IBM Beamspring 5251/Great White low profile alternative shell

[lot_lizard]

Several months ago, I put together something that has since sat. The 5251 Beamspring is probably my favorite to type on amongst the few I have, but the outer shell profile is HORRID. I constructed a model for the inner assembly (top/bottom plates, controller, keys, etc), and started building a minimal case around it. I still need to clean up a few things about it:

• increased front radius (more rounded)
• platform to hold the solenoid (now above the controller since there is no cutout for it like other beamsprings)
• bottom plate
• integration (fasteners) with the bottom plate
• hole for the cable

Sticking this thread out here as a placeholder, and to see how much interest there would actually be. I know there have been quite a few 5251 finds as of late. I have no idea the cost, materials used, etc... but do know it would work well. The KEY profile pitch would now be very close to the Model M or F. I also like the idea of planning ahead for an optional wrist wrest to bring down the height of the beamspring switches in relation to the "desk".

If you have a 5251, Great White, or other beamspring where you would appreciate this design, post something here so I can get an idea if it is worth pursuing once my other projects die down. Thanks

5251 top profile for scale

IMG_6969.JPG (490.28 KiB) Viewed 10626 times

Refactored top shell with keys

beamspringCaseTopWithKeys.png (158.23 KiB) Viewed 10626 times

Refactored top shell

beamspringCase.png (11.35 KiB) Viewed 10626 times

5251 side profile for scale

IMG_6970.JPG (433.92 KiB) Viewed 10626 times

Refactored side profile

beamspringCaseSide.png (5.64 KiB) Viewed 10626 times

[seebart]

Ah I remember you telling me about this and it's still a great idea! The empty space on the sides can be cut making it much more modern and a lot more sleek.

[Slom]

that looks really cool would the reduced angle mean the top of the keycaps are now pointing away from the user?

[snuci]

I'm in if you do this. It's a great idea.

[lot_lizard]

that looks really cool would the reduced angle mean the top of the keycaps are now pointing away from the user?

Not at all... but the row pitch is definitely less severe. I have a few beamsprings, and all of mine have identical keycap profiles by row number. IBM seemed to be really experimenting with how to transition users away from typewriter key cap profiles (think Selectrics) throughout this era. The 327x's are actually quite row pitched, but the assembly is low in relation to the desk. The 3101 pitch is quite a bit more flat (similar to that of the refactored design), but the assembly itself sits substantially higher. The 5251 reverted pitch back more closely to the 327x's, but kept the assembly height of the 3101. They were definitely debating what to do, and I am sure modified based on feedback. Many of us have tried using the beamspring without a case, but I feel the keys need quite a bit more row pitch to be usable as a daily driver (to your point, the keycap pitches are then "off"). This is substantially taller in the back than having the assembly just sit flat on the desk.


I might in the end raise the numrow by 1/8 inch (maybe...), but I used without a case adjusting stilts for about a week testing pitches... and settled on this. The caps are definitely not inverted, but the home row is taller in relation to our conventional M when compared to the extremes (M/F have a curved backplane, where this is flat). It seemed to be an ideal trade-off all the way around, and my typing speeds (the part that matters), remained consistent. I by no means blaze as a typist, but can consistently crank out 90 wpm on a Topre (80 wpm on a beamspring.. they are just slower no matter what anyone tells you).

We'll continue to experiment on the row pitch if there is interest to really dial it in, and I should go back and add the key profiles to the renders so cap pitch itself is more obvious (thanks for pointing that out)

[XMIT]

Looks great. If you can path it for a Tormach CNC mill I can try building one here.

The endgame for me is the BSSK: Beamspring Space Saving Keyboard. With a Displaywriter board we're most of the way to having the right keys for a good layout. So with a PCB, and a case, I would have something I can actually use.

[seebart]

The endgame for me is the BSSK: Beamspring Space Saving Keyboard. With a Displaywriter board we're most of the way to having the right keys for a good layout. So with a PCB, and a case, I would have something I can actually use.

Oh now you're talking XMIT! Make my dreams come true. While searching old DT threads recently I saw another one of those:

keyboards-f2/speculation-analysis-of-be ... t2513.html

Spoiler:

35Jahre.jpg (53.85 KiB) Viewed 10550 times

[webwit]

You gotta up that game: the Beamspring HHKB.

Great show on that Display Station replacement case. My least favourite beamspring keyboard after the Displaywriter, which is just fugly.

[E TwentyNine]

Love this idea, was something I've been thinking about since the first time I opened up a beam spring and realized how much open spaces is in there.

[lot_lizard]

Thanks... I think in order to be cost effective, we might need to get creative with how it is constructed. If we do it correctly, we should be able to make one top plate pattern through a series of produced parts that can be assembled for really ANY beamspring.

What I have in my head will make more sense later when we refine the design (with pictures), but imagine all steel of various thicknesses... with the sides being universal end caps, the front and back are universal pressed plates (chopped down as needed), and the vertical slots separating the key areas are placed in after the outer rim of the shell is assembled. Then some light welding (could do myself in the low numbers we are talking about), everything is ground smooth, and then powder coating professionally. It is a little effort, but certainly not bad. If all the universal parts are cut and pressed ahead of time professionally... I could chop, assemble, weld, and polish each top shell in under an hour. Solid steel (better than any cast by far). The bottom shell would likely be a flat thick steel plate, potentially pressed with ~4-6 ribs if we see it needs it (the more ribs, the thinner the metal needed... So a balancing act there on cost).

I am almost positive we would never get enough volume to offer molds, but we should be able to produce these in reasonable bulk in an artistic way, and end up with something of better quality than was originally produced in any beamspring.

