ReSKCM - Optical rework, and cloning of Alps SKCM/SKCL switches

[Lynx_Carpathica]

Hi there.
I just got to the point when I can't take my mind of of the cloning scene. I MUST make an at least as good, or even better clone of Alps SKCx switches.
My idea is:
-Either recreate the original switchplate design
-Or make an either drop in, or clone with optical sensing mechanism.

If you think about it, it's pretty easy to see that SKCL/SKCM switches can be very easily made optical. We just need a PCB with an IR led and a photo diode or something to detect the change in distance between the switchplate and slider. Some SMD stuff and it can be easily achived. In addition, it can be made analog if we apply a gradual change pattern into the slider that can detect the electonics. This can be made a drop-in replacement for old switches (keep in mind, you'll need a new mobo as well, but you can keep your precious SKCM Blues). If we do it right, we can make it 2 pin I think.

First, we need the exact materials and internal meausrements. Tactile/cliclk leaf spring bend angles, leafspring weighting. We have to find a company to reverse-engineer the switch housing and slider materials. What companies do this?

I am politely asking anyone who can tell me anything... Let's make this a reality, shall we? We recreated Model F's. Don't think SKCMs are impossible, especially theesedays. I'm not good at CAD, nor am I an engineer. Any kind of help is indeed, welcome.

We could try writing an email to Alps to license us the original design as well if they still have it.

[joebeazelman]

Your youthful naïveté and bravado is invigorating and inspiring, but you're painfully unaware of the magnificent scope of such an endeavor. Recreating these beloved switches has been a recurring topic of numerous threads here on DT and elsewhere. Every serious attempt has been a lone winding road littered with dried up corpses slain by broken promises, demoralization and dead ends. Even the most earnest attempt has resulted in less than stellar results. Matias Keyboards is the only one that has achieved any level of success and even so, its recreation has been half-hearted in its exacting of Alps' prized sound and feel.

It may be possible to recreate Alps to a high level of precision, but it's only half the battle. The rest requires assembling the switches. If human labor is employed, the resulting unit cost would greatly exceed what the market will bare. Automated assembly machines would be required to drive down the costs. Unfortunately, such a system is capital intensive and outsourcing the assembly--not without its own set of risks--isn't enough to drive down the costs signficantly.

To even consider putting a foot down on this perilous road, the market must be clearly established. While the SKCM has many enthusiasts as exhibited here on DT, its prospects in terms of size and potential growth is unknown. I would wager the market is very small, otherwise it would have already been attempted by at least one of the numerous Asian switch manufacturers who instead see more profits in cloning MX switches.

Even with a significantly large and growing market, there are other risks posed. It is possible the resulting product turns out lackluster in its quality, especially for a product demanding an exact reproduction of sensory characteristics, which are highly subjective. Any seemingly minor deviation from the original may result in market rejection. The only way to ensure success is a long process of testing and refinement along with complete supervision of the manufacturing process which also makes it incongruous to outsourcing or contract manufacturing.

[TheInverseKey]

Spoiler:

Your youthful naïveté and bravado is invigorating and inspiring, but you're painfully unaware of the magnificent scope of such an endeavor. Recreating these beloved switches has been a recurring topic of numerous threads here on DT and elsewhere. Every serious attempt has been a lone winding road littered with dried up corpses slain by broken promises, demoralization and dead ends. Even the most earnest attempt has resulted in less than stellar results. Matias Keyboards is the only one that has achieved any level of success and even so, its recreation has been half-hearted in its exacting of Alps' prized sound and feel.

It may be possible to recreate Alps to a high level of precision, but it's only half the battle. The rest requires assembling the switches. If human labor is employed, the resulting unit cost would greatly exceed what the market will bare. Automated assembly machines would be required to drive down the costs. Unfortunately, such a system is capital intensive and outsourcing the assembly--not without its own set of risks--isn't enough to drive down the costs signficantly.

