Buckling spring WITHOUT a membrane - is it better?

[ksd5]

Is it better to have a BS mechanism where the spring is part of the switch circuit, and there is no need for a membrane? I'll explain:

• The spring would be analogous to the top conductive layer which would normally be attached to the rubber dome.
• There would be a layer of a conducting substance on the inside of the plastic cylinder that the spring resides in. This layer of metal would be analogous to the bottom conductive layer which a rubber dome would activate.
• When the key is pressed, the spring buckles against the side of the cylinder, which would complete the circuit.

I think the advantage would be that the tactile/auditory event would be synchronized with the key signal event.

The springs and metal layer would be oxidation resistant and corrosion resistant.

[ch_123]

An interesting idea, and you're not the first person to suggest it - the original IBM buckling spring patent described such a mechanism. I have a rough description of it in this thread.

The thing was never used in a production keyboard. I think the problem with such a thing would be electrical bounce issues. The patent has some variations that would solve this, but made it comically overcomplicated. The capacitive buckling spring system used in the Model F (where the hammer creates a capacitor between itself and two electrical pads on a PCB below it, which the keyboard senses as a keypress) was a much smarter idea, at the expense of being relatively expensive. Thus why they moved to a membrane, which did pretty much the same thing, except was much cheaper to make.

In reality, the membranes in Model Ms take a lot of wear for them to fail.

[ch_123]

The PCB in the Model F is nice in that you can wipe it down with a damp cloth and it has no effect on it (assuming you dry it afterwards of course)

[clickclack]

Wear is not the problem, spills are.

This is basically how Topre does it but with a coiled spring.  The spring completes the electrical circuit.  The rubber gives the tactile feel.  The design is fairly spill proof although I  wouldn't go pushing it.

+1
that is exactly what I was thinking.
I am also very impressed that these keyboards can be as durable as they are!

[ch_123]

When I opened my HHKB, I decided to test it with a Model F spring/hammer to see what effect it had (none I'm afraid). However, when I touched the spring directly I could have sworn that I got a slightly shock off it... I don't think the Topre's contacts are sealed in the same way that the Model F's one is. Then again, it doesn't really make sense for them not to be sealed... I might repeat the experiment with a volt meter to verify.

[JBert]

As far as I have seen on ripster pictures, the Topre spring makes contact with the solder-covered traces. I think the spring gets "charged" as well as the pads beneath it, that's why you felt a shock.

Why the model F pads were sandwiched between those layers of fiberglass instead of just covering them with solder is beyond me. I would file it under typical IBM over-engineering or under a production limitation (maybe they couldn't solder-dip the curved board).

[ch_123]

Of course... The metal plate is meant to be a capacitor which stores charge... Silly me.

I think though that the capacitance in the Topre must be greater than the IBM if I was able to feel a shock when touching the spring, but never experienced that with an IBM. The patent for the Beam Spring states that the capacitance when the fly plate lays directly on top of the insulation is 4pF, which is probably quite small. Perhaps the Topre uses higher values which requires less sensitive measurements.

[quadibloc]

I think I can see why this was not used. While stainless steel is corrosion resistant, and springy, it isn't used for making contacts. Only two metals are normally used for electrical contacts: gold and tin. Other corrosion-resistant metals, like aluminum, usually work by forming a very thin and transparent - but insulating - oxide layer on their surfaces.

Iron, though, is the metal for making springs. And springs flex. So the plating on a gold-plated spring would normally end up flaking off.

This has given me an idea, though, but this idea is also unworkable. If springs are made of iron, while iron is not a good electrical contact material, iron has another useful property: it can be magnetized.

So have a magnetized spring, and when it buckles, it causes a reed switch to close, or actuates a Hall-effect detector!

However, while those types of magnetic switches have been used in keyboards, they all required the magnet to move vertically with the plunger. In a buckling-spring design, the magnetic element wouldn't move by as large a distance when the key is pressed, so it probably wouldn't be possible to make a keyswitch with reliable "closed" and "open" positions.

[Phaedrus2129]

But ferromagnets lose their magnetic properties when struck, bent, or otherwise mechanically deformed. It upsets the magnetic domains. Your magnetized spring could have a lifetime shorter than that of the worst rubber dome keyboard.

[ksd5]

But ferromagnets lose their magnetic properties when struck, bent, or otherwise mechanically deformed. It upsets the magnetic domains. Your magnetized spring could have a lifetime shorter than that of the worst rubber dome keyboard.

You could have a system that sends a brief but powerful charge through each spring at the touch of a switch on the bottom of the keyboard, or at the press of a hotkey combo.

[ch_123]

I think I can see why this was not used. While stainless steel is corrosion resistant, and springy, it isn't used for making contacts. Only two metals are normally used for electrical contacts: gold and tin. Other corrosion-resistant metals, like aluminum, usually work by forming a very thin and transparent - but insulating - oxide layer on their surfaces.

Iron, though, is the metal for making springs. And springs flex. So the plating on a gold-plated spring would normally end up flaking off.

This has given me an idea, though, but this idea is also unworkable. If springs are made of iron, while iron is not a good electrical contact material, iron has another useful property: it can be magnetized.

So have a magnetized spring, and when it buckles, it causes a reed switch to close, or actuates a Hall-effect detector!

However, while those types of magnetic switches have been used in keyboards, they all required the magnet to move vertically with the plunger. In a buckling-spring design, the magnetic element wouldn't move by as large a distance when the key is pressed, so it probably wouldn't be possible to make a keyswitch with reliable "closed" and "open" positions.

The IBM patent went on to describe a variation where the spring hit against some sort of capacitive contact which presumably sensed the distance between the spring and itself... In a manner suspiciously similar to the way of a Topre switch works.

Either way, the capacitive system used in the Model F was a much better idea.

[quadibloc]

But ferromagnets lose their magnetic properties when struck, bent, or otherwise mechanically deformed. It upsets the magnetic domains. Your magnetized spring could have a lifetime shorter than that of the worst rubber dome keyboard.

Another good reason not to try that design. While one could put a magnetized ball bearing in the middle of a spring to avoid that problem, the short distance it would move makes the idea impractical in any case.

[brkz]

I think the advantage would be that the tactile/auditory event would be synchronized with the key signal event.

In a well-adjusted model M this is the case already. When the spring buckles it turns the hammer which closes the circuit on the membranes.

Sometimes you get keys on Ms that need a bit of extra pressure after the spring buckles to register, this indicates a problem with tolerances (usually plastic rivets gone awol)

[ch_123]

Well-adjusted? Any Model M will do that, that's the whole point of the clicking and the tactility - it coincides with the actuation of the membrane.