Project: SSK revival

[mkawa]

UPDATE 11.28
sorry, i've been super derelict in keeping this OP updated. here's the latest update on where this project is at: http://geekhack.org/index.php?topic=33430.msg724146#msg724146

<insert link to SSK wiki here>

I will be starting a community-driven project to revive the SSK, with limited help from Unicomp (as in they are not aware that this is happening, but they will be providing parts regardless )

Current status: there is a board en route to my CAD guy, who will disassemble, take measurements, and design a new machinable aluminum (or acrylic, depending) casing for the SSK BS internals. we are aiming for ease of manufacture and low target cost.

Project goal: brand new SSKs, constructed by hand, for everyone who wants one.

[mkawa]

reserved

[boost]

Can't wait.

[Djuzuh]

Is the goal to have a TKL BS board, or to have a replica of the SSK?

The first one seems more interesting, as we can customize it, and make it smaller.

[mkawa]

definitely the former. an aluminum replica of the original SSK casing would be prohibitively expensive, and defeat the purpose of a ground-up rethink.

[wcass]

i started a similar project at DT during the down time here. model F (capacitive BS) is my #1 switch choice, so i started with an XT and will re-use everything but the matrix and barrel plate - this to keep the cost as low as possible. already decided on a tenkeyless key layout. i would be very happy to help with this as there is quite a bit of overlap. my 2D CAD skills are pretty good.

[fartq]

interesting. your project is much more extensive, as you need to fab a pcb, barrel plate, etc. for this SSK project nearly all the internals are either directly available from unicomp or modifiable from parts that are. the scope of this project is largely limited designing a new economical (and probably smaller) casing for the standard SSK internals.

[Djuzuh]

unicomp has all the SSK internals for sale ? 

[fartq]

not directly, but with some relatively minimal hacking, internals can be constructed from what they have available or tooling for.

[wcass]

the XT had separate barrels and barrel plate, so all i need is a plate with holes cut and will reuse the old barrels. this allows for lots of layout options. i ordered a prototype plate last week ($25 + ship). i expect that the plates would be much cheaper if i ordered 20 or so.

 i was planning on doing a compact 16x5 layout next. i had an idea of replacing the bottom membrane with very thin PCB. each column of 5 keys would have one of these http://www.digikey.com/product-detail/en/MMBD4148PLM-7/MMBD4148PLMDICT-ND/1964793 to give full NKRO. putting it at the top of the column or on the back will keep it out of the way of the hammers.

[fartq]

this is very interesting. i suspect there will be overlap between the projects. any way to move your project thread over here?

[wcass]

the XTant project is established at DT, but i would be happy to discuss/plan the compact keyboard project here.

what do you think of this layout? the keys left of the left arrow and up arrow are stepped "caps lock" keys. this should help isolate the arrow block.

[hashbaz]

I support these projects and can't wait to get my hands on one or both.  A custom BS board would be even cooler than a vintage SSK.

The layout above needs a tilde key IMO.

[ironman31]

I am definitely interested.

[wcass]

The layout above needs a tilde key IMO.

i believe that all keyboards should be 100% programmable. only the key layout is set in stone, and even that can be converted to ANSI/ISO enter and left shift. i know a lot of people like the Ctrl in the caps lock position and they will be able to have that. want a Fn on both sides of the board? you can have that too.

[mkawa]

i'm starting to warm up to the tilde placement on the pure. maybe the scrlock in your layout should turn into a tilde?

also the two keys above it pgup/pgdown

naturally since this will need a custom controller it will be programmable, but let's face it, you can't just release a programmable controller and tell people to "have at it"

[hashbaz]

I like the idea of ditching scroll lock for tilde. Don't think I have ever in my life used that key. But as a linux user I use tilde dozens of times a day.

This is all just my perspective of course and it's your design ultimately.  And I suppose with model M key caps it's mostly irrelevant isn't it?  They are all the same profile.

[mkawa]

wcass, i've never seen an XT's barrel plate. do you have some pics you could link or upload?

also, the problem with a think pcb (and the reason why Ms use a membrane) is that the single-key-profile design more or less requires a curved backplate, which is incompatible with most pcb materials. (in particular, cheap ones  )

[wcass]

i'll upload some pictures when i get home.

I measured/calculated the curvature on the XT to be pretty close to 12" radius. The PCB in the XT is .031". that thicknes and thinner will work, but is a curve really nessessary? look at the new Apple keyboards. not having the curve will make the case fab MUCH easier, cheaper, and make the keyboard thinner and easier to transport.

[harrison]

looking forward to seeing some design mock ups of the casing.

[Parak]

As another side note, the singular barrels are all curved as well, so I can see how they wouldn't work all that great with a flat plate. Ideally, a new barrel for flat plates would have to be designed and molded, and here improvements can be made for diy - exact outer 1/2" barrel diameter drill, smaller drill under that for the stabilizing pin, etc. Molding new barrels out of something like uhmwpe is pretty expensive, though. Not to mention that the capacitive feet would still need to be sourced from vintage boards..