To even consider putting a foot down on this perilous road, the market must be clearly established. While the SKCM has many enthusiasts as exhibited here on DT, its prospects in terms of size and potential growth is unknown. I would wager the market is very small, otherwise it would have already been attempted by at least one of the numerous Asian switch manufacturers who instead see more profits in cloning MX switches.

Even with a significantly large and growing market, there are other risks posed. It is possible the resulting product turns out lackluster in its quality, especially for a product demanding an exact reproduction of sensory characteristics, which are highly subjective. Any seemingly minor deviation from the original may result in market rejection. The only way to ensure success is a long process of testing and refinement along with complete supervision of the manufacturing process which also makes it incongruous to outsourcing or contract manufacturing.

11/10

Couldn't have said it better myself.

Costs too much, little interest. Alps still can be found in decent condition.

[Polecat]

According to the wiki Matias placed an order for a million (!) switches to keep Alps SKBM in production until their own replacement switches were available. That's no small chunk of change, even at a fraction of a dollar per switch (it's anybody's guess what the actual price was). Keeping in mind that Alps SKBM was a simplified (literally) replacement for the SKCM complicated switches, which themselves had undergone changes in order to compete in the PC market, it's probably not realistic that early complicated Alps switches will ever be faithfully reproduced, no matter how much we'd like to think so. In addition, the mystique, or placebo effect, or whatever you want to call it, of blue Alps will not be there with any reproduction switch, and that's largely what fuels the (inflated) prices of blue Alps switches and keyboards. Then again, perhaps someone out there has millions of dollars burning a hole in their pocket and is willing to pay to get them back in production, you never know!

[ZedTheMan]

An SKCM recreation is one of those potential money pits I would throw my money in if I won a very large lottery, but not otherwise.

[Lynx_Carpathica]

I think a kickstarter project would solve the issue. At least part of it
And I have to buy a few switches, even if they are just loose. Specifically blue alps and Mathias switches.

[Lynx_Carpathica]

No... Haven't tried it. Since I can't get hold of any. But, I have tried SKBM clones, and I genuinely thought that they were tactiles. And I keep hearing that Mathias is disappointing.

Anyway, I'll try creating some optoelectric stuff for any alps switch. That will probably be more interesting than anything. And if it doesn't work out, I'll let you know. I have "raw" PCB lieing around, so I will definietly try to make some optoelectric switches, and see if they work. But I have to find the suitable SMD IR LEDs and photodiodes that are not too thin, nor too thick.

Also, if I can get hold of some, I'll take them appart and make a tactile leaf comparison. and swap parts around the switches (Blue, and White/Ivory and Orange or Salmon) I asked around in the Alps appr. topic if someone could compare it for me, but noone did. And I haven't been able to find out the difference in the tactile leaves. I'll be back with this topic after I have been able to hunt down a blue alps keyboard that will not cost me an arm and a leg.

EDIT:
But you're right... I have no right to talk about recreating what I don't even know how the given thing feels. And I must apologize for that. I'll try gathering something, and then think about doing this later when I'm a bit more mentally mature regarding this huge byte that is cloning of a switch.

[ZedTheMan]

One reservation I may have in particular with this idea, though I do encourage you to try to make prototypes for the sake of it, is that the complicated switchplate and contact leaf actually provide a part of the alps feeling, especially in linears. That isn't to say that an opto-electric version wouldn't be good, but, even with all else somehow being equal, it won't feel the same, and a lot of people will not be happy with that.
SKCM, especially blues, I have once heard compared to the wings of Icarus. Flies like the gods, but it was not made to last.

[Findecanor]

You could take a look at Aimpad: a similar project which was originally converting Cherry MX to analogue optical, and read about the project's hurdles along the way.
The Flaretech uses a prism in a special slider to direct IR light back to the PCB. Then there is at least one Cherry MX-like switch using the centre shaft in the stem for on/off sensing.

