Maglev keyswitches...

[Snarfangel]

This is a rather off-the-wall question, but since some people like linear switches, and neodymium magnets are ridiculously powerful for their size, does anyone know of attempts to make keyswitches using magnetic repulsion between two north (or south) poles of a magnet? Your keys would be floating lightly above the keyboard (perhaps held in place to keep it from flying off), rather than sitting on a spring or rubber dome. It would be one less mechanical thing to go wrong (though you would be trading it for a couple of magnets per switch that could go equally wrong).

Not sure if it could be made with Hall effect or reed switches, or if it would cause the keyboard to stick to paper clips or refrigerators, either. There are probably other difficulties involved, too, but it would be fun to try one out if they existed.

[Melvang]

If you want to try this with MX switches .125 or 1/8" diameter magnets would fit inside the spring.  I actually suggested this to Mkawa a couple days ago and while physically it should work, there is a very real possibility of inducing voltage through the matrix.  This would be on the principle that passing a magnetic field across a wire you will induce an electric current.  While I feel that the progressive "spring rate" would feel real nice I don't think it would be electronically viable. 

If you do want to try this I would suggest n52 neodymium magnets.  This grade has the highest magnetic field for standard available magnets, the magnetic force degrades the least over time including physical shock and heat, but it would also pose the greatest risk of induced electric current.

[fohat.digs]

I am not a physicist or electrical engineer, but wouldn't all that magnetism swarming (because it would be a swarm of discreet fields, no?) over a small area be an electronic nightmare?

On the other hand, some people like magnets for arthritis, so you could get your therapy as you typed!

[Melvang]

I am not a physicist or electrical engineer, but wouldn't all that magnetism swarming (because it would be a swarm of discreet fields, no?) over a small area be an electronic nightmare?

On the other hand, some people like magnets for arthritis, so you could get your therapy as you typed!

Yes and no.  Static magnetic fields *should not* induce electrical currents, however when you put them into motion (actual typing) I feel it is possible.  Again, I am in the same boat as fohat in that I am neither a physicist nor an electrical engineer but I do have a gravitation towards physics and mechanical engineering.

[Daniel Beardsmore]

Here's something with magnets in:

http://deskthority.net/photos-videos-f8/univac-f-1355-00-t6489.html

Not what you had in mind, but still, you're not the only person to consider using magnets in that sort of way.

[Snarfangel]

Here's something with magnets in:

http://deskthority.net/photos-videos-f8/univac-f-1355-00-t6489.html

Not what you had in mind, but still, you're not the only person to consider using magnets in that sort of way.

That is awesome! Thanks for the link.

[thesentinel]

i'm sorry

[berserkfan]

I am not a physicist or electrical engineer, but wouldn't all that magnetism swarming (because it would be a swarm of discreet fields, no?) over a small area be an electronic nightmare?

On the other hand, some people like magnets for arthritis, so you could get your therapy as you typed!

How can this be good for your other electronics peripherals? As a perpetual klutz I am extremely sensitive to the possibility that I screw up my computer doing something silly. One big electromagnet sounds like a hard disk killer to me - and I sometimes put hard disks on my desk.

[Melvang]

I am not a physicist or electrical engineer, but wouldn't all that magnetism swarming (because it would be a swarm of discreet fields, no?) over a small area be an electronic nightmare?

On the other hand, some people like magnets for arthritis, so you could get your therapy as you typed!

How can this be good for your other electronics peripherals? As a perpetual klutz I am extremely sensitive to the possibility that I screw up my computer doing something silly. One big electromagnet sounds like a hard disk killer to me - and I sometimes put hard disks on my desk.

http://www.techhive.com/article/116572/article.html

First we have not suggested using electromagnets.  The ones I suggested are permanent n52 neodymium rare earth magnets.  For the most part you will not find a magnet in your house strong enough to kill a hard drive.  The one exception I could possibly think of would be a magnet from a high end 12"+ speaker.  The magnetic fields we would be talking about here would only extend a couple inches at most.  Besides the magnet that actually moves the head in a platter type hard drive are usually these n52 neodymium magnets anyway.  Reason being, reducing the power to this mechinism is actually the best way to reduce the power draw of the hard drive.  Also increasing the acceleration rates in this head both in speeding up and slowing down (slowing down is technically considered acceleration) is the easiest way to increase random access times. 

