Custom PCB + case + chiclet keycaps: how much will it cost?

[vemv]

Hey there,

I'll want to commission (to any suitable manufacturer) a keyboard with a custom physical layout.

With which I mean:

- there's a custom PCB with a unique arrangement of keys (particularly modifier keys)
- there's a custom case which can contain such a PCB. Don't care about its appearance.
- keycaps need to have custom sizes. I don't want mechanical keys - Classic Apple-like chiclet would be perfect. But I need non-standard dimensions of those chiclet keycaps. I envision I'll need 6 different keycap 'types' (dimensions). _{I don't entirely discard the idea of mechanical keys. But 6 custom sizes sounds hard?}
- Predictably, this custom keycap size affects the proportions of the PCB, so there has to be a holistic design of the solution;
- Otherwise no bells and whistles: I don't need bluetooth, programmability, etc. Just a dumb keyboard, with a custom arrangement and size of keys.

And obviously, someone will have to design, manufacture, test and deliver that stuff to me. I'm only interested in the final product, as a prototype, for validating the idea. i.e. I'm not a hardware engineer or electronics aficionado.

Or at most I could order 100 or 1000 units in an attempt to make the project have a cost of zero to me.

How much money should I be willing to spend? What's the absolute minimum I can spend for having a single 'experimental' prototype at my door? And what recommendable amount I should invest for having e.g. 1000 'near-production-grade' units at my door?

I live in Europe, and am open to trade with specialists/manufacturers from anywhere in the world.

I'm kind of willing to spend e.g. 20k euros in the project (operating under the assumption that people will buy 999 units of my experiment). But optimistically, I'd like to think that keyboards aren't rocket science and maybe some hacker can deliver an MVP of my idea for e.g. 2k euros.

Any hint appreciated. Thanks!

[Techno Trousers]

I don't have any advice, but I'm watching this thread because it's an interesting topic. How do people spearhead the creation of new keyboards from scratch, including the case and key caps?

[suicidal_orange]

You're thinking chiclet is cheap but there are actually more custom parts involved (mechanical you only need a PCB and a case, a plate is usually included but optional) while for a chiclet you need a case, plate and rubber dome sheet as well as the PCB.  Then assembly, you only need to solder for a mechanical board but for a chiclet you need to attach all the little clips for each switch and stack all the layers together.  If you're making thousands of boards the setup cost is spread out so the product can be cheaper but for a small run it won't be.  You're not going to find a hacker with the machinery required to make a rubber dome sheet or keycaps (at this price - kurplop manually milled some for his amazing Planet 6) either.

You'll need to find a keyboard manufacturer in China to ask but for my guess is by the time someone's designed all the parts you'll have spent half your 20k on the prototype if you go for completely custom everything.

Good luck, and that's not sarcastic but I think you'll need it

[vemv]

Good hints, thank you.

Would the narrative change much if I went for mechanical instead?

[suicidal_orange]

I've re-read this multiple times and am not sure but I think what you want in the end is one fully functional keyboard, and you have loads of money in the bank that you could spend but you'd like to get most of it back (or never spend it at all...)

In this case a one off custom layout mechanical PCB and a sandwhich case (looks like the board in this first post) to put it in will only cost a couple of hundred (depending how big it is), switches another 100 and that just leaves the keycaps.  I think it would be useful to know what you're thinking of doing with this board - can you try and explain the layout or make it on keyboard-layout-editor.com so we can see just how 'custom' you're thinking? 

[Findecanor]

- keycaps need to have custom sizes. I don't want mechanical keys - Classic Apple-like chiclet would be perfect. But I need non-standard dimensions of those chiclet keycaps. I envision I'll need 6 different keycap 'types' (dimensions). _{I don't entirely discard the idea of mechanical keys. But 6 custom sizes sounds hard?}

What kind of custom keycaps? Do they really need to be shaped that way or could you use standard sizes
laid out in a special way?