Concept needs time to harden, but the general idea.

EDIT: Plus this approach has the upside of you being able to construct custom layouts like you mentioned previously, but TKL wouldn't work unless you cut different end caps for the sides. 5 rows is the magic number in the universal version. The end caps would be the cheapest part to have one offs of though. I'm sure you could have a 7 row version (essentially what TKL is) cut by Big Blue Saw or the like for under 30-35 for the set.

[derzemel]

this is a great idea. I have also been having some ideas to make a wood case for my 5251, similar to the one you have designed.

But your steel sheet idea is much better, especially if you could make artistic welds, like walking the cup

BTW, do you havethis image in a larger format? I want a new wallpaper

[photekq]

Looking great so far. I think getting this made will be a nightmare though. I hate to imagine the CNC costs for that 3278-style curved top

[seebart]

True, the curve top might have to go for production if it drives up the price.

[derzemel]

True, the curve top might have to go for production if it drives up the price.

Looking great so far. I think getting this made will be a nightmare though. I hate to imagine the CNC costs for that 3278-style curved top

I think lot_lizard is thinking of press sheet metal and then welded, not CNC. Personally I do not see any issue with this, as long as the metal sheets are thick

[photekq]

Yeah, that definitely seems to be the most sensible approach to me. However, you've then got the issue of mounts, screw holes, etc. Lots of difficulties with any approach, I can tell you that! Not impossible though.

[seebart]

Either way I'm looking forward to this case. Would make the 5251 much nicer.

[Nuum]

For my 5251 I had the idea of a simple bend piece of sheet metal as top case and a plate as bottom, with e.g. wood as side walls. The side tabs of the beamspring assembly would then rest on the side walls.

Beamspring 5251 case schematic.png (3.07 KiB) Viewed 10354 times

Beamspring 5251 case schematic wood.png (2.82 KiB) Viewed 10354 times

If you variant is not too costly I'd be interested.

[lot_lizard]

If you can path it for a Tormach CNC mill I can try building one here.

I completely missed this before for some reason. You own one (if so... jealous), or have access to at reasonable rates? If you could let me know the model number, I can look up the dimensions for the working envelope and see what we could do. Also, is the controller driven by PathPilot, or something else? Even if this project falls outside the scope of its use, knowing it's availability in general could prove very interesting later (ie... Model MF cases).

For my 5251 I had the idea of a simple bend piece of sheet metal as top case and a plate as bottom, with e.g. wood as side walls. The side tabs of the beamspring assembly would then rest on the side walls

This would be a very similar assembly process as you described. The only variances would be in the materials used and methods of producing each assembly part enables some artistic liberties. Very similar concept in construction design though

[y11971alex]

Since we're on this subject, is it true that the internal assembly of beam springs are largely identical in profile?

[lot_lizard]

Since we're on this subject, is it true that the internal assembly of beam springs are largely identical in profile?

I can't speak for all beamsprings, but I can say the following:

• 3278
• 3279
• 3101
• 5251

All have very similar inner assemblies (webwit... EDIT: and Seebart... would be able to speak to some that I am missing). The width and key layout of each obviously varies, but the plates themselves are very close (all 5 rows with a consistent switch). Some will not take advantage of the bottom row for all key columns though, so we would have to be able to cover those gaps when missing.

The one notable difference is the solenoid cutouts at the back of the top plate. The 327x's have a cutout for approx. half of the solenoid to recess back into the top plate. The 5251 has the solenoid completely off to the side (this is the need for the mounting plate I mentioned in the first post). I would have to open the 3101 up again to remember how the solenoid was situated there. But other than the width, key layouts, PCB and controllers themselves, they are all largely identical.

[seebart]

Since we're on this subject, is it true that the internal assembly of beam springs are largely identical in profile?

Good question, I have never openend three of mine. I could imagine my 5281 "Beamfoot" is quite similair to the 5251 internally although the case has a different shape. I'll have to check.

[Scottex]

this is a great idea given that this is one of the most common (i think so) beamspring boards

[Invisius]

Really love this idea, I've grown fond of the 5251 layout after using it for some time. The bottom plate does raise the board up quite a bit, so I think doing a full case vs just the top plate would be a lot easier to make standard across all the different models.

Either way, I'm definitely interested! Sheet metal would look really cool with these keycaps.

[snuci]

I was originally thinking polished stainless steel but that might be pushing it.

[Invisius]

I was originally thinking polished stainless steel but that might be pushing it.

Well maybe not if we can find a few different shops to give us quotes on the metalwork. Wish I had a buddy with access to a CNC machine right about now.

[lot_lizard]

I was originally thinking polished stainless steel but that might be pushing it.

Stainless is actually very doable. When I get the pressed plates for the Model MF project back, you'll have an expectation of the result here. Once cleaned up, it could be polished or brushed (or whatever) pretty easily.

Well maybe not if we can find a few different shops to give us quotes on the metalwork. Wish I had a buddy with access to a CNC machine right about now.

Unfortunately curves like the dip of the top plate at the back absolutely suck on CNC's since the spindle likes to remain vertical in most setups (even pitches like we are proposing throughout the top plate are not ideal). Pressed metal, welded, really is the way to go. 75 amps on stainless with a mig/tig is quite easy. If a video would help, I could find something. Let me know

[seebart]

I'd love polished stainless steel for mine.

[Phenix]

now I regret I did not got an 5151 from the bulk from retrosammler (or his previous one)

[seebart]

now I regret I did not got an 5151 from the bulk from retrosammler (or his previous one)

I'm pretty sure he will have more for of those in the future.

[mecano]

What are you talking about? I did an username search with retrosammler and found nothing :/