Unicomp SSK... for real this time?

[bhtooefr]

The problem with discrete BS switches is, how BS works.

Ultimately, you're adding a lot of thickness and complexity to make a discrete BS switch that either is conductive or appears to be conductive. (Frankly, I don't trust the energized spring implementation of BS, and I think there's a reason why IBM never used it.)

Now, the Brother approach is an interesting one, because while Brother used it for membrane, THAT could be applied to a capacitive switch, and now we're back into the realm of custom PCBs (albeit ones with weird shapes drilled in them for mounting the switch).

[Muirium]

Got some info on live spring versions? Although surely a hack back in the day, we have the controllers at our disposal now to overcome a lot of issues. And saving height in a keyswitch that will already have an MX mount on top is well worth considering…

[bhtooefr]

The original BS patent (in 1971, IIRC) was for a live spring version. And it's not the controller that I see as the issue (although I do see this being a particularly bouncy switch, so the controller can compensate for that, at the expense of lag), it's the durability of the electrical paths.

http://www.google.com/patents/US3699296 is the patent, for what it's worth.

And, I wouldn't go MX mount, but that's just me. Alps mount is much better matched with the buckling spring mechanism as far as low profile goes, and it IS possible to get Alps mount caps, it'll just cost a fair amount the first go around.

[Muirium]

Nice, thanks for the link. You're doubtless right about the dynamics of a charged spring smacking into metal over the long run. There's probably some artful way to do this indirectly, without breaking from a discrete switch design.

The trouble with any other cap mount besides MX is that's where all the sweet caps are at. Even Alps just doesn't get a look in for GBs. MX is nicer to pull, too; which I consider another fair advantage having had some fights with tight Alps of my own! In fact, I think the smartest thing Matias could do with his switches is go MX mount on top.

The deck is rigged in favour of MX when it comes to caps, that's for sure!

[7bit]

...
MX is nicer to pull, too; which I consider another fair advantage having had some fights with tight Alps of my own! ...

You have to put the key puller wires left and right under the cap and wiggle sideways.*

-----------------
*) This tip I got from the number one keyboard expert on planet earth, a/k/a Ripster.

[Muirium]

Okay, so I was using my fingers. But still!

[Hypersphere]

Nice, thanks for the link. You're doubtless right about the dynamics of a charged spring smacking into metal over the long run. There's probably some artful way to do this indirectly, without breaking from a discrete switch design.

The trouble with any other cap mount besides MX is that's where all the sweet caps are at. Even Alps just doesn't get a look in for GBs. MX is nicer to pull, too; which I consider another fair advantage having had some fights with tight Alps of my own! In fact, I think the smartest thing Matias could do with his switches is go MX mount on top.

The deck is rigged in favour of MX when it comes to caps, that's for sure!

All true. However, I think the nicest design is the IBM 2-part stem and cap. All the keys on the IBM Model M are the same profile, and all single-unit caps are interchangeable. Nevertheless, custom caps for BS stems are in limited supply. Unfortunately, as you have pointed out, the greatest variety is for Cherry stems.

[Muirium]

Indeed, I do like what IBM did with uniform caps on a curved backplate (let alone the fully international barrel system) but when talking about making stuff ourselves, that curved backplate is the obstacle.

IBM caps are nice, but I'm well spoiled by my sphericals now. Really want to put them on everything! Perhaps beam spring caps are as good.

[7bit]

For Model F and Ms we need to use all row 3 (or all row 2 keys).

If you drill out the stem and gule them on the IBM stems it should also work out, somehow.


ps: I'm speaking about SP sphericals here, not genuine Cherry caps!

[Grond]

Apparently there was much more interest for Unicomp SSK when it was just a proposal/mockup/dream than now that it's going to happen. Our love for keyboards may be platonic after all.

[Muirium]

Platonic as in ideal or platonic like friends without "benefits"?

[Grond]

Both, you know. We love the idea of a Unicomp SSK, but when it comes down from the world of ideas and it turns real we don't quite like it anymore. So in the end most people who wanted this keyboard when it didn't exist won't buy it when it goes on sale.

[7bit]

The reason is that we dreamed a different dream, each.


My dream is a BS-switch module.

[Muirium]

I haven't heard anyone against the idea of a true-to-its-roots SSK remake. We all agree that'd be a great board, because the originals already are. A couple of extra mods to bring it up to the modern format, and frankly we are all set. Every other proposed improvement on the TKL is just a bonus, built on top of that solid foundation.