Magnets also have zero effect on flash based storage which would include thumb drives, memory cards, and solid state hard drives.  The reason for this is because the data is not written using magnetics. 

With respect to credit cards, I have a wallet with a money clip that is of the magnetic type and all my cards still work.  Well with the exception of having worked in a foundry for 22 months doing a multi-million dollar install project.  But this was because of the dust (both metal and mold casting sand).  The reason I say it was from the environment and not the wallet is because I had already had the wallet for 5 years or so with not one card failure.

[Kamen Rider Blade]

Opposing magnets would make very neat key switches IMO. But I wonder how much would they cost?

[Melvang]

Opposing magnets would make very neat key switches IMO. But I wonder how much would they cost?

These are 1/8" diameter (which should fit inside an MX spring) and 1/16" thick, pulling force against another of the same magnet is 1.04lbs.
https://www.kjmagnetics.com/proddetail.asp?prod=D21B-N52

These are 1/32" thick and rated for the same amount of pull
https://www.kjmagnetics.com/proddetail.asp?prod=D201-N52&cat=168

Keep in mind that dropping the grade down from the n52 will substantially reduce the price but will also drop the magnetic force.

Looked again and not that much of a drop in price.

[Kamen Rider Blade]

Those are expensive

You need $1 per magnet, meaning you need 2 for a switch.

Ouch.

[Oobly]

Will definitely feel different to springs. Springs have an almost constant increase in resistance, whereas opposed magnets have an increasing rate of increase of resistance (square law). You can tune this "rate of increase" by adjusting the distance and magnet strength (stronger magnet at further distance = same initial force, but slower increase in force).

It may be hard to get a set of magnets with consistent enough strength between them and it'll take some experimentation to get the right distance and magnet strength, but it's possible. Not sure I'd like the feeling, though. I like the linear increase of springs. It could feel really odd, with a soft initial press and rapid increase in resistance. Kind of organic...

You'd need around 2g (stem and keycap mass) + desired initial press strength at the top and will have to have some kind of retention system for the caps. An MX slider and case may work well. Magnets could be mounted somewhere on the slider and under the case bottom. If the magnet is 2mm or smaller and less than 0.8mm thick it will fit on the pole that goes into the tube in the lower case. I suspect it won't be strong enough at that size, though, unless the lower magnet is really strong. You could also try a thicker magnet (or even a cylindrical one) and trim the end of the pole to match.

[Melvang]

Those are expensive

You need $1 per magnet, meaning you need 2 for a switch.

Ouch.

Those prices were for a pack of 10 magnets.

Will definitely feel different to springs. Springs have an almost constant increase in resistance, whereas opposed magnets have an increasing rate of increase of resistance (square law). You can tune this "rate of increase" by adjusting the distance and magnet strength (stronger magnet at further distance = same initial force, but slower increase in force).

It may be hard to get a set of magnets with consistent enough strength between them and it'll take some experimentation to get the right distance and magnet strength, but it's possible. Not sure I'd like the feeling, though. I like the linear increase of springs. It could feel really odd, with a soft initial press and rapid increase in resistance. Kind of organic...

You'd need around 2g (stem and keycap mass) + desired initial press strength at the top and will have to have some kind of retention system for the caps. An MX slider and case may work well. Magnets could be mounted somewhere on the slider and under the case bottom. If the magnet is 2mm or smaller and less than 0.8mm thick it will fit on the pole that goes into the tube in the lower case. I suspect it won't be strong enough at that size, though, unless the lower magnet is really strong. You could also try a thicker magnet (or even a cylindrical one) and trim the end of the pole to match.

While I don't have any issues with the usual lot of Cherry springs I would love to try some very progressive rate springs.  The magnets I posted should fit inside an MX spring.  If I were to try it I would start with some linear switches and a light spring. 