Personally, I don't think Alps' lineage is interesting for anything but for how to achieve tactile/clicky feel.
The upcoming Input Club Keystone keyboard and its upcoming Silo analogue hall effect switches will have a tactile variant that is supposed to resemble Alps' tactile switches in its feel, but we'll see whenever it ships.
Edit: I should also mention that specs, schematics, firmware and switches have been promised to become available right after the Keystone has shipped.

[TheInverseKey]

I'll be back with this topic after I have been able to hunt down a blue alps keyboard that will not cost me an arm and a leg.

Good luck and happy hunting!

[kmnov2017]

I've had 5 Alps based boards over the years (2 SKCM blues, 1 white, 1 orange, 1 salmon), they were so disappointing it'd be such a waste of time, money and resources to try get them remade.

[Polecat]

No... Haven't tried it. Since I can't get hold of any. But, I have tried SKBM clones, and I genuinely thought that they were tactiles. And I keep hearing that Mathias is disappointing.

Anyway, I'll try creating some optoelectric stuff for any alps switch. That will probably be more interesting than anything. And if it doesn't work out, I'll let you know. I have "raw" PCB lieing around, so I will definietly try to make some optoelectric switches, and see if they work. But I have to find the suitable SMD IR LEDs and photodiodes that are not too thin, nor too thick.

Also, if I can get hold of some, I'll take them appart and make a tactile leaf comparison. and swap parts around the switches (Blue, and White/Ivory and Orange or Salmon) I asked around in the Alps appr. topic if someone could compare it for me, but noone did. And I haven't been able to find out the difference in the tactile leaves. I'll be back with this topic after I have been able to hunt down a blue alps keyboard that will not cost me an arm and a leg.

EDIT:
But you're right... I have no right to talk about recreating what I don't even know how the given thing feels. And I must apologize for that. I'll try gathering something, and then think about doing this later when I'm a bit more mentally mature regarding this huge byte that is cloning of a switch.

Since you're on a budget, like most of us, I'd suggest finding an early white Alps keyboard instead of blue Alps. The early white switches are very similar to blues, perhaps even the same internally, and with a bit of luck and patience you can pick up a nice one for a much cheaper price. Look for something from 1989-90, like an early Focus or Monterey, and try to find one that's clean inside.

It's been a while since I played with opto stuff, but back in the day we used to open up the slotted photo interruptor modules as a source for phototransistors and IR LEDs. Some of those have discrete transistors and LEDs inside, you just need to open up the housing and slide the parts out, very much like an Alps switch in fact.

[m_raggie]

I've had 5 Alps based boards over the years (2 SKCM blues, 1 white, 1 orange, 1 salmon), they were so disappointing it'd be such a waste of time, money and resources to try get them remade.

Disappointing as in you had your hopes up and/or they were all poor quality? I'm satisfied with the key feel I've experienced via Alps switches. What were your expectations?

[Lynx_Carpathica]

Since you're on a budget, like most of us, I'd suggest finding an early white Alps keyboard instead of blue Alps. The early white switches are very similar to blues, perhaps even the same internally, and with a bit of luck and patience you can pick up a nice one for a much cheaper price. Look for something from 1989-90, like an early Focus or Monterey, and try to find one that's clean inside.

It's been a while since I played with opto stuff, but back in the day we used to open up the slotted photo interruptor modules as a source for phototransistors and IR LEDs. Some of those have discrete transistors and LEDs inside, you just need to open up the housing and slide the parts out, very much like an Alps switch in fact.

Got one. My chicony is decked out with them, from a KB-6251EA. Unbranded, shiny tops, very tactile, but sadly they bind just a bit. I'll try NeKtaring it when we find out the best stuff for that. Anyway, I really liked it so far, and can't go back to MX blues, but can't use it either because it binds... Somehow, I didn't noticei it a at first. Anyway, for now I'm using a futaba MA Clicly board

[hellothere]

The upcoming Input Club Keystone keyboard and its upcoming Silo analogue hall effect switches will have a tactile variant that is supposed to resemble Alps' tactile switches in its feel, but we'll see whenever it ships.

