741.52 KiB) Viewed 14034 times

5251-14.jpg (695.67 KiB) Viewed 14034 times

5251-15.jpg (495.03 KiB) Viewed 14034 times

My last mistake was to not pay enough attention, when I put the xwhatsit controller on. The plug of the xwhatsit controller isn't exactly the same size as the connector of the keyboard pcb, so it can be attached a bit to the right or left. You have to make sure yourself that the contacts line up correctly.
At first, after attaching it wrong, it was not possible to calibrate the keyboard, because a lot of rows where practically shorted.

Tips & Tricks:

- If a switch does not stay in the top plate - only held by it's own two hooks – the foam is probably too thick.
- That idea is not from me, but from this forum and is worthy for being mentioned once more. It saved me a lot of trouble: When you want to disassemble a beamspring switch, but can't get the metal stem out, heat it up with a lighter.
- The only things that were a bit rusty, were the stabilizer wires. I didn't want to tread them with vinegar or similar stuff, because that would probably have removed the black colour as well. Black paint, applied by me, would certainly have rubbed off – making a little mess over time.
Instead I simply rubbed the wires with try toilet paper. That removed most of the rust while still retaining most of the original black surface.

Finally - the absurd size of the keyboard compared to a 27" monitor.

5251-16.jpg (1.15 MiB) Viewed 14034 times

(Although the tactile feel of a beamspring switch is really sublime and the sound is a bit quieter and a lot more refined than the one of a Model F, and all switches work reliably, I'm not using it as a daily driver. Both at work and at home, this job is done quietly by keyboards with silent red switches, these days.)

[darkcruix]

How beautiful is that?

The foam is quite thick - the one I used is only half of it I would assume. Still my keys are stiff - maybe you can make a "wobble video" when you have time. The final picture let's my heart jump - wonderful!

My daily driver is a 3278 Beam Spring and I agree - the sound can be annoying to some people. Compared to a Model F it is actually whispering symphony in my ears

[matt3o]

I've featured your post in homepage, nice job on your refurb!

[AJM]

@matt3o: Oh, thank you.

@darkcruiz: Thanks. There is no wobble (at the moment). After cutting around on the foam to make it thinner (by about a good third), I hit the sweet spot: Thin enough to enable a correct assembly, and thick enough to hold the switches properly.

[//gainsborough]

Man, what a fun read! Great job on the restoration. Your pictures look really good too!

[Lbibass]

Wow, this looks incredible. Nice job!

[Hypersphere]

@AJM: I am humbled by your diligence, ingenuity, and craftsmanship. Excellent job!

Is it not difficult to give up such refinement and elegance for something as lacking in character as mx silent reds?

Does your VIBE have silent reds? If so, how do you like that keyboard?

I like the look of the VIBE, but I find it difficult to type on keyboards with flat profiles.

I also don't like the Vortex way of achieving programmabilty, which is only partial, and it is not possible to remap the Right Control key using its built-in remapping. Fortunately, it is possible to use remapping software or a Hasu USB-USB converter to do this.

[AJM]

Wow, thank you all again for your kind words.

@Hypersphere:
I think I have to start at the beginning: I'm not a fan of flat profiles, also. And finding out, that there are so many different keycaps with different profiles and especially sculpted ones like SA and MT3, was the incentive for me to try out MX (compatible) switches. (Because SA keysets are usually not very ISO friendly I started with MT3, fortunately, and I'm pretty sure I'll now just stick with them. I love them.)
The reason to start with a Vortex keyboard was for one simply the availability and secondly the layout of the ViBE, which comes close to an XT layout. I'm not really fond of low-profile cases, but that couldn't be helped (and I've already thought a lot about putting the ViBE somehow in an IBM XT case ....).
Being familiar only with tactile "switches" in the form of rubberdomes and buckling springs (with the faint memory of a linear Nixdorf keyboard, which I had to use in my first job sometimes, on which I had to hammer the keys multiple times to get a reaction), I would never have thought to buy a keyboard with linear switches. So my first ViBE actually had MX brown switches! I was quite shocked, that this scratchy feeling is being called tactility.
But still - it's not so bad, if one types swiftly and I was very surprised that, when I tested my speed for the first time, I was immediately a lot faster than on my Model F XTs and typing was just so easy with far less typos.
I'm also a very shy and quiet person, so if I type on my Model Fs or the 5251, it gets louder the faster you type and I then do that with gritted teeth - wondering, if I'm annoying my neighbours, which is very uncomfortable and probably unconsciously slows me down again. The (not silent) MX browns were also very relaxing in that regard.
Then I got me the 63 switch tester and had to find out, that I liked none of the so-called tactile switches. I liked Tealos and Gateron silent reds, but - again because of availability - got me then a Vortex Race 3 and later ViBE with MX silent reds. They're a tiny bit scratchy compared with Gaterons, but let's call that "character".
I absolutely agree that the programmability is very crude, but in practise it's surprising how much you can achieve with it. Somehow it just works, although - in theory - you think it's actually not logical. Only one small example: I use the Race 3 in Mac mode on my Windows PC, because I prefer to have the media functions on the standard layer of the F-keys. This would mean that I have no AltGr key. But when I "reprogram" (while being in Windows mode) the key "AltGr" with "AltGr" (sic), that key then also works in Mac mode as AltGr.
Concerning the Right Control key I don't notice a problem: I can reprogram Right Control with any other function and reprogram any key as Right Control.

