Any Market For A Modern Capacitive Buckling Spring?

[Paralel]

Does anyone think there is a decent sized market for a modern capacitive buckling spring or beam spring keyboard?

[tlt]

If you could argue that it's good for gaming you have a winner

[Findecanor]

Indeed. The Cherry MX Blue is popular among gamers (for some reason...) but you don't see many clicky keyboards in offices. I have seen a few, but I would classify those workplaces as being more progressive than the norm.

[robo]

I think it's hard to say - it depends a lot on price and how it's marketed. Price could be on the high end but you'd still need to recoup considerable development costs (nevermind the licensing...). Marketing wise you'd have to have something that appeals to people beyond the current DeskThority crowd, which I think would be doable with a clean modern looking design.

It's clear that buckling springs are some of the best keyboards around for typing, and there's a market for typing-focused keyboards (WASD CODE models sell out, Topre's are more about typing feel than gaming, etc). The gaming market may be bigger, but possibly balancing that is the fact that people willing to buy a really nice keyboard for coding/writing are probably on average a bit more grown up and more willing to pay more (hence Topre prices, IBM SSK prices).

IMO the big problem with current buckling spring adoption is that you have to use a large and very old fashioned looking board, either a genuinely old one (which some of us enjoy, but most people are put off by), or an ugly and somewhat shoddily built new board from Unicomp, a company with all the aesthetic charm of an unwashed Bill Gates sweater. Yes, SSKs exist providing a more compact form factor, but are so rare and expensive as to not be an option for the regular buyer, and again, the 80's look is only appealing to a few. So effectively, most people never consider buckling springs when looking for a keyboard upgrade, because buying used is scary for people and buying new is unappealing. So they all go Cherry.

If Unicomp improved their manufacturing tolerances, paid attention to trends a bit more (tenkeyless...) and hired a good designer, I suspect that their boards would be much, much more popular. The fact is that nice keyboards are luxury items, and a big part of success in that market is making something that is aesthetically pleasing (even in an industrial kind of way) and that feels well made. Unicomp fail miserably at this.

If they cleaned up their act I suspect that membrane buckling springs could really give Cherry and Topre boards a run for their money and could gain a following that would then be obvious buyers for the next step up, to capacitative buckling springs.

I wonder how much Unicomp is worth as a company. Maybe DT can band together and buy them out

[Muirium]

Ah, Unicomp. There's a spark of creativity there lately, which is just as needed as Robo says. But you can never quite tell…

http://deskthority.net/keyboards-f2/uni ... t7739.html

Model Ms aren't capacitative buckling spring. A subtle technicality, of course, but my 60% Model F Kishsaver is the genuine article:



The greatest potential of any keyboard I have found so far to be "my only one".

Anyway, the grown up / keyboards are for typing crowd does definitely exist. The Code keyboard, Matias, even Das to an extent is all about them. That's where I came to the community from, too. My Kishsaver is unlikely to get much more FPS action from me than it has any time since it was made in 1986.

But buckling spring? Hmm. It's a hard sell, because they are what they are: old-school loud. Even a smart, brushed aluminium topped, thoroughly modern looking buckling spring board sounds like the 1980s. They were retro already, decades ago.

MX blue has the reputation of being the SO LOUD switch these days. The one you use to send a message to coworkers, or a spouse! But buckling spring is a whole other level on top. I love it, but it's a long way from mainstream. It really is closer to typing on a Selectric typewriter than a computer. An elegant weapon, for a more civilised age.

[quantalume]

I don't get the aversion to loud keyboards. I remember terminal rooms in the 70's where you'd have four to six keypunch machines and a dozen 327x terminals, each with a beamspring keyboard, clicking and clacking and thumping away. Nobody complained about keyboard or card punch noise, partly because it was drowned out almost entirely by the sound of the massive blowers keeping the IBM System/360 machine room down the hall cool.

And yet, people drive by my house with their stereos so loud I can hear them a block away even with all the doors and windows closed, and there is no widespread social pressure to control this behavior.