That intrigued me, so I spent a few minutes looking into them. They're a year behind schedule. They're hoping to have an update within a week ...

[Rayndalf]

I think that the Matias/simplified Alps angle is valuable. If you can determine what these switch designs got right, and what they got wrong, you will know which parts you can source from Matias and which parts are missing or need to be redesigned.

Complicated Alps switches are renowned for their click/tactility, but I found Salmon Alps to be more pleasent when linearized. That said, marketing an Alps clone as a smooth linear is suicide, their are dozens of smooth MX clones (most of which are made by JWK/Durock ). Your goal (if you want people to throw money at you) is to recreate blue or brown Alps

Designing improved click leaves similar to complex Alps that could be installed in Matias or simplified Alps could be a very cost effective way to start reverse engineering Alps switches. I would focus on the parts that were cut from later Alps/Matias. If you can find a way to take simplified switches and "complicate" them you have a winning product. Focusing on stems and leaves first seems like a safer bet than trying to also manufacture and prototype the housings as well (assuming Matias got the housings mostly right).

Good luck!

[Lynx_Carpathica]

You are absolutelly right. You literally wrote down my absolute first idea that was:
-A recreation of SKCM Lime green tactile switches - since they are so rare.

And then I rolled the idea forward, and got where we are right now.

EDIT:

If Xiang Min KSB switches are any similar to Mathias switches regarding the keyfeel, that's enough for me to know that they are not the best linear switches. No linear switch sould be AS tactile as theese are. I was literally convinced that they were tactile.
However, I'm convinced that this is because it takes way more to press together the thicker contacts, than to press a leafsrping over a plastic tab over a copper foil over a hard brass/copper plate to bridge a contact.

But, if we'd remove the leafpsing álá' optical mod, we'd probably return to the linear curve of for example MX swithes. If we take a look at any SKCL/SKCM graph, we can see that after it clears the leafspring of the slider, it flatens the force curve, and it stays low.
Why is it only probably? The click and tactile leaves could have the same (let's call it the "braking", since it's literally friction) effect as the leaf spring of the contacts.

Theese are only possibilites. What are your opinions?

[SneakyRobb]

You are absolutelly right. You literally wrote down my absolute first idea that was:
-A recreation of SKCM Lime green tactile switches - since they are so rare.

And then I rolled the idea forward, and got where we are right now.

EDIT:

If Xiang Min KSB switches are any similar to Mathias switches regarding the keyfeel, that's enough for me to know that they are not the best linear switches. No linear switch sould be AS tactile as theese are. I was literally convinced that they were tactile.
However, I'm convinced that this is because it takes way more to press together the thicker contacts, than to press a leafsrping over a plastic tab over a copper foil over a hard brass/copper plate to bridge a contact.

But, if we'd remove the leafpsing álá' optical mod, we'd probably return to the linear curve of for example MX swithes. If we take a look at any SKCL/SKCM graph, we can see that after it clears the leafspring of the slider, it flatens the force curve, and it stays low.
Why is it only probably? The click and tactile leaves could have the same (let's call it the "braking", since it's literally friction) effect as the leaf spring of the contacts.

Theese are only possibilites. What are your opinions?

Always ask yourself 2 questions whenever presented with possibilities

1. Is it possible?
2. Is it practical?

It's very easy to get stuck on the yes of question one, saying "of course it's possible." A yes to question 1 is almost a given. The second one? Not as much, but. But, you should proceed.

That pivot to the second question is where production is made.


I would recommend you get some FDM/resin 3d printers. Learn cad and just make it. If you have a vision of a new alps switch, just do it.

There is no finished good that starts perfect. This is iterative. simplified alps didn't just appear. It was a simplified skcm. Which itself reused skcc parts. Do they even all match the patent drawings? Idk, but they changed. You need to make the switch.

