deskthority

Greetings!

Just over a week ago I received a package and in that package was wcass' unique Model F compact SSK. I've been using it as my daily-driver for about a week now. I've now gotten enough information to give a good, solid review and consent to tear it down and comment on its construction.

Geek Out time

First things first - this one:



Keyboard Layout / Feel / Sound

It feels glorious. In theory the keys should be slightly stretched on the y-axis but if they are it isn't perceptible at all. Those of us who own multiple versions of the Model F can attest to the different models feel and sound different. This keyboard has the key 'snap' and 'feel' of the XT and the 4704 more than the more 'loosey-goosy-hippie' F-122.

I'll upload audio in post #2 later today.

My usual daily driver is the F-122 at full height (think typewriter rake) so I did prop the keyboard up for my comfort. It's still glorious.

The Teardown, a tale of three parts.

USB micro connection on the top face as you might expect... ... and HOLY JYR&^F - how many fasteners? ... needless to say, the back isn't going anywhere. The second thing I noticed was that each of the fastners released with a 'snap' which tells me that wcass either owns a tension driver or is just obsessive with attention to detail.

About 15 snaps in, I noticed something else which I always have forgotten to do:
The fasteners have been perfectly countersunk. When they're in they're completely flush - not perceptibly above or below the metal case.

Just before I let loose the last few I remembered that it's wise to remove all the keys before opening a Model F lest it explode into a million pieces:
Oh Hello - let's take a closer look at that spacebar retainer... Yes, even that was countersunk(!).

... Let's look at the back!
Nice - it's subtle and hard to see here but the two screenholes you see in the PCB are ground and plated which means that they're making electrical contact with the case itself. You can also see that the USB socket is on a daughter-board which is nice - means one can modify either or both at will with a minimum of fuss.

Epiphany

Anyone who has ever worked on a Model F knows how hard it can be to re-assemble them. You have all the feet in tension against the plate, they all need to be aligned perfectly - you slide the PCB and base-plate against the other until they align... It's hard because the whole thing is under tension due to the use of foam padding etc.

The different approach: The base of the barrels and the PCB are countersunk into the case. This results in the PCB resting on top of the barrels with no pushing and shoving required. Speaking of pushing and shoving, what about that daughter-board?
The daughterboard is also countersunk to the depth required such that it makes direct physical contact with the PCB when the PCB is put on top. The daughter-board can move a little when the main PCB isn't there but the posts on the interfacing component alight directly with holes in the main PCB which index the connector perfectly.

Did I mention earlier that there was no foam and no index pins in the barrels? Then how do they keep aligned so we don't have the barrels rotating? If you hadn't guessed already:
The barrels themselves are also countersunk.

Final View
The final photo which I think says it all for anyone that has ever struggled to re-assemble a Model F. See all those barrels in place? See all those feet in perfect alignment? To re-assemble I just dropped the PCB in. The PCB fell into the cutout that is perfectly sized for it with no fighting or tension or anything. Just add the fasteners and done.

Common Sense

Another way that this design differs from your typical Model F designs is that the PCB is four layer and the controller is directly connected (as opposed to via wires etc from a cypress devkit).
The signal to noise ratio for this board and this design is significantly better than even my original IBM produced beamspring PCB. That's food for thought right there.

Final Notes

Attention to detail is an understatement. I have no idea at all how wcass was able to get all of these parts to align at all, let alone first time. His PCB outperforms the original IBM keyboards. His board doesn't use foam because he doesn't need the 'futz-factor' the foam provides during manufacture.

AND, it's a keyboard that has the snap of the old XT and is a pleasure to type on.

Now if you'll excuse me, I have to go change my name, move house, and lose his return address

wcass samples:
Flat: Propped Up: This does not do it justice.

For reference - other keybs in my collection:
XT: Unmodified 4704: Modified 4704 with craft foam: F-122:

I had no idea this even existed #dta2019

wcass' work is just plain fantastic; my only complaint here is about the layout (then again, isn't that what I always complain about?)

Now, can we commission wcass to make a 75% with this same level of craftmanship?

This is a truly wonderful project. I was fortunate to take a look at and tear down his earlier (non-functional) prototype, and this final product is truly light years beyond that one. I'm so impressed by all the little touches, like the countersunk barrels.

Great job, wcass!

Indeed this is just amazing! How much does that thing even weigh? Anyway that's some #dta2019 material right there

Now on to the big question. How do I get my own on my preferred layout

A masterpiece of craftsmanship! Truly humbling.

If it were possible to have something like this made to order for me, I would go for a 60% form factor without the top two rows of keys, and a split right shift and split backspace to enable a HHKB layout. Would that I could!

Spectacular work!

Could we get a typing video with sound?