[Muirium]

Noise is a very complex thing. I really do loathe overhearing other people's music, and I truly detested the noise that hard drives and cooling used to make in computers too. But the sound of a good keyboard, or a typewriter, is sweet to my ears. Even when it's a family member's at night or just a stranger's, nearby. I understand why people can be upset at either thing, given the invasive property of sound that you mentioned. It's all shades of grey, beyond the mountain monastery…

No, not *that* kind. Filthy minded outsiders!

[robo]

People's reactions to noise depend a lot on how they feel about what's making the noise. That's why the car cruising down the street blasting rap music is obnoxious to some people but (apparently) attractive to others. Same goes for engine noise in a supercar (people love it) vs in a boring car (people want to minimize it).

Keyboard enthusiasts enjoy noisy keyboards because it's the sound of something they enjoy - both keyboards themselves and the nostalgia associated with clickety ones. For others it's a boring piece of equipment attached to a scary machine from a place they're forced to slave away in for most of their adult lives, and one that's noisy just makes it obnoxious rather than boring.

[pyrelink]

It's interesting. I find that with most sound, as long as it is my own noise pollution, it doesn't bother me. While I tend to be more respective at stoplights and neighborhoods, I don't mind loud music. But someone else's loud music? Infuriating.

On the opposite spectrum, I have always really liked the loud ass sound of typewriters, and keyboards. Whether it be the sound of someone typing away in the other room, or an office filled with clicky keyboards, I actually really like it.

What I really need to test, is how I feel when someone drives down the street, blaring amplified buckling spring, audio clips....

[Paralel]

I think it's hard to say - it depends a lot on price and how it's marketed. Price could be on the high end but you'd still need to recoup considerable development costs (nevermind the licensing...)...

It's interesting you mentioned licensing. I looked into the patents covering beamspring and buckling spring capacitive, as well as membrane, and they either were never approved, have lapsed, or expired. The beamspring design by IBM was rejected for patent here in the US. The patent that explicitly covers buckling spring membrane keyboards lapsed in 1997 with no hope of being resurrected. The patent that covers the capacitive buckling spring is more than 20 years old, which according to patent reform laws that were instituted in 1995, is the maximum term for a patent of that type that were granted before 1980. So, technically, anyone can make any of those types of keyboard and not pay anyone anything. I was quite happy when I saw that.


The controllers for the keyboards are quite a bit more tricky, all that code is copyrighted which lasts much, much longer. However, the keyboard enthusiast community seems to have done a good job of reverse engineering that code and producing original, modern alternatives that are completely functional.

[bitemyweewee]

I think it's hard to say - it depends a lot on price and how it's marketed. Price could be on the high end but you'd still need to recoup considerable development costs (nevermind the licensing...)...

It's interesting you mentioned licensing. I looked into the patents covering beamspring and buckling spring capacitive, as well as membrane, and they either were never approved, have lapsed, or expired. The beamspring design by IBM was rejected for patent here in the US. The patent that explicitly covers buckling spring membrane keyboards lapsed in 1997 with no hope of being resurrected. The patent that covers the capacitive buckling spring is more than 20 years old, which according to patent reform laws that were instituted in 1995, is the maximum term for a patent of that type that were granted before 1980. So, technically, anyone can make any of those types of keyboard and not pay anyone anything. I was quite happy when I saw that.


The controllers for the keyboards are quite a bit more tricky, all that code is copyrighted which lasts much, much longer. However, the keyboard enthusiast community seems to have done a good job of reverse engineering that code and producing original, modern alternatives that are completely functional.

That's fantastic! Why is this not common knowledge? So many people in the tech community, keybeards inc, think that it's still under patent and that unicomp own the rights to it. Heck, even I thought that.
I think we need to suggest CM look into this, they seem pretty open and innovative toward the keybeard community

*edit: Sent off a pm to CM Bram to see what he thinks of the idea

[Paralel]

Thinking about it further, I realize that some of the original controller code may indeed be in the public domain after all.

