Deep thoughts on Hall Effect switches

[wendell]

I was thinking about Hall Effect switches and started wondering why anyone would want them to be linear. I mean, think of where they are used, in military craft, medical devices, and really dangerous machinery, things that could get you killed from a typo. I would think people would want tactile feedback, and not just the puny "click" of a Model M, but a resounding "crack" like you broke something.

And what is it with this obsession with umpteen billion keystrokes? Nobody is ever going to write a novel on those keyboards. They just get used for occasionally entering short commands, sometimes under very distracting conditions, like getting shot at or having people scream at you.

[CommunistWitchDr]

Let's say the Soviets have launched their first wave of missiles. It takes out all automatic retaliation, without being caught by the computers of the era, though civilian casualties are still small. If we can take out the rest of their nuclear weapons now we stand a chance. You go to one of the missile control computers to send back whatever you can, and as you start to enter the launch code you realize the "3" doesn't work. The second wave destroys us.

It's not because the keyboard is used a lot that it has to be stupidly durable, it's because of the possible results of failure.

[oTurtlez]

I was thinking about Hall Effect switches and started wondering why anyone would want them to be linear. I mean, think of where they are used, in military craft, medical devices, and really dangerous machinery, things that could get you killed from a typo. I would think people would want tactile feedback, and not just the puny "click" of a Model M, but a resounding "crack" like you broke something.

And what is it with this obsession with umpteen billion keystrokes? Nobody is ever going to write a novel on those keyboards. They just get used for occasionally entering short commands, sometimes under very distracting conditions, like getting shot at or having people scream at you.

They're built to last, which is why they can be used for so many keypresses. They're durable as all hell, and they need to be. Something that important NEEDS to always work and always work perfectly.

As for linear, I agree to some point. A tactile or clicky design would be better, but nothing severe. And they take a decent amount of force to bottom out so that's another deterrent. Then again, the general design of the switch might not allow for a tactile or clicky mechanism to be implemented.

[davkol]

At least industrial keyboards with linear switches have speakers to beet when key press is activated. Why not the same thing in the military?

Tactile mechanism is another thing that might break. It's pretty hard to get a stuck switch if it's linear, tactile ones on the other hand...

[oTurtlez]

At least industrial keyboards with linear switches have speakers to beet when key press is activated. Why not the same thing in the military?

Tactile mechanism is another thing that might break. It's pretty hard to get a stuck switch if it's linear, tactile ones on the other hand...

Imagine a hall effect with the grainyness of a brown... *shivers*

[Daniel Beardsmore]

Springs are linear. All keyswitches were linear back in the day, with the exception of IBM. There are some who still maintain that the buckling spring is the only way to get tactility, i.e. a tactile mechanism with no friction. (Though my Model M is pretty scrapy — you still have a slider/stem, after all.) Buckling spring doesn't lend itself all that well to fitting magnetic sensing in, though.

Tactility is a very difficult objective. Alps Electric's common tactile switches are pretty dreadful, and Cherry never really pulled anything off either (though I find MX brown to be tactile enough, but just too light). SMK second generation switches are nice and tactile, but they're made of shrapnel and baling wire inside. (You also contemplate suicide when trying to reassemble them.)

It's probably easiest for anyone making ultra-robust switches to simply leave out all this weird mumbo jumbo of trying to get a tactile bump into a switch that is inherently linear. (And frankly, membrane buckling spring isn't particularly tactile anyway.)

Besides, if our only protection against nuclear winter and mass extinction is a computer system that is one typo away from global thermonuclear war, I think we're doomed.

[Wall Street]

This is pretty silly.  It would seem that the benefits of hall effect switches center around the fact that the sensor doesn't need electrical contact and hence you can minimize friction within the device.  If you and going to have two things hit/rub each other at the activation point, may as well just use a leaf or capacitive switch.  The problem is that the hall effect itself produces no mechanical feedback, so you find yourself adding friction back into the system which is counterproductive from a longevity standpoint.

[wendell]

What I'd really like to know is whether linear switches became the norm in critical situations because user testing showed they were best or because it just made life easier for the engineers.

[HaaTa]

Check the wiki on Honeywell hall effect, a tactile version was discovered by dorkvader not too long ago. Mind, it's ultra rare.

[dorkvader]

I noticed tactile hall effects were added to the wiki, as well as that honeywell pdf I found.

as far as linear hall effect switches, they are designed for reliability above all. A linear system won't wear out as fast as a tactile one (one of my tactile switches is now a linear, but still works. The tactile mechanism is broken, but the switch actuates fine) They are rated for 30,000,000,000 keypresses, though in reality, they'll become "gritty" long before that. Vintage hall effect switches have sliders that are probably ABS and switch housings that are a much harder polymer. In well sued vintage hall effect keyboards, you'll get some grittiness as the shaft has wear marks from off-axis hits. some lube would fix that though, and the switch still works.

From what I've seen, much of the very important actions were preformed with covered toggle switches and those fancy switches that require keys, which are very tactile. Hall effect keyboards may have been used to input some commands, but as far as I know, the missile targets were pre-programmed. A different scenario occurs on a miltary vehicle, where a hall effect keyboard might be used to program in a new target's coordinates. My general dynamics rubberdome with NKRO keyboard was part of a land vehicle called a SBCT, in a system called ABCS.


I tried to spread the 30 billion keystrokes number around, and I don't think other's have perpetuated it much. I was impressed with the number: like much of the gear form the time, and especially millitary equiptment, it's high reliability. It's a very impressive number!

Speaking of which, I don't think the fourth pin was used for redundancy in many designs. Form the PCB's I've looked at, pin2 is mainly wired horizontally, and pin3 is wired vertically. in a sort of odd offset matrix.

If I'm worried about missile attacks, I'll bottomout just to make sure.

I think linear switches are sort-of the norm because they are easier to make reliable (less parts to fail (except futuba complicated linears) ) and have been in production for quite a while. I mean, after typewriters, most non IBM switches were linear for quite a while.

[Krogenar]

And what is it with this obsession with umpteen billion keystrokes? Nobody is ever going to write a novel on those keyboards. They just get used for occasionally entering short commands, sometimes under very distracting conditions, like getting shot at or having people scream at you.

And yet, no one cares if the military spends $2 million dollars on a single toilet seat.