[Hypersphere]

@AJM: Thanks for the very complete answer!

Now I regret that I never got an MT3 set. Apparently, they are only available through Massdrop, and so I suppose that new drops for this set would only occur rarely.

Regarding the RCtrl on your VIBE, are you saying that you can remap it to any other key and remap any other key as RCtrl? My problem is that I want to remap RCtrl as Fn. The only way I found to do that on a Vortex board is not to use its built-in programming, but to use an external remapping software or an external Hasu USB-USB converter.

Among Cherry mx or its clones, I found two that I sort of like -- Gateron Yellow (linear; intermediate weigth between Cherry red and black) and Zilent v1 67g. I have tested some individual Zilent v2 switches in various weights -- they are more tactile than v1, but I am not sure if I like them better.

My best accuracy and speed are with Topre 45g switches in my RF and HHKB boards.

[tron]

Looks like it's in great condition

You may want to consider making a dust cover for it now that the contamination shield is removed. I had a seamstress make a couple prototypes including the one below. When my schedule clears up I'll have another one made like a tight fitting shell (without the spandex bottom).

Cover gallery:

https://imgur.com/a/AEs7Bz2

[mcmaxmcmc]

Very nice! Though please retrobright the 8 keycap so that it'll blend with the rest. Otherwise, perfect! Nice job!

[AJM]

@Hypersphere:
Yeah, I'm happy, that I found MT3 keycaps to fill both of the Vortex keyboards relatively quickly. (Patience is not my main virtue.)
You're right about the programming problem: <FN> is special and apparently can't be programmed to RCtrl.
I would be interested to try Topre some day, but the choice of keycaps and layouts is just too limited for my taste.

@tron:
That dust cover looks really nice.
While I had the 5251 on my desk, I just put a small towel over it, when I didn't use it. At the moment it's well protected "in storage".

@mcmaxmcmc:
I've thought about that. I'm a bit afraid that the grey colour of the keycap might brighten up during the process as well, which would make matters worse.

[Hypersphere]

@Hypersphere:
<snip>
I would be interested to try Topre some day, but the choice of keycaps and layouts is just too limited for my taste.
<snip>

Well, I've managed to enliven my Topre-switch boards, at least with respect to colors. There are also high-form Topre caps on some models, but I prefer the standard Topre profile and cylindrical tops -- practical for typing.

rfr2afc.jpg (76.8 KiB) Viewed 13732 times

As for layouts, although I do not need to change the default layout on my HHKB, I change the RF layout to HHKB using a Hasu USB-USB converter. For those who want any number of different layouts on their HHKBs along with the ability to program macros, there is a Hasu replacement controller available.

[John Doe]

Nice beam spring in good condition, congrates.

But 1250usd becomes 1600usd, 350$ for shippng? Can't imagine that even.

[darkcruix]

Mine did cost me with all my cleaning and refurbishing processes, shipping, extra keys etc. way more.
You can find them on ebay for $3995 at the moment in a complete un-restored state.

I don't think, this is a reasonable price if you just look at it from a objective standpoint. I bought it for the history, for the unmatched typing feel, for my own excitement when I type on it, my collectors heart, the beauty to look at it etc. I bought something that makes me feel good. Hope that makes sense to others

[AJM]

... But 1250usd becomes 1600usd, 350$ for shippng? Can't imagine that even.

Shipping alone would have been something around 80, but to all of that then comes 19 % of import tax.

[digital_matthew]

You could try sanding the 8 key down to the original whiteness since they're double-shot. It will wreck the texture though.

[JP!]

Looks like it's in great condition

You may want to consider making a dust cover for it now that the contamination shield is removed. I had a seamstress make a couple prototypes including the one below. When my schedule clears up I'll have another one made like a tight fitting shell (without the spandex bottom).

Cover gallery:

https://imgur.com/a/AEs7Bz2

That turned out really nice. The logo is a nice touch

[tron]

Looks like it's in great condition

You may want to consider making a dust cover for it now that the contamination shield is removed. I had a seamstress make a couple prototypes including the one below. When my schedule clears up I'll have another one made like a tight fitting shell (without the spandex bottom).

Cover gallery:

https://imgur.com/a/AEs7Bz2

That turned out really nice. The logo is a nice touch

Thanks JP I wanted it to loosely resemble a Selectric dust cover with the piped edges and IBM badge. I'll post a separate thread in the photo/video section when I receive the shell version from the seamstress.

[Darkshado]

You could try sanding the 8 key down to the original whiteness since they're double-shot. It will wreck the texture though.

Could a drop of peroxide cream applied only on the legends with a toothpick work?

[darkcruix]

You could try sanding the 8 key down to the original whiteness since they're double-shot. It will wreck the texture though.

Could a drop of peroxide cream applied only on the legends with a toothpick work?