Any code incorporated into any device, including keyboards, made before March 1989 when the Berne Convention Implementation Act came into force that did not have a copyright mark on the chip nor proper copyright notice where the code is "fixed" needed to be registered with the US Copyright Office for it to be considered copyrighted. Authors of fixed code that was not properly marked or noted were given until April 1st, 1994 to correctly register their works with the US Copyright Office before the code automatically entered the public domain.

As such, if IBM didn't mark the chips that contained the code properly, didn't provide any notice of copyright concerning the code, and didn't subsequently register it with the US Copyright Office, it is in the public domain. All it would take to know for sure is to check the chip and search the US Copyright Office for any registrations, if there are none then it is not copyrighted under the law.

Roland Corp. ran into a situation that highlighted this particular loophole when they tried to say that the ROM code in the MT-32 synthesizer was copyrighted. However, the ROM code was "fixed" in 1987, they didn't mark the ROMs with copyright marks, provided no notice of copyright, and didn't register it by April 1994, so, without them even realizing it, their MT-32 ROM code passed into the public domain, and just like a genie, once it's free, no subsequent claim of ownership can be asserted on it.

[Muirium]

The controller is no problem. We already have our own replacement developed!

http://deskthority.net/marketplace-f11/ ... t7993.html

I use Xwhatsit's Model F controller every day now. It's a lot more configurable than IBM's original. There's at least one more project out there that has succeeded, too.

Oddly enough, the hard part is the buckling spring mechanism itself. It's not as similar to existing needs – like PCBs – to easily manufacture for us.

[lowpoly]

Thinking about it further, I realize that some of the original controller code may indeed be in the public domain after all.

Why go through the trouble of reverse-engineering a non-programmable controller?

[bhtooefr]

I'd say we're to the point of having the controller problem solved.

As far as the switches... buckling spring is fine, although requires some model-specific configuration that Cherry MX does not (but anything plate-mount is about as model-specific, IIRC, as far as manufacturing requirements). Beam spring, however, is a very, very tall switch that is complicated to assemble.

[Paralel]

That's fantastic! Why is this not common knowledge? So many people in the tech community, keybeards inc, think that it's still under patent and that unicomp own the rights to it...

I guess no one ever looked into it? The funny part is, none of the relevant patents are actually owned by Unicomp. At most, they were licensing them. All the relevant patents were either owned by Lexmark or by IBM themselves, until they expired, according to the US Patent and Trademark Office database.


You can check it out here:

http://www.google.com/patents/US4528431#v=onepage

That patent is lapsed with no hope of recovery, and even if it hadn't lapsed, it would have expired by now anyway. All subsequent patents from IBM and/or Lexmark that reference this mechanism are lapsed with no hope of recovery and/or are expired. Any patents that it references are expired. As such all the mechanisms are in the public domain.

This is true for any switch mechanism patented by any company, Alps, etc... that were granted a patent on or before May 23, 1994, as that was 20 years ago today.

I'm looking forward to Jan. 1st 2020, that will be the day that every patent of the 20th century will be in the public domain, and you have to admit, that will encompass an unbelievable amount of technology.

[bhtooefr]

I will note that around 1997-1998 or so, the Chinese started cranking out downright atrocious knockoffs of membrane buckling spring boards.

And, Alps clones have existed for over a decade. Cherry clones have existed for a few years now.

To expand on my point about the height of beam spring... ISO 9241-4:1998 specifies that the home row height of the keyboard shall not exceed 35 mm (and preferred height is not greater than 30 mm). That's from the strike surface of the unactuated key to the desk. 0 degree slope is available to us (5 degree being the minimum recommended slope), but even then, I'm fairly sure a beam spring switch and cap is quite a bit taller than 35 mm.

[Paralel]

I will note that around 1997-1998 or so, the Chinese started cranking out downright atrocious knockoffs of membrane buckling spring boards.