The thing you need to do, is the doing part. Everyone wants to make this or that happen. Everyone here would love skcm, model f, beamspring etc to reappear in full production at a nice cost but they need force behind them. They need money and project management to happen. Sometimes, like model f. That force exists.


I do not want to discourage you, but your enthusiasm must be transferred into sweat. I mean this in a very positive way. You need to make models, try them out. Then take steps to determine costs of parts etc, interest checks, and then getting money.

Think of it like a pizza shop in your city. We can all say it might be good to open the pizza shop and agree, but who is paying. What is the market for it.

On your switch thoughts:
I would say that changing the sensing method of alps switches, removing the switch plate, would change the feel. Because the sensing relies on a horizontal movement that has friction. Its a bit reductive, but imagine a linear alps where you remove the switchplate. Will it be smoother? Yes because there is less friction, but it now wont actuate. Now, It might be possible to replace the switch plate with a skcm brown style blank that still has a actuation metal leaf. Thus recreating the same friction event.

How could we confirm that? Someone needs to test it. To try it to measure it. That person is you!

Recreating a switch while making fundamental changes can be a challenge.

I would not discourage your endeavour. Far from it, please proceed. Too many times do we hear the word impossible. I would say that It is best to start the project.

I would encourage you to try to think of a way to modify the slider such that you can install a sensor that fits such that you can maintain that drop in replacement. Is it a magnet like the SILO switch at the bottom. But then remember if its a drop in replacement, what on the old PCB can interface with the new sensor. How would a drop in work with an old pcb? etc etc.

What benefit does that sensing method offer over switch plate? Is it cheaper or easier?

Basically what I really mean, is that to do this you need to try. Get the tools, the printers, the cad software. Experiment. There are a million failed designs for the most successful inventions, and most of the failed designs were also made by the successful inventor.

If your question is "I wonder what it would be like to be on the other side of the stream?" The answer, is to put your foot in that water and go.

[SneakyRobb]

Also if you need any files for skcm I have been making and can help to provide you with them!

[Lynx_Carpathica]

As soon as I'm finished with the school's... I'll be working on it, and after I bought my fist car... (the 3d printer I mean) I won't abandon the project. But until then, I can make measurements, and learn cad. That's also infinitely useful. Especially modeling the different leaves. So I started looking for an M0115 and a blue alps board so that I can start tinkering.
Even if nothing comes out of it, as far as the optical sensing, or the cloned complicated switches go, I want to contribute at least someting to the "legacy" of Alps switches, and determining what makes each switch different. Like blue alps to white alps, or salmon to Ivory. That sort of thing,

Thanks for the good words.

[PlacaFromHell]

Hi. I see you have quite a lot of inspiration and decision to actually do (at least) some really interesting experiments. If you don't mind I would like to suggest another similar switches to put an eye on, like Fujitsu magnetic reed tactile, HTK and ProWorld MX. It's clear that the leaf is the core of an Alps switch, followed by the switchplate, which may be the only problem here as it affects the feel too but you want to remove it. If I were you, my first attemt would be fuse both the switchplate shape and the click leaf and use its movement to close and open an optical sensor. An optical brown Alps would be the easiest first approach.
Said that, CAD might be scary, but for something like a switch, with no organic shapes at all, you can easily develop decent skills. I personally use Solidworks, so if you go with the same route just tell me and I'll give you some help.

[Polecat]

Since you're on a budget, like most of us, I'd suggest finding an early white Alps keyboard instead of blue Alps. The early white switches are very similar to blues, perhaps even the same internally, and with a bit of luck and patience you can pick up a nice one for a much cheaper price. Look for something from 1989-90, like an early Focus or Monterey, and try to find one that's clean inside.

It's been a while since I played with opto stuff, but back in the day we used to open up the slotted photo interruptor modules as a source for phototransistors and IR LEDs. Some of those have discrete transistors and LEDs inside, you just need to open up the housing and slide the parts out, very much like an Alps switch in fact.