I wonder, if it is really oxidation that caused the yellowing.
I'd love to hear the thoughts of @Chyros (a PhD in Chemistry can be so handy at times).

[digital_matthew]

Could a drop of peroxide cream applied only on the legends with a toothpick work?

I wonder, if it is really oxidation that caused the yellowing.
I'd love to hear the thoughts of @Chyros (a PhD in Chemistry can be so handy at times).

I'm no PhD, but I'd say it's UV exposure that causes the yellowing. Peroxide may work, but it would be difficult to mask the area around the legend and the grey plastic could be discolored.

[hansichen]

UV exposure can be the reason for yellowing but sometimes there are other reasons. It depends on the used plastic. On my Beamspring only engraved keycaps were yellowed, I don't think that peroxide would help then. I'd rather keep it the way it currently is instead of taking the risk

[zrrion]

I think beam springs were made before the additive that makes ABS yellow was required, so they don't yellow with age. I've never heard of engraved caps for beam springs, but if they are newer than the doubleshot caps then they might use new ABS with the additives that cause discolouration

[SneakyRobb]

I think beam springs were made before the additive that makes ABS yellow was required, so they don't yellow with age. I've never heard of engraved caps for beam springs, but if they are newer than the doubleshot caps then they might use new ABS with the additives that cause discolouration

I believe the keycaps were made of mostly POM and sometime ABS. The white keys specifically were usually POM and not ABS. ABS of course famously having flame retardants added that overtime react with UV light to cause yellowing.

Some keys were all abs, some all pom, some both. Depending on legends and customization it was easier to engrave more unusual keycaps for customers than to make a mold.

See this thread

viewtopic.php?t=17318

Peroxide will reverse the yellowing on specifically ABS that is treated with the boron flame retardants. POM plastic does not react well to peroxide and it is highly recommended to not try using peroxide on POM.

Also the bottom of beamspring keys might look messy with little rough tabs around the edges but these are actually rough tabs to keep the layers of plastic together as POM is very hard to bond with other plastic. Without the tabs you could easily separate the shots of the beamspring caps.

You can easily tell the difference between ABS and POM. ABS just feels like ABS. The white POM keycaps feel weirdly greasy. This is why the surface of the keycaps is slightly roughed up on the beamspring. POM is a good bearing plastic material.

I think POM was used on the keycaps because POM is very easily machinable.


Also one would imagine that near the end of the beamspring lifetime when they made replacement keys they switched to the ABS plastic, which might be why there are occasionally keys that have yellowed legends.

[darkcruix]

Thanks a mil for the info - wasn’t aware at all. I thought they all have been ABS.

[SneakyRobb]

Hi no problem. I would say that I am not exactly sure what they are. I believe them to be some kind of POM like material but many companies have slight variants.

I have been trying to find out what specific material they are made out of. So far I have found that the service manuals recommend you do not use alcohol in the cleaning of any selectric or displaywriter keycaps. I believe the beamspring caps are basically the same. It says only the specific IBM* brand cleaning solution or mild dish soap.

That said the spacebar does say PEI in it which could stand for Polyetherimide

[inmbolmie]

@AJM, Thanks a lot for the " little hooks were bent" tip, I had one switch just affected by this that I thought was somehow broken and now feels perfect after reseating the hooks.

[SneakyRobb]

So I am now in the not sure what the keycaps are made of camp. I am waiting to hear back from the IBM archives guy. IBM developed acrylic injected keycaps in 1967 but I am fairly certain they are the clear relegendable ones on displaywriters, model f/ and pingmasters.

I tried the acetone on one black with white legend beamspring key and it sort of did something to the black outerplastic. No colour came off though onto the cloth. I could easily scrape off the slight discoloration and make it shiny again though on the black part so I don't know if it just has a laquer on it and the acetone did something to that laquer.

On the white plastic doubleshot inside of the same key, the acetone did nothing.

So the blackpart could be ABS although I have found no references to ABS being used in any manuals. It does say to never clean beamspring keycaps with alcohol of methyl chloride. This doesn't help much as we are not trying to tell the difference between pbt and abs. Most solvents will melt many candidates pretty equally. Also pbt wasnt really used until the 1970s.


The candidates I have been searching for are abs, acrylic, bakelite, polyimide, melamine, and other formaldehyde plastics. They could also be some variant of polycarbonate.

As for the PEI logo, I believe it is some kind of internal mold marking possibly a subdivision of ibm another company or an outdated material code. Most of the keys have mold marks like M-15 which is probably the specific mold to trace issues. The PEI mark doesn't look like those at all. From what I can tell Polyetherimide wasn't invented until the early 80s.

IBM did update their two-shot injection molding process in the early 80s for replacement keycaps, but it seems more likely they would have just used ABS. As well for some reason they decided to shoot the outer part of the key first then the insert. Perhaps there was some plastic bonding reason for this. Or maybe just nicer legends.

I do know that earlier IBM electric type writers had cellulose and bakelite keycaps that had laquer finishes. So I will wait for the archives guy to see if he has anything.

[Wingklip]

Does this fix the rattling that beamsprings with old foam make?