If you are building only one or a few then it would be very feasible to get ordinary flat keycaps and have laser-cut some flat pieces of textured acrylic that you would just glue on top of these. You could get the legends laser-engraved at the same time. You could also get keycaps 3D-printed.
You could also have the enclosure be a laser-cut sandwich of acrylic. There is also a type of acrylic that is topped with metal foil but not as many firms allow that and I am not sure about the durability. You could of course order metal parts instead from a machine shop but that is significantly more costly than acrylic.
Laser-cut acrylic is very affordable.

One upside with mechanical switches for DIY projects is that they are discrete modules that you can lay out any way you want.
Kailh makes at least three families of low(er)-profile mechanical switches and I think that chiclet keycaps and stabilisers for them should be available. Feel is supposedly close to full-size mech switches but with lower key travel.

[Leslieann]

I've dealt with low volume manufacturing and prototyping...

If you are willing to do some legwork, you can contract it all out and get one for $2k… Maybe
Have someone design a pcb, another design the case, send the pcb plans to have made, send the case to a local machine shop or China to be machined or laser cut from whatever. For keycaps, you could hire an artisan to make you a set, it won’t be doubleshot thick pbt or ABS, but you would have them. You could also have someone design them in Cad and send them off to a company like Shapeways and have them 3d printed using their powder system ($$$).

That’s the good news.
The bad news is that it rarely works the first time, especially when coordinating so many people and companies, much less working with people and companies you don’t know or trust. Particularly China, where you run the risk of them just simply walking away with your money and you having no recourse. The keyboard creation tools Suicidal_Orange mentioned will help, but with you wanting custom sizing and such, it's probably going to need some manipulation in cad.  Figure 6months to a year before you have what you want in your hands with 6 months being seriously rushed.

One offs and prototypes are not cheap, and shops tend to push them to the end of the line to make way for higher paying repetitive stuff so it could be 6 months before a shop even started on a case, depending on what you want. It's just how the industry works.  I generally tell people $1000 to make just a case, and that is with you doing the leg work and if things go right, but it can go up to several thousand really fast, and you still would need your keycaps.


If you want production quality and be able to recoup your investment your 20k would get you either a few keyboards or the caps but not both. Honestly, the keyboard is the easy part, keycaps are tricky and require several molds. Molds EAT money like Pacman eats pellets. There is a reason good keycaps cost as much as they do, you need lots of rather precise molds, it’s time consuming and labor intensive.

The only chance you have of doing 100 or 1000 at this price is to coordinate with a Chinese manufacturer and leave them the rights to it to white label it to others (sell it to others), which you don’t seem opposed to. On the surface this sounds great, basically you subsidize their R&D. GREAT! Not so fast. The last guy I knew who did this sent a 3d printer over to have cloned, they quoted him a price, made his printers stuck them in a container and shipped them out on a boat. A week later, before he got his printers, mind you, those printers were on Ebay, Amazon and others for about $40 more than he paid, and they included shipping. Basically once he shipped them, he made nothing on each printer if he wanted to compete, 6 months later they were for sale for quite a bit less than he paid and again, with shipping.

For this, time depends on money and many other things, it could be 6 months or it could be 2 years.

[vemv]

@suicidal_orange

I've re-read this multiple times and am not sure but I think what you want in the end is one fully functional keyboard, and you have loads of money in the bank that you could spend but you'd like to get most of it back (or never spend it at all...)

Sure thing, I want to scratch my own itch first and foremost. If other people get to enjoy my design, cool, win-win. As for money, the mentioned figure is something I'm willing to achieve (as a consultant, via a loan, etc) but certainly not something I have ready to burn right now First I should evaluate the actual cost and risks.

I think it would be useful to know what you're thinking of doing with this board - can you try and explain the layout?

Here it is: https://i.imgur.com/E8bW4yW.png . In the background you can see an original Apple wireless keyboard (v1) as a size comparison. Recap:

- Only slightly wider than an Apple WK1, that's an active goal I pursue.
- Columnar
- Chiclet preferrable, but mechanical OK.
- Unique physical arrangement
- 6 keycap sizes: A-Z, Control, F1, Esc, Space, Command. The size variety has ergonomic impact and also helps with symmetry, beauty, compactness etc. In particular I enjoy small keycaps for certain tasks.
- I hope that the custom keycap sizes are feasible to stabilize.