And, Alps clones have existed for over a decade. Cherry clones have existed for a few years now.

To expand on my point about the height of beam spring... ISO 9241-4:1998 specifies that the home row height of the keyboard shall not exceed 35 mm (and preferred height is not greater than 30 mm). That's from the strike surface of the unactuated key to the desk. 0 degree slope is available to us (5 degree being the minimum recommended slope), but even then, I'm fairly sure a beam spring switch and cap is quite a bit taller than 35 mm.

Interesting about the ISO requirement. I guess that leaves capacitive buckling spring. I'm thinking something like a kishsaver, less of a border around, with maybe Ti replacing the steel plates to make it light but strong.

[bhtooefr]

Ti isn't necessary, there's quite a bit of weight reduction available just by switching to Al (which is what the 5155's Model F uses).

[Hypersphere]

I thought I had acquired the ideal keyboard when I was able to find a NIB IBM SSK to replace my full-size Model M. Then I discovered capacitive buckling springs; this opened a new world for me. Now I am in the process of restoring and remapping an IBM XT, and I have plans to do the same for an AT and F-122. I much prefer the sound and feel of the capacitive buckling springs over any other type of switch I have tried, including all flavors of Cherry mx, Matias quiet and tactile, Topre, and membrane buckling springs, as found in the classic Model M.

It would be hard to beat a genuine IBM, but I think there are many who would buy a new product with similar quality as the original but in a standard and more compact layout.

[quantalume]

ISO 9241-4:1998 specifies that the home row height of the keyboard shall not exceed 35 mm (and preferred height is not greater than 30 mm).

Is that measured from the bottom of the keyboard or the top of the desk? If the former, it's a ridiculous requirement, since anyone skilled in the art of ergonomics knows that keyboards should be on their own, adjustable platform and not on the desktop. Bureaucrats.

[quantalume]

Ti isn't necessary, there's quite a bit of weight reduction available just by switching to Al (which is what the 5155's Model F uses).

Interesting. i did not know that. In my opinion, one of the few flaws of the Model F is the use of plain steel for the top plate. People will surely spill stuff on their keyboard, and plain steel will surely rust in time.

[Hypersphere]

I don't get the aversion to loud keyboards. I remember terminal rooms in the 70's where you'd have four to six keypunch machines and a dozen 327x terminals, each with a beamspring keyboard, clicking and clacking and thumping away. Nobody complained about keyboard or card punch noise, partly because it was drowned out almost entirely by the sound of the massive blowers keeping the IBM System/360 machine room down the hall cool.

And yet, people drive by my house with their stereos so loud I can hear them a block away even with all the doors and windows closed, and there is no widespread social pressure to control this behavior.

As they say, "I hear you!" I don't mind noise so much when it is being made for some useful purpose and if it is machine noise of some kind akin to white noise. Sounds that I cannot stand include people talking, other people's "music", crying babies, and barking dogs. In contrast, although I prefer silence, I can tolerate lawn mowers, building ventilation fans, or mechanical keyboards.

[bhtooefr]

The exact language is:

height from the centre of the strike surface of an unactuated key in the home row to the support surface

And, thickness does matter even on a keyboard tray, because it affects how close to the legs the home row can be.

[Muirium]

Anywho, IBM already solved all that with buckling spring. The Kishsaver's even lower profile than the XT: the typical rounded pyramid profile of the other Model F cases is instead a rounded rectangle on this one, with a flat rather than a sloping top.

"Low profile" is IBM's own term for them:

The Model F was the first design to make use of IBM's buckling spring mechanism. It was a follow on to the earlier Beam Spring keyboards. In addition to being cheaper, they were also more compact and lighter than their predecessors – to the extent that IBM's sales literature described the keyboards as "low profile" and "ergonomic". The Model F made use of the same design of capacitive contacts as the beam spring keyboards, and thus many early Model F designs were buckling spring versions of beam spring keyboards, although the exact electronics differed slightly.

http://deskthority.net/wiki/IBM_Model_F ... _Features/

A modern day Model F would be an amazing thing. Especially if the switches were discrete, so we could move them around at will for ourselves. Custom keyboards are a big draw for me, and, despite my taste, MX owns them: both for modularity and cap selection. Two big problems for buckling, and beam, spring.