Hypersphere wrote: ↑Could we get a typing video with sound?

After the kid goes to bed.

That's just phenomenal. I'll take mine in an M2 form factor please...

EDIT: It's a shame that you aren't in the EU. I'd love to see Chyros do a review of this.

DOUBLE EDIT: I really enjoyed your teardown and review though

__red__ wrote:

Hypersphere wrote: ↑Could we get a typing video with sound?

After the kid goes to bed.

Be sure to put some cotton in his or her ears. ;p

Wow. What else can you say but nice!!

...I'm looking at this and thinking about how a five axis mill would be able to countersink Model F parts into a curved machined surface. Very nice work.

XMIT wrote:...I'm looking at this and thinking about how a five axis mill would be able to countersink Model F parts into a curved machined surface. Very nice work.

Now you're talking: a full TKL model F SSK. {Drool}

Indeed! I tremble at the likely cost, but a solid metal SSK or HHKB is a compelling thought, to the point of downright mesmerising.

Hey Red: does this board really lose nothing by dodging foam entirely? Are there any downsides at all in feel and sound? I kind of worry for the lifespan of the flippers, whacking against bare metal and a solid backed PCB. I’d be happier if we also made those.

Flippers in a regular model F are bouncing off the PCB and underside of the barrel, so I don't think there's any difference here. If anything the sound should be louder and crisper without foam surrounding the barrels and deadening the sound.

Muirium wrote: ↑ Hey Red: does this board really lose nothing by dodging foam entirely? Are there any downsides at all in feel and sound?

I uploaded some samples above. I may re-record them because I'm not really happy with how they sound compared to reality. The keyboard sounds amazing. much better than in the recording in my estimation.

The biggest indicator of the effect that foam has I think is to listen to my converted and non-converted 4704s. The unconverted has deteriorating foam in it so loose as hell. The converted one has the craft foam that doesn't want to compress at all so everything is under tension.

I need a better microphone to catch the whole range, esp the bass. It sounds deep irl.

Shouldn't we start calling this (these, hopefully) unit(s) the Model W keyboard(s)?

The only material upgrade I could imagine would be using a 4 axis mill to tool everything the same way but with the classic profile curve.

Indeed. And a nice classic style case. And Bluetooth. And a pony. I always want a pony!

I've got a few hundred flippers and barrels, so please someone make it so! Would it help if I begged?

And a logo, done with IBM's font, that says "WCASS" (because using just the first two letters would make for undesirable jokes).

From the recordings as they are, I like the sound of the XT the best. The XT case provides a nice resonating chamber.

Nevertheless, wcass' work is beautiful. Given that they aren't using it anymore, maybe we could buy the former platinum-iridium meter from the French for machining the ultra-deluxe model of the Model W. Hmm, French-milled keyboards. Every luxury hotel room should have one.

Well, I’m all for it… if it comes to $100 or less for the whole keyboard!

Muirium wrote: ↑Well, I’m all for it… if it comes to $100 or less for the whole keyboard!

I'll up the ante and offer $110!!! (and I'll procure the caps myself as well!)

Hypersphere wrote: ↑From the recordings as they are, I like the sound of the XT the best. The XT case provides a nice resonating chamber.

Nevertheless, wcass' work is beautiful. Given that they aren't using it anymore, maybe we could buy the former platinum-iridium meter from the French for machining the ultra-deluxe model of the Model W. Hmm, French-milled keyboards. Every luxury hotel room should have one.

I'll settle for a solid slab of titanium.

and a slab of technetium for a certain someone...

Sangdrax wrote: ↑The only material upgrade I could imagine would be using a 4 axis mill to tool everything the same way but with the classic profile curve.

Ok, I'll bite and make the dumb question myself. Just by looking at the pictures of wcass' Model W (W1?) keyboard, it's clear that milling another case with a profile curve is absolutely within the realm of possibility. The problem would be getting a curved PCB... and IIRC, for the FSSK project, the provided PCBs were flat, which then had to be manually heated and beaten into the curved shape. Now, given that the curvature only changes when going from one row to another...

Would it be at all possible to make, for a lack of a better term, a set of PCB strips, all to be connected to the "main" one or to a daughterboard of sorts? That would solve the issue of making curved PCBs (and would simplify the milling part a bit, as the internal curve would only need to be a (simpler) open polygonal line).

That sounds a compelling idea.

Yeah that PCB strips idea should work easily so long as you don't have any keys taller than 1u. I've had the same idea for making curved profile Cherry boards.

Difference being that with Cherry curve is an experimental oddity, incompatible with most all of the best caps, while with IBM it is mandatory.

The technology has existed for some time to make roll-up musical keyboards, and now foldable phones and displays are making debuts. Should it not be possible to use the same technologies to make flexible keyboard PCBs?