@Findecanor

If you are building only one or a few then it would be very feasible to get ordinary flat keycaps and have laser-cut some flat pieces of textured acrylic that you would just glue on top of these.

Sounds ace, hadn't thought of that. Except for having smaller keycaps maybe? (my arrow/function keys)

@Leslieann

Got it, thanks for the insights. Tough world!

Perhaps you can recommend me some other plan given the extra info I provided above?

Thank you all.

[Findecanor]

Ah. I think all available low-profile mechanical switches are too large for the half-height keys.

There exists a commercial keyboard line with orthonormal layout and low-profile keys: the TypeMatrix. You might not like the style or the layout though.

There are a few PCBs available for DIY low-profile Kailh switches, with orthonormal layout but smaller size...  A few more, including one larger is in the works.
Someone also made their own PCB to put  Kailh switches into a Plum keyboard, slightly larger than yours and only 1u and 3u keys.

[Leslieann]

Got it, thanks for the insights. Tough world!

Perhaps you can recommend me some other plan given the extra info I provided above?

It's the reality of manufacturing, low volume means lots of labor, high volume means a high buy in but a low repeat cost.

One problem you are going to face with this design, as mentioned is key spacing, your caps are smaller than low profile mechanical switches. While you can work around this, it won't be easy or cheap. Keyboards with these sorts of layouts are almost always scissor or rubber dome for packaging reasons and about the only way you can do those is with injection molding. The reason mechanicals allow for customizing is that they are self contained, rubber domes and scissor switches are made as a whole, you can't just pop out the switch and move it around, the plate and matrix are built together and need to be precision injection molded. And if you think the precision is not a big deal at this level, take a look at the issues Apple is having right now with their new ones.

I'm not saying it's impossible to do, but it's difficult, which means expensive. Worse, no 3d printed part or resin cast part will hold up long term on parts the size we are talking. Actually, it can, but it will be EXTREMELY expensive and regardless it won't be as smooth without hand finishing. Rubber domes and scissors were created as a way to make keyboards cheap, but that only works because of volume/injection molding. It doesn't work small scale like you want, mechanicals do, but they come at a size disadvantage.


Also, you seem to be confused, chicklet or island style is the keycap style, not the switch type. You can make island/chicklet keycaps for mechanical switches, a few companies even make some that are sort of in between chicklet and normal for MX switches.

[Lanrefni]

Ah. I think all available low-profile mechanical switches are too large for the half-height keys.

There exists a commercial keyboard line with orthonormal layout and low-profile keys: the TypeMatrix. You might not like the style or the layout though.

There are a few PCBs available for DIY low-profile Kailh switches, with orthonormal layout but smaller size...  A few more, including one larger is in the works.
Someone also made their own PCB to put  Kailh switches into a Plum keyboard, slightly larger than yours and only 1u and 3u keys.

I'm giving some serious thought to buying another set of BFO-9000 pcb's and some Kailh switches to put in my Plum Ortholinear case,looks like I'd just have to beak off one column from each pcb to make it work,then engage in some creative gluing to attach the 3u stabilizers.

[ErgoMacros]

Hi,
   You really, really need to read the Kickstarter updates for the Keybordio Model 1.
   Here: https://www.kickstarter.com/projects/keyboardio/the-model-01-an-heirloom-grade-keyboard-for-seriou/updates

* They raised $652,001 on a $120,000 goal
* They are totally obsessed with quality
* Each keyboard retails for $329.00
* They had all the manufacturing done in China including:
    * Custom shaped keys
    * Custom engraving of legends
    * Custom cases (wood)
    * much else as well.
* They are 1200+ days into the project, have shipped all the KS backer's keyboards
* have shipped more (maybe way more) than 2,000 keyboards.

The updates give you some idea of the things that can go wrong, and right (not nearly so often) in a manufacturing effort. Really amazing to read.

With that as background you may have a greater appreciation for the array of skills you're looking for, and how talented people have to be to overcome the inevitable hurdles.

All the best on your project, it sounds interesting!