deskthority

Are there any other key switches besides buckling spring (and beam spring, from what I can discern looking at the figures and force diagrams) that actuate at the exact same moment one hears the click?

I would say that the clicky Cherry MX switches do, actually. It is the actuation mechanism that propels the inner slider to make the sound when it hits the bottom.

Yup. I've seen some switches that click only on release (unlike buckling spring which clicks on activation *and* release) but they are the odd ones out.

I'd love to see someone try to out react any theoretical delay in MX's click mechanism. It's not as direct as IBM's, but nor is it as indirect as you think!

I've just tried to discern a difference between the moment the switch clicks and the moment it actuates, by pressing down really slowly. I tried Cherry MX blue and blue Monterey switches.

With MX blue, there was no discernible difference. I wasn't able to make the switch actuate without clicking at the same time, or click without actuating.

With Monterey switches, I could make the switch click without actuating, but I had to be unrealistically slow to do that.

So, a keyboard based on the Cherry Green MX switch should be fairly similar to the buckling spring for actuation and sound?

Broadly similar, but quite different in many ways. Buckling spring is super smooth, and suddenly goes SNAP at the actuation point. (Then pings a lot.) Clicky MX switches (blue and green) have a tactile mechanism in them that feels crunchier as you trigger the click. Also, MX clicks are higher pitched.

Both definitely clicky though. If you want everyone to know you're typing, either will do!

Paralel wrote:So, a keyboard based on the Cherry Green MX switch should be fairly similar to the buckling spring for actuation and sound?

As a buckling spring aficionado, I actually find the Cherry clears more to my liking. There is no audible click as in the green switches, but the magnitude of the force bump is greater. I believe the greens are simply blues with a stronger spring; the piece of plastic that provides the tactile bump is the same.

Muirium wrote:...Buckling spring is super smooth... Also, MX clicks are higher pitched...

I'm glad you mentioned those two things as they are major factors for me. The nice low tone of my Model M is much more to my liking that the higher pitch of most other models, including the Model F.


I guess the Model M really doesn't have any substitutes.

If that's what you like, then you're probably right. I love Model F and Model M, for me the F has the better sound and feel, but the M has the better stabs (hence my Kishsaver conversion project). I merely like MX green / blue. Clicky Alps can be interesting, though. Lower pitched than MX, that's for sure. And I'm looking forward to trying a board of clicky space invaders too.

Depends how much thrill you find in the chase, of course.

My ducky with XM ALPS actuates on click.

At least, it does on the good switches.

It sort of works backwards with MX blue/green-- if you work it right, you can make it acutate without a click-- press once, let it rise up enough to release again, but you'll still be below where the click mechanism resets and so you can re-activate without another click.

With Space Invaders, the click mechanism is literally bolted onto the side of the switch; any relationship between the sound and switch closure will be coincidental at best.

What about Matias switches?

Alps Platespring actuates when it clicks, but it's basically a simplified beamspring that doesn't feel quite as nice.

My mind, it is spinning!

Alps are nice, but they're no beam spring. Hell, even buckling spring is no beam spring!

Beam spring clicks at the same point of actuation. Buckling spring too as mentioned above.

My wife has a MX Blue (a Leopold numpadless thing) and I can't handle the disconnect between actuation and click. It's even worse than the scratchiness of the movement. I find I miss a lot of keys as I feel the bump then lift off. It makes the keyboard feel `slow' and unresponsive.

The beamspring is nice (I keep shelving my Model F PC-AT and 122-key terminal even though I should be testing the controllers) as it's very `weighted', so you get a nice swing, feel the actuation, release as smooth as butter. Very good for getting a rhythm going. Even better with the solenoid click, as you get a shudder through the whole keyboard when the key actuates—this is closing the loop with software inside the keyboard if you like.

I'm a control systems guy in my day job, and to give a metaphor in that language I find I can't type `open loop', and trying to `close the loop' where your keyboard has a non-linear transfer curve—or too much backlash within that loop—throws me off. I either have to turn my `typing gain' up too high or I have to have soft typing performance. Actuation at same point as noise/tactile feel is absolutely key to get a really good control loop between keyboard and brain.

That's exactly the thing with beam spring: it swings like weighted piano keys. My desire to have a fancy weighted Roland keyboard, and a beam spring, is one of those things that keeps my wallet such an open loop…

Halvar wrote:I've just tried to discern a difference between the moment the switch clicks and the moment it actuates, by pressing down really slowly. I tried Cherry MX blue and blue Monterey switches.

With MX blue, there was no discernible difference. I wasn't able to make the switch actuate without clicking at the same time, or click without actuating.

step 1: actuate the switch
step 2: release just a little but not the full length
step 3: press again
result: actuation without click. I actually use this "feature" often to press a key repeatedly (such as the arrow keys).

xwhatsit wrote:Beam spring clicks at the same point of actuation. Buckling spring too as mentioned above.

My wife has a MX Blue (a Leopold numpadless thing) and I can't handle the disconnect between actuation and click. It's even worse than the scratchiness of the movement. I find I miss a lot of keys as I feel the bump then lift off. It makes the keyboard feel `slow' and unresponsive.

The beamspring is nice (I keep shelving my Model F PC-AT and 122-key terminal even though I should be testing the controllers) as it's very `weighted', so you get a nice swing, feel the actuation, release as smooth as butter. Very good for getting a rhythm going. Even better with the solenoid click, as you get a shudder through the whole keyboard when the key actuates—this is closing the loop with software inside the keyboard if you like.

I'm a control systems guy in my day job, and to give a metaphor in that language I find I can't type `open loop', and trying to `close the loop' where your keyboard has a non-linear transfer curve—or too much backlash within that loop—throws me off. I either have to turn my `typing gain' up too high or I have to have soft typing performance. Actuation at same point as noise/tactile feel is absolutely key to get a really good control loop between keyboard and brain.