The n52 grade for neodymium magnets are very strong.  At this size it would be very difficult to separate them.

Just to give an idea on how strong these things are, Mythbusters actually successfully altered a bullet trajectory a significant amount using these.  Granted they had to use 9 magnets to do it.  The magnets they used were 2 inches across and 6 of them were 6" tall and I don't remember how tall the other 3 were but they were a different size.  They were using a 9mm pistol for the test but they accomplished it.

[Kamen Rider Blade]

Those are expensive

You need $1 per magnet, meaning you need 2 for a switch.

Ouch.

Those prices were for a pack of 10 magnets.

That's far more reasonable then.

[HaaTa]

Haha, nice.
I have a few "magnetic" keyboard switch projects I'm working on. Even did some magnetics research near the end of my undergrad (mostly to get comfortable with finite element modelling).

Using opposing magnets for simple levitation is very doable (and if using rare earth magnets could be made quite small).
http://www.kjmagnetics.com/calculator.repel.asp is a very nice "simple" force calculator for this.
Keep in mind that magnetic force increases exponentially with the distance for levitation.
The next thing you have to think about is how the actuation will take place (and at what point during the press). One option is a hall effect sensor (which, tbh in this case is a positioning nightmare; you have to understand exactly what the fields look like, and how they change when using more than one magnetic field source).
Another, more reasonable option is to use a dome switch. This works by having the lower magnet on top of the dome and once enough force is applied it will actuate.

Anything more than this, you start to need serious EE and/or ME background. EE for manipulating the fields, and ME (Mechanical Engineering) for instrumenting a tactile/click sensation.
Electromagnets bring in a whole 'nother world of hurt 

I'm working on reproducing the Univac switches linked right now. Should have an experimental prototype switch ready in the next few weeks (once my ski-cation is finished ).
Bug me on IRC if you want to discuss.

[niubio]

I guess I'll add my 2 cents here...

I was thinking about reed switch boards I had posted about a month ago from Zbrojovka Brno. They use a pretty simple linear mechanism - a single spring and a reed element soldered on a small "elevated" plate. It is activated by a slider-mounted magnet. I guess it would be possible to combine this with maglev concept (both ideas utilize magnets) thus obrtaining an ultimate solution Electronics in a keyboard like this would be 100% isolated from the above, thus making it 100% spill/dust resistant. Here's my quick MS paint job:



It's a very simplified schematic without paying much attention to small details... And there should be "isolation" instead of "insulation", but I think you get the idea After the cap is pressed, the magnet descends down and activates the reed switch. You could use different magnets to obtain different forces, and maybe even add something to achieve tactility. You can use different "caps-latch" mechanism (I've copied the "space invader" more or less in my gif).

[Daniel Beardsmore]

I imagine that would feel really weird!

(Hm, when did MS Paint get anti-aliased graphics? Was that Windows 8, or an earlier version?)

[niubio]

I've added caption + leading lines in PS Good eye for details!

[mougrim]

I guess I'll add my 2 cents here...

I was thinking about reed switch boards I had posted about a month ago from Zbrojovka Brno. They use a pretty simple linear mechanism - a single spring and a reed element soldered on a small "elevated" plate. It is activated by a slider-mounted magnet. In Unitra-Polam switch they've put a magnet... inside the cap. I guess it would be possible to combine this with maglev concept (both ideas utilize magnets) thus obrtaining an ultimate solution Electronics in a keyboard like this would be 100% isolated from the above, thus making it 100% spill/dust resistant. Here's my quick MS paint job:

Show Image

It's a very simplified schematic without paying much attention to small details... And there should be "isolation" instead of "insulation", but I think you get the idea ;) After the cap is pressed, the magnet descends down and activates the reed switch. You could use different magnets to obtain different forces, and maybe even add something to achieve tactility. You can use different "caps-latch" mechanism (I've copied the "space invader" more or less in my gif).

Dat cool. But calibration of this thing will be a pain in the ass - you need reed switches activated only when magnet directly beside it...

Still - I'd love to see it. It could be reed switches keyboard re-imagined all through.