[mkawa]

interesting note parak. injection molding is hard, to say the least, and doesn't work very well at small scales. there is 3d printing, but damorgue raised some pretty valid concerns re: surface smoothness. we know that unicomp can provide curved plates and barrel frames, so for this initial SSK project at least, I'd like to stick with them.

what would be really nice actually is if we could look into getting custom membranes made. further, is it possible to print a diode-like device onto a membrane? now _that_ would be cool.

[Parak]

uhmwpe is actually not injection molded, so there are some pretty nice dimensional tolerances in the process. Unfortunately, because it's not injection molded, expect even higher costs

[alaricljs]

I thought MS's NKRO board had diodes printed on the membrane...

[fartq]

then the question becomes can _we_ print diodes on a membrane?

[wcass]

as promised, pictures of the barrel plate.

the AT and 122 barrel plates have notches at 30 degrees below horizontal; the XT barrel plate has a hole for a pin centered .4" below the center of the barrel.

[fartq]

to be clear, is the last picture measuring the flat bit of the individual barrel plate? and the other dimension of the ind. barrel plate has the same curve as the underside of the metal barrel insert plate?

[wcass]

to be clear, is the last picture measuring the flat bit of the individual barrel plate? and the other dimension of the ind. barrel plate has the same curve as the underside of the metal barrel insert plate?

the last picture is a side view of an individual barrel. i am still not sure if it was born curved. i think there is a .0005 gap at center curve now, but that might just be from being pressed into a curve for 30 years. when my XTant test plate comes in i will check fit with 84 barrels while the plate is still flat.

for DIY membrane, you can get polyester or mylar sheets from art/drafting supply stores. the M uses .005 for bottom and middle, .003 for top. go to a hobby store for 1/16" wide model masking tape. and instead of silver silk screen, try this:
http://www.amazon.com/CircuitWriterTM-Precision-Pen-silver-based-grams/dp/B0002BBVQO

http://www.hobbysuperstore.com/browse.cfm/masking-tape-1-16/4,66417.html
http://www.pearlpaint.com/shop-Mylar-Sheets_6935_6934.html

[fartq]

no need for full DIY membranes unless we have some way to print a semiconductor on

BUT, i just realized how we could add components to the membrane circuit: machine more holes in the backplate, place components in holes, connect leads of components onto appropriate points on backside of membrane (which needs to have some kind of makeshift via)

it may be possible to program the mill to make the appropriate holes (just large enough to fit an SMT diode), and vias through membranes are definitely possible, so it may actually be possible to make an NKRO SSK...

[telepete]

This stuff is pretty interesting. I'm glad to hear that you have some plan on how to make changes to the membrane circuit. That would definitely give me some problems.

[mkawa]

so, my CAD guy has the board and is making some initial drawings and measurements. i'll post them as soon as he gives them to me.

[rao2100]

This is pretty cool, I will be in for one.

[keyboardlover]

Will they be 2-kro?

[Icarium]

Hopefully we can make them NKRO or at least 6KRO. (Though going from 6 to N is just a programming nuissance.)

[mkawa]

Will they be 2-kro?

unless my completely insane idea with embedding the diodes in the backplate works (which i highly doubt it will), yes, it will use the unicomp membranes and controller and hence be 2kro.

[Icarium]

Okay, sorry I was talking about the XTant efforts. (Still would have preferred FUK for the name. )

[keyboardlover]

I am tentatively interested.

[486]

I really like the idea. I hope it comes into fruition.
Any way, with this new project, will vintage SSK prices increase or descrease?

[REVENGE]

Will they be 2-kro?

unless my completely insane idea with embedding the diodes in the backplate works (which i highly doubt it will), yes, it will use the unicomp membranes and controller and hence be 2kro.

Better go license resistive multitouch...

[jnsjr]

Count me in! BS TKL is my dream.

[SmallFry]

Just so that you know Mkawa, I have my AIKON project...sort of. I have to reorder the circuit boards, but I think that they are just about functional. I'll get back to you on it. Shoot me a PM though.

[mkawa]

PM shot

[mkawa]

fyi, we're going to start by getting a precise model of an SSK backplate, and then seeing if there's some easy to integrate a casing into something that's equally compatible with SSK frames

that didn't make any sense. bottom line: one piece backplate and casing. possible? only autocad knows

[Elrick]

fyi, we're going to start by getting a precise model of an SSK backplate, and then seeing if there's some easy to integrate a casing into something that's equally compatible with SSK frames

that didn't make any sense. bottom line: one piece backplate and casing. possible? only autocad knows

If you do manage to complete this, please sell it as a kit, so that the rest of us can piece it together.  I would love the opportunity to build an SSK from parts delivered from this venture.

It would remind me of the early PC kits that were sold in the late 1980's.  Buying a kit from Altronics here in Perth then running home to build it up.  Took several weeks to complete but it was done.  Damn I was proud of myself back then.......

[jkercado]

What's the projected price for this keyboard?

[deleet]

Definitely in!

[mkawa]

my hope is that we can get the case to come in at 200 or less, including custom powdercoating. at current unicomp prices, the rest of the parts are around 100$. i have no idea what will happen to unicomp's parts pricing though if the projects succeeds and the kit becomes remotely popular (personally, i'm just hoping to build 5-10 new units as a proof of concept right now)

[slueth]

interested!

[hoggy]

Interested.

[do_Og@n]

  EXCITED