I would like to work on a custom keyboard for the upcoming Framework 16. As the stated in the announcement event, there is only 3.7mm of thickness allowed for input modules.
Does anyone know of keyboard switches I can use that would fit? Even the Kailh Notebook Switches which are the thinnest I have found so far appear to be too thick by a decent margin.
Cherry ULP is only 3.5mm high according to the data sheet, but that is the mechanism only and would presumably still need to be mounted onto a PCB or baseplate, and have keycaps on top.
The quoted 3.7mm height feels a bit optimistic even for any sort of reasonable membrane key switch if it has to include the underlying PCB and keycaps. Instead, I am inclined to think the similarity of “keyboard module” height to that of the Cherry switch suggests this refers to switch height and they have aimed at compatibility with this specific type (allowing 0.2mm for tolerances).
Certainly Framework are likely to be aware of the desire for mechanical switches amongst its users, given how the rest of their recent product announcements reflect demands on this forum.
I think the company should at least offer blank modules of all sizes for custom mapping. Because a DIY input module that can blend in is going to be so difficult to produce…