[niubio]

Hmmm... Why mougrim? I mean in Zbrojovka boards and in MK45 / Elwro Junior they activate after a magnet reaches a certain proximity point in space, and they "release" after said magnet leaves this space - and only after that they are ready to register another "key press". You don't have to be that precise I guess...?

[blackbox]

Very interesting concept niubiu. It would make ip44/47 certified mechanical keyboards a lot easier to make.

[mougrim]

Hmmm... Why mougrim? I mean in Zbrojovka boards and in MK45 / Elwro Junior they activate after a magnet reaches a certain proximity point in space, and they "release" after said magnet leaves this space - and only after that they are ready to register another "key press". You don't have to be that precise I guess...?

Yeah, but in those boards you don't have second magnet at the bottom of switch. But again, maybe it really doesn't matter.

[Melvang]

As far as experimenting with this I was thinking about using magnets inside of cherry switches.  From the measurements that I got from mkawa 1/8" diameter magnets should fit inside the spring.

[SpAmRaY]

As far as experimenting with this I was thinking about using magnets inside of cherry switches.  From the measurements that I got from mkawa 1/8" diameter magnets should fit inside the spring.

I just happened to be looking for some magnets for something figured I'd share a link (no idea if these costs are reasonable etc, just a site I was browsing)

http://www.kjmagnetics.com/products.asp?cat=13

[HaaTa]

So rather than calibration, I think there might be a problem with the magnetic field direction in respect to the reed switch.

I was just going to explain it, but since I do finite element stuff anyways...might as well show you some actual simulations of the magnetic fields between two equal strength magnets.
These are 2D slices through the center of 2 cylinder rare earth magnets setup for levitation (opposing North and South poles).

Close


Far


For either a hall effect sensor or reed switch you are using a change in the magnetic field strength (reed switches use a much stronger change). As shown above, the magnetic field around where you want to place the reed switch doesn't really change at all... Yes the magnitude changes, but overall, there is a magnetic field already going in the direction you want whether the magnets are close or far (so the reed switch will always be closed ).

Using a reed switch is much easier overall to implement (I'm familiar with the hall effect issues though), so keep trying 

[mougrim]

So rather than calibration, I think there might be a problem with the magnetic field direction in respect to the reed switch.

I was just going to explain it, but since I do finite element stuff anyways...might as well show you some actual simulations of the magnetic fields between two equal strength magnets.
These are 2D slices through the center of 2 cylinder rare earth magnets setup for levitation (opposing North and South poles).

Close

Show Image

Far

Show Image

For either a hall effect sensor or reed switch you are using a change in the magnetic field strength (reed switches use a much stronger change). As shown above, the magnetic field around where you want to place the reed switch doesn't really change at all... Yes the magnitude changes, but overall, there is a magnetic field already going in the direction you want whether the magnets are close or far (so the reed switch will always be closed ).

Using a reed switch is much easier overall to implement (I'm familiar with the hall effect issues though), so keep trying 

And if you place reed switch a little more up?

[HaaTa]

May be sufficient, but it's highly dependent upon the strength of the magnets and peak magnetic field required for the reed switch. Switching out magnets will likely be hard.

[mougrim]

May be sufficient, but it's highly dependent upon the strength of the magnets and peak magnetic field required for the reed switch. Switching out magnets will likely be hard.

Yeah. I see need for a serious modeling here.

[niubio]

Thanks for the feedback guys! Great pictures HaaTa! I have a quick question: what is the distance scale on them? I mean if you could draw 1mm distance on it, that would make a good point of reference I guess it isn't that much of a problem, because you can always move the reed switch in horizotal space wherever you want to to suit your needs. It can be even placed behind a very thin plastic wall, so when the magnet descends, it would almost "touch" the reed... Thanks  to your explanation, now I understand why you can hear metallic "click" sound on some Unitra-Polam switches! That's because some of the magntes are indeed touching reed elements! A good way to be sure that it qould register

Now I'm ready to make new concept sketch, this time with actual dimensions. Gosh, now I wish I would have a caliper! That way it would be easy to measure reed switch dimensions or plastic thickness for example I wonder which kind of magnets would give a "nice" kind of feel... Now I'm thinking where I could find ANY magnets and experiment a little with some rubber dome switches - you know, the cheapest ones, I think they are best suited for experimenting thanks to the very simple design!