What got us posting back and forth all the time was the potential to really change things. Unicomp's caught that bug as well! This is a good thing. Especially if they get the message that a classic SSK remake is at the heart of what we want.

As for buckling spring modules, I was talking to Matteo about this the other day. Here's my basic idea:

Say we used regular Model M flippers and springs. And made our own prototype barrel housing for them, with MX plate mount compatible attachment points. And then hooked up one contact to a metal sleeve inside the barrel, and another to a wire attached directly to the spring? Then finally topped it off with a slider with an MX mount on top?

There would be problems. But a proof of concept for a discrete BS switch would be quite something!

My exact idea could be a load of crap. But there's a lot of ways to do it. I really think we could make modular buckling spring if we wanted.

[matt3o]

this is what is going to happen.

- Unicomp will make a new SSK design
- Unicomp will screw it up because of Unicomp
- Everybody will whine about how lame the new keyboard is
- Everybody will buy it anyway

win-win situation for Unicomp

[scottc]

Okay, so I was using my fingers. But still!

This is a nice way to have bloody fingers... and I speak from experience!

damn you, Dell AT102W...

[Muirium]

Sore, but not bloody. (I got a key puller eventually, so no more.) It's computer insides that get me, every time. One of the reasons I prefer messing with keyboards than that stuff.

[rindorbrot]

I hope for a classic (ISO) SSK with added Win-keys, short right shift and Trackpoint!

[bhtooefr]

For what it's worth, beam spring sphericals are my favorite key caps. They blow the wimpy SP sphericals away, and in fact the dished F and J are pyramidical, not even spherical.

[Muirium]

Even Signature Plastics' tall SA family sphericals? Those match up fairly well against authentic Honeywell caps:



http://deskthority.net/post125355.html#p125355

Let's see what you've got!

[Dubsgalore]

I mean..if this comes out..and is reasonably priced..I'd probably grab it just because

[scottc]

I have to agree with bhtooefr here: nothing beats those beautiful beamspring caps.

[bhtooefr]

I was comparing to SA, not DSA, for what it's worth.

The SA deep F/J is ALMOST as deep as the normal beam spring caps.

(Also, the Selectric doesn't have as aggressive of an F or J as the beam springs, but all other Selectric keys are the same shape on a given row.)

[Muirium]

Yeah, the Honeywell Hall Effect caps I compared SA to above are a fair bit deeper dished too. You need a ton of plastic depth to pull that off, as evidenced underneath.

[Hypersphere]

Apparently there was much more interest for Unicomp SSK when it was just a proposal/mockup/dream than now that it's going to happen. Our love for keyboards may be platonic after all.

As others have said, a dream represents an ideal. When reality comes, it usually falls short of the ideal and therefore causes disappointment. Many fear that Unicomp will not choose the best design, and even if they did, it seems probable that the execution will not have the build quality and fit and finish of bona fide IBM products of yesteryear.

In any event, today I sent Unicomp a detailed set of comments. Briefly, I told them that they ought to produce two distinct keyboards: (1) a classical TKL design like the original SSK; and (2) a true 60% design like the HHKB, perhaps adding 3 additional keys on the bottom row. However, I suggested that if they were to persist with a hybrid design that they make certain improvements to their April 1st concept rendering, including some that I have seen in the comments in this thread.

[Miko]

Still absolutely nothing about it on their website, this is a aprils fools. It would have been so nice.

[Muirium]

That's a pity. If it was an April fool after all, something tells me the customers wanting the return of a classic product (that still sells used for more than the new stuff Unicomp makes today) aren't the fools in the equation.

[Daemon Raccoon]

«Je suis un poisson d'avril.»

[Hypersphere]

Still absolutely nothing about it on their website, this is a aprils fools. It would have been so nice.

If it is an April 1st prank, they are prolonging it. I sent them an inquiry a few days ago and they replied indicating the special email address to use for sending comments about the proposed new keyboard. Today I sent my comments and they replied with thanks. I would hope that if this were all a joke, they would admit to it. I would suggest that we send comments anyway, assuming that the project is real; it it is not, then perhaps a deluge of comments will provoke genuine action.

If the project is not real or if it is real and the product is a disappointment, there are still IBM SSKs available. I am grateful to have one and to know that it is sufficiently adaptable to be reconfigured with a HHKB layout!

[8bit]

As they are a company, there can be also a 3rd choice - a "let's check interest" announcement - half true, half false - if they arouse enough excitement (quantity, people), then it becomes true, otherwhise it becomes false. Net result looks like a group buy, but in a commercial skin.