Sorry, it took me a while to dig these out of storage. Cheap and easy source of IR LEDs and phototransistors. Photointerruptor module and reflection sensor. Many brands and part numbers available, all work pretty much the same way electrically, but some like these two are much easier to open up to salvage the internal parts.

[Lynx_Carpathica]

Hi. I see you have quite a lot of inspiration and decision to actually do (at least) some really interesting experiments. If you don't mind I would like to suggest another similar switches to put an eye on, like Fujitsu magnetic reed tactile, HTK and ProWorld MX. It's clear that the leaf is the core of an Alps switch, followed by the switchplate, which may be the only problem here as it affects the feel too but you want to remove it. If I were you, my first attemt would be fuse both the switchplate shape and the click leaf and use its movement to close and open an optical sensor. An optical brown Alps would be the easiest first approach.
Said that, CAD might be scary, but for something like a switch, with no organic shapes at all, you can easily develop decent skills. I personally use Solidworks, so if you go with the same route just tell me and I'll give you some help.

Yes. The leaves are a true concerns about the Alps switchfeel. That's why I have to perform a lot of testing, and comparing.

Polecat: What I imagined is all based on SMD components, and a simple PCB. Idea: Check the reflections of the slider. Anyway... For now, I have to stop thinking about it, and focus on my studies. I'm behind scheulde quite a bit, and I don't want to fail.

Anyway... I put the gus into the WTS thread, and hope someone will get it.

[SneakyRobb]

Hi. I see you have quite a lot of inspiration and decision to actually do (at least) some really interesting experiments. If you don't mind I would like to suggest another similar switches to put an eye on, like Fujitsu magnetic reed tactile, HTK and ProWorld MX. It's clear that the leaf is the core of an Alps switch, followed by the switchplate, which may be the only problem here as it affects the feel too but you want to remove it. If I were you, my first attemt would be fuse both the switchplate shape and the click leaf and use its movement to close and open an optical sensor. An optical brown Alps would be the easiest first approach.
Said that, CAD might be scary, but for something like a switch, with no organic shapes at all, you can easily develop decent skills. I personally use Solidworks, so if you go with the same route just tell me and I'll give you some help.

Yes. The leaves are a true concerns about the Alps switchfeel. That's why I have to perform a lot of testing, and comparing.

Polecat: What I imagined is all based on SMD components, and a simple PCB. Idea: Check the reflections of the slider. Anyway... For now, I have to stop thinking about it, and focus on my studies. I'm behind scheulde quite a bit, and I don't want to fail.

Anyway... I put the gus into the WTS thread, and hope someone will get it.

Ive been working on some skcm models, shoot me a dm and we can discuss

[Lynx_Carpathica]

Hi. I see you have quite a lot of inspiration and decision to actually do (at least) some really interesting experiments. If you don't mind I would like to suggest another similar switches to put an eye on, like Fujitsu magnetic reed tactile, HTK and ProWorld MX. It's clear that the leaf is the core of an Alps switch, followed by the switchplate, which may be the only problem here as it affects the feel too but you want to remove it. If I were you, my first attemt would be fuse both the switchplate shape and the click leaf and use its movement to close and open an optical sensor. An optical brown Alps would be the easiest first approach.
Said that, CAD might be scary, but for something like a switch, with no organic shapes at all, you can easily develop decent skills. I personally use Solidworks, so if you go with the same route just tell me and I'll give you some help.

Yes. The leaves are a true concerns about the Alps switchfeel. That's why I have to perform a lot of testing, and comparing.

Polecat: What I imagined is all based on SMD components, and a simple PCB. Idea: Check the reflections of the slider. Anyway... For now, I have to stop thinking about it, and focus on my studies. I'm behind scheulde quite a bit, and I don't want to fail.

Anyway... I put the gus into the WTS thread, and hope someone will get it.

Ive been working on some skcm models, shoot me a dm and we can discuss

PM sent.