[mougrim]

Thanks for the feedback guys! Great pictures HaaTa! I have a quick question: what is the distance scale on them? I mean if you could draw 1mm distance on it, that would make a good point of reference I guess it isn't that much of a problem, because you can always move the reed switch in horizotal space wherever you want to to suit your needs. It can be even placed behind a very thin plastic wall, so when the magnet descends, it would almost "touch" the reed... Thanks  to your explanation, now I understand why you can hear metallic "click" sound on some Unitra-Polam switches! That's because some of the magntes are indeed touching reed elements! A good way to be sure that it qould register

Now I'm ready to make new concept sketch, this time with actual dimensions. Gosh, now I wish I would have a caliper! That way it would be easy to measure reed switch dimensions or plastic thickness for example I wonder which kind of magnets would give a "nice" kind of feel... Now I'm thinking where I could find ANY magnets and experiment a little with some rubber dome switches - you know, the cheapest ones, I think they are best suited for experimenting thanks to the very simple design!

Good luck to you When you'll become a greatest keyboard designer of the world, just give us 1% of the shares

You have damn fine ideas.

[Kamen Rider Blade]

Maybe you should apply for a patent on this type of switch.

If it's different enough from the Halls Effect, it should be able to get it's own patent, and you can create a new superior switch.

[niubio]

Good luck to you When you'll become a greatest keyboard designer of the world, just give us 1% of the shares

Are you crazy? I'm faaaar from being a "designer", I'm not even a keyboard power user. That's just a wild idea - nothing more. Thanks anyway

Maybe you should apply for a patent on this type of switch.

Mhmmm... Maybe instead of applying for a patent, which is way to expensive for me right now, I could open source it - this way it's going to be impossible for anyone to steal it and - which I'm more afraid of - to shelve it, as often the story goes with many new cool ideas...

Anyway, I've just purchased 110 neodynium magnets online - the cheapest ones They'll arrive the day after tomorrow I guess, I'll install them in one of my rubber domes in a simplified fashion - two magnets would simply "squeeze" the membrane beneath, or sth like that. If I have any success with it, I'll post some photos and a video. Then, if it makes any sense at all, I'm thinking about sending this concept board to fellow geekhackers, if anyone's interested in trying maglev

[peterstock]

Very cool idea. I think the idea of an x^2 rate of force would be good - a rapidly increasing force after a tactile point is what everyone seems to agree is desirable to stop bottoming out. But linear springs can't offer that, with lightness up to the activation point.

[simon_C]

now I'm thinking about prototyping this. Possibly with some brown switches.

[niubio]

It's been a very busy week... Got a short news update:

1) 240 neodynium magnets arrived
2) I've purchased 10 reed switches online to prototype my idea, they haven't arrived yet.
3) 5 minutes ago I've finished installing one pair of magnets in a beaten up "Turbo Plus" rubber dome from KME.
4) Two words: IT WORKS.
5) Switches are 100% contact-less, without the need of any reeds.
6) I've to clean this board and install all magnets, so expect some pictures + a video tomorrow at best. Maybe I'll run into some new problems... It's not that easy :/
7) It's not a "magnet touches reed element" sound that I was hearing in my MK45 and Elwro - it's the sound of the reed element itself, when it's getting into a close proximity to a magnet. I was a total idiot and had no idea what I was talking about.
8 ) I've tried to search something up on those vintage magnetic switches... Now I know A LITTLE more I think... Zbrojovka Brno-s utilize Honeywell HAL effect clones ("a holy grail of keyboards" ), MK45 and Elwro use reed switches. I know, for some of you out there is common knowledge, but it wasn't that obvious for me up until now!

Looking forward to hear from you simon_C! Let's get this on!

[Snarfangel]

It's been a very busy week... Got a short news update:

1) 240 neodynium magnets arrived
2) I've purchased 10 reed switches online to prototype my idea, they haven't arrived yet.
3) 5 minutes ago I've finished installing one pair of magnets in a beaten up "Turbo Plus" rubber dome from KME.
4) Two words: IT WORKS.
5) Switches are 100% contact-less, without the need of any reeds.
6) I've to clean this board and install all magnets, so expect some pictures + a video tomorrow at best. Maybe I'll run into some new problems... It's not that easy :/
7) It's not a "magnet touches reed element" sound that I was hearing in my MK45 and Elwro - it's the sound of the reed element itself, when it's getting into a close proximity to a magnet. I was a total idiot and had no idea what I was talking about.
8 ) I've tried to search something up on those vintage magnetic switches... Now I know A LITTLE more I think... Zbrojovka Brno-s utilize Honeywell HAL effect clones ("a holy grail of keyboards" ), MK45 and Elwro use reed switches. I know, for some of you out there is common knowledge, but it wasn't that obvious for me up until now!

Looking forward to hear from you simon_C! Let's get this on!

That sounds awesome! I can't wait for the pictures.   

[simon_C]

Hm.. Interesting.
Looking forward to more

[blackbox]

Awesome! Good that someone is trying it out. Looking forwards to see the results of the prototyping.

[mougrim]

Keep good job, dude! I'm looking forward to videos!

And yep, you right - that distinctive sound was sound of reed switch itself - as magnet even don't touch switch - it have no need to

[niubio]

I've just finished it. Haven't catched any sleep in more than 24 hours. I think I'm now obsessed with this maglev idea. So I've just made a linear "switch" board out of a simple rubber dome. I have a concept how to make a tacticle version. Anyway, it's getting dark now so no good quality photos nor a video today. I think I'm starting talking to myself, so goodnight everyone! A couple of quick photos (a teaser I guess!):

[blackbox]

Looks good! Cant wait for the results!

[simon_C]

What about just putting two magnets in an existing mechanical switch, and just using it as a spring?

[tricheboars]

post a video so i can hear it!

[Daniel Beardsmore]

"Tactile" not "tactical" :-) (I am sure Lt Reed will be happy to show you his tactical array …)

You sure it's a linear switch? It should be exponential: easy to press at the start of travel, increasing to the point that it should be virtually impossible to bottom out.

Also, don't you have a job/classes to go to? :-P

[tooki]

Great work! Really interested in seeing the final results!

So I was thinking -- ignoring the EM field nightmare it would be -- why not make one with electromagnets? It'd be expensive as f**k, but you could create adjustable keyforce keyboards, possibly even with an electronically-generated tactile point. (E.g. once the reed switch or hall effect sensor trips, reduce the magnetic field strength. Or increase it. Whatever you want.)

[Kamen Rider Blade]

Great work! Really interested in seeing the final results!

So I was thinking -- ignoring the EM field nightmare it would be -- why not make one with electromagnets? It'd be expensive as f**k, but you could create adjustable keyforce keyboards, possibly even with an electronically-generated tactile point. (E.g. once the reed switch or hall effect sensor trips, reduce the magnetic field strength. Or increase it. Whatever you want.)

I think you answered your own question.

The cost part isn't worth it, along with the other high EM field issue.

[niubio]

Hi guys! I'm back from the dead

Got some photos and a video for you - I'm uploading it right now, and I'll process pictures later today. So, let's explain it all a little