[quantalume]

The exact language is:

height from the centre of the strike surface of an unactuated key in the home row to the support surface

And, thickness does matter even on a keyboard tray, because it affects how close to the legs the home row can be.

I can see this being a problem for Lilliputians.

A modern day Model F would be an amazing thing. Especially if the switches were discrete, so we could move them around at will for ourselves. Custom keyboards are a big draw for me, and, despite my taste, MX owns them: both for modularity and cap selection. Two big problems for buckling, and beam, spring.

How about a buckling spring mechanism with MX cap compatibility? One would need to design an MX-compatible stem, and it might add a bit of height to the switch. Better be careful not to upset any bureaucrats. I don't know how doable a discrete switch would be, given how sensitive capacitive switches are. Why not a hall-effect, buckling-spring switch! You could put a small magnet on the end of the flapper.

Actually, I think a membrane BS switch could have just as nice a feel as a capacitive one if it were designed properly (no rubber mat, and a sitting over a heavy plate).

[Muirium]

Membranes are a design flaw, honestly. Susceptible to spills, general physical wear, and right buggers to get access to in the Model M, at least. Besides, they're the antithesis of custom keyboards. Discrete switches are so the way to go if inventing new ones.

You're right about Hall Effect, though. It's easily the equal of capsense in every positive regard, plus it's inherently modular. Behold!


http://deskthority.net/photos-videos-f8 ... t6296.html

A hall effect buckling spring module with an MX mount… oh yeah! It'd be tall, but it would make me happy!

[xwhatsit]

Hall effect is cool, and it gets around the same problem that capacitive switching so neatly solves—the ability to get rid of the scratchy horrible mess that is physical contacts (c.f. Cherry MX/Alps). The Model M gets around this mechanically, but requires a flipper to mash the membrane pretty hard to ensure contact. This is, I assume, why the Model M feels much more clunky and forceful than a Model F.

I don't know if I quite get the modular switch thing, aside from people like us being able to do things like build keyboards without PCBs.

That said, there's nothing stopping you from making a modular capacitive switch, in the same manner that the hall effect switches were made. You could damn near crank out something right now using an existing Model F barrel and spring/flipper and make a tiny key-sized PCB that fits on the bottom of the barrel that just has the sense pads on it and two terminals to solder to. Make it out of really thin PCB material to make it thinner, mould a nice bottom piece to clamp it all in place and you're done. All you need then is a capacitive controller in the same manner as the ibm-capsense-usb boards (you could shrink it way down as you don't need the 3.96mm spacing for the 30-pin ribbon cable).

You could integrate an individual capacitive IC into each switch (Atmel/ST/TI make that kind of thing, for stuff like touchbuttons on modern TVs, they're little SOT23-sized things) to allow you to use something like a Teensy to scan them easily, but it would be far easier to do a centralised capacitive controller as per the status quo. As I said you could shrink this down, make it dumb so it doesn't do its own USB stuff, and let people plug in their Teensy-style thing to do their own scanning, or just centralise it as per the ibm-capsense-usb boards.

[bhtooefr]

I think you could really do Brother-style BS modules, though, clipping into a plate and PCB assembly. Yes, you need to get plates made, but plate mount Cherry has the same requirement.

And most customs are using custom PCBs, not point-to-point wiring. (Custom membranes are certainly harder, though.)

[Muirium]

Hall effect, like capsense, doesn't even need diodes as I understand. So hand wiring one of those would be a cakewalk!

Custom PCBs are cool (and much easier to source than custom membranes, let alone membranes that have a chance in hell of lasting) but there's a certain solder fume induced magic to hand wired matrices. Let me tell you…