Yes, I get the same impression when using a Cherry mx keyboard (I have a Filco with blues and a CM with greens). Although when I test the actuation point it seems that the actuation coincides with the click, I tend to bottom out with every stroke, so that it feels as if there is a disconnect between the tactile feel and actuation. I also don't like the "scratchy" feeling of Cherry mx switches.

I have not yet tried a beam spring, but to me the keys on my Topre-switch boards (HHKB Pro 2 and Leopold FC660C) feel like weighted piano keys. If a beam spring is even more so, I should like it very much.

Thus far, my favorite switches are IBM capacitive buckling spring, as in my XT, and Topre 45g, as in my HHKB and FC660C (I think I would probably prefer silenced 55g Topres, but I have not yet tried these). I also like IBM membrane buckling springs, as in my SSK, but I prefer the crisp metallic sound from the XT. All of these switches provide the right aural and tactile feedback to enable me to achieve good speed and accuracy. I don't get such a favorable feedback loop when using a Cherry mx board.

The one beam spring I have tried, several occasions now, feels like a metal, clicky Topre. So you probably want to pick one up! It's my favourite key action.

xwhatsit wrote:Beam spring clicks at the same point of actuation. Buckling spring too as mentioned above.

My wife has a MX Blue (a Leopold numpadless thing) and I can't handle the disconnect between actuation and click. It's even worse than the scratchiness of the movement. I find I miss a lot of keys as I feel the bump then lift off. It makes the keyboard feel `slow' and unresponsive.

The beamspring is nice (I keep shelving my Model F PC-AT and 122-key terminal even though I should be testing the controllers) as it's very `weighted', so you get a nice swing, feel the actuation, release as smooth as butter. Very good for getting a rhythm going. Even better with the solenoid click, as you get a shudder through the whole keyboard when the key actuates—this is closing the loop with software inside the keyboard if you like.

I'm a control systems guy in my day job, and to give a metaphor in that language I find I can't type `open loop', and trying to `close the loop' where your keyboard has a non-linear transfer curve—or too much backlash within that loop—throws me off. I either have to turn my `typing gain' up too high or I have to have soft typing performance. Actuation at same point as noise/tactile feel is absolutely key to get a really good control loop between keyboard and brain.

I feel the same exact way about click/actuation alignment. I guess the only switch types that have the kind of click/actuation synchronicity I'm looking for are beam spring and buckling spring.


What is the "solenoid click" phenomenon you mention in conjunction with beam spring keyboards? It is some kind of solenoid action in addition to the inherent click of the beam spring switch? Why was this included? To make it sound/feel more like a different type of keyboard, like a typewriter or something?

Paralel wrote: What is the "solenoid click" phenomenon you mention in conjunction with beam spring keyboards? It is some kind of solenoid action in addition to the inherent click of the beam spring switch? Why was this included? To make it sound/feel more like a different type of keyboard, like a typewriter or something?

Check out xwhatsit's other threads here and here for a full explanation. Basically, it's another level of impulse feedback which you can hear and also feel in the body of the keyboard, much like the feedback provided by electric typewriters and keypunch machines, the other typing interfaces most early terminal users were familiar with. When a keypunch machine fired, there was absolutely no doubt that the key press registered.

Yes, the solenoid is a noise maker and shaker that some (or all?) beam springs had to give even more feedback than their inherent click. I assume they were there to sound and feel reassuring to IBM Selectric users migrating to their first "cold" computer keyboards. Xwhatsit drives them with his beam spring / Model F controller, which now lies at the heart of my most prized keyboard! Though my Kishsaver only has a beeper instead.

That sounds quite nice. These beam springs are probably fairly difficult/expensive to come by?

Here's one:

http://www.ebay.com/itm/IBM-Displaywrit ... 4ad5cf411f

They can go for over $100 quite easily, and the "great white" Displaywriter model in particular is very, very large! Here's the one I've seen in person:



That's an AT Model F on the left for comparison! IBM wasn't kidding when they marketed those as "low profile".

Wow. That is a HUGE keyboard... and I thought my Model M was a monster...

It's easy to see why they switched to buckling spring capacitive.

I think the Displaywriter had a speaker instead of a solenoid, and I believe some other beam springs may have been solenoid-less. For one, the Displaywriter (and some other Beam Springs) had a plastic case, so the solenoid striking off the side wouldn't achieve the right effect.

The idea with the solenoid (which could be turned on and off as required) is that it strikes the side of the case when the terminal (or perhaps the terminal controller) has registered the key that has been pressed by the user. As such, the solenoid strike did not occur straight after the key was pressed, the key press signal had to go to the terminal, and then the terminal would send back a signal to click the solenoid.

While it could have some role in diagnostics if there was a communication failure, I believe the main purpose of it was to do with IBM's form-based user interfaces, which ignore user input if the cursor is positioned on something other than a field, or if the system is busy processing the contents of the form, etc. In those situations, a little X would appear at the bottom of the screen to signify that input was not allowed, but I could imagine that the sudden absence of the solenoid click would be a more useful indicator to a fast typist.

As for why they used solenoids when everyone else (including themselves later on) used beep speakers - I think, as described by others, that IBM felt it would be a more 'natural' noise to people trained on keypunches and teletype machines.

Indeed. We forget with our fancy personal computers how different things were with timesharing and generally slower hardware, back in the beam spring days. Typists were already fast, but the systems weren't. Closing a feedback loop was especially important. Even if it was but a little bipper like my Kishsaver.

Which reminds me, I should record it again with the wee bugger beeping away. And I'd like to hear a solenoid comparison from one of the revived beam springs!