1) Yeah, it's "tactile" not "tactical" Daniel You're a purist, are you? Well, it's about two weeks I even know the meaning of the word, so don't be too picky on me!
2) The switch is not linear of course - it's expotential but it's possible to bottom out without any problem.
3) Your concept tooki is the ideal solution - you could create an universal modular version - with interchangeable caps and fine-tuned switches. You could choose with software if you want tacticle/linear/or whatever switch, it's power, etc. That would be the ultimate solution!
4) Ok, so back to my switch... One magnet sits on the top of the membrane "contact", the other can be clearly visible inside the cap. The cap levitates above the membrane. When the key is pressed, the expotential force between the magnets works pretty much like a standard rubber dome - it pushes the membrane sheets together and you got an actuation.
5) You don't have to bottom out the switch. You can assemble it in such way, that it will actuate after traveling about 1mm! I had an excelent shot of it today and lost it somewhere (pressing scroll lock and observing as LED goes on and off).
6) This idea is a pain in the ass to implement. The devil's in the detail. The trick is you have to choose the right distance between the magnets - and it all depends on their strenght. I had no choice, as I just picked the cheapest square neodynium magnets that were available online - so that leaves me only one option - tweaking the distance. The bottom magnet is just lying flat on the membrane - nothing to do here. And here's a problem - you have to choose the right distance for a magnet that is installed inside the slider. If you mount it to low, you'll have to use extreme force to press it. The switch feels VERY stiff then. And if you mount it too high inside the slider, it won't actuate.

I've used... nails as magnet "spacers" in the sliders. They were all 13mm, but unfortunately distances vary greatly inside the caps... So some keys feel stiff, some feel a little like cherry REDs. In some of them I got very lucky, and they feel as something I've never felt before - just awesome! I think that's probably the cheapest way to make yourself a gaming board!
7) Why KME Turbo Trak? It's a really ****ty board. But it's a prototype to see if this works AT ALL. It's just a proof that it can be done, and those square magnets fit KME's sliders perfectly. It's also my point, that you can take total crap and change it into something of a value.
8 ) I wouldn't recommend fitting magnets inside the mechanical switch. You don't have much option in terms of tweaking the distance, so you have to use magnets with specific strenght to achieve your "feeling" target. Also, I don't think that magnets with radius matching cherry spring would be sufficent to create enough maglev force.
9) Next time I'll pick round magnets. For the bottom (membrane mounting) you should pick magnets with 1mm thickness (matching more or less rubber dome thickness) and radius that matches "membrane contact" radius, or even 1mm smaller ones. For the slider you should pick round magnets that match slider's radius (even if it's a square one). SQUARE MAGNETS SUCKS and require much precision to mount them in the right spot on the membrane.
10) I think it's possible to change any rubber dome into a maglev board. I'll try to mod some of them (including the crappiest "flatbeds" I own) to prove my point

Enjoy!

https://www.youtube.com/watch?v=JBGiA6O4RgM

[Daniel Beardsmore]

Does the exponential curve make the switch really bouncy when you get the distance right?

Electromagnetic would be interesting. I guess there are two ways to do this. You could have all the switches powered permanently, and when you turn off the PC, all the sliders would drop, and then shoot back up when the PC is turned on.

Otherwise, I presume that you'd need a static permanent magnet, a static cancellation electromagnet, and finally a movable permanent magnet. Once you've reached the tactile point, the electromagnet would be powered up and would cancel out the static permanent magnet.

[niubio]

I guess you can call it "bouncier" than regular rubber dome or mechanical switch, but it doesn't make that much of a difference... for me at least!

Holy crap, that's just another awesome cool idea Daniel... Didn't think of it! Dammit, if it would be possible to prototype even not fully functional unit with electromagnets and reasonable material costs... I must make yet another post on polish electronics forum to ask where could I obtain cheap electromagnets.

For now I have developed an impresive backlog - have to add pictures & descriptions to my site & deskthority, send some parcels and bite a big chunk of this giant keyboard pile in my basement... I've messed up my whole apartment - magnets, screwdrivers, board elements and other stuff everywhere. This chaos has to be tamed right now.

Back to the topic... There's one another board with electromagnets I think has a lot potential  - the one that was posted by HaaTa (Univac F-1355-00). This simple non-tactile maglev-over-membrane design just doesn't apply to my taste. I would like to make something that really makes you go WOW!

Ah, almost forget, an idea for tactile version: upper round magnet descends down THROUGH another magnetic RING (slides through it). After breaching the critical point it goes down even faster - with help from gravitiy and opposite magnetic polar repell from ring above - and bottom out HARD. The problem is how to make it go back Simple answer - a spring or another - stronger - magnet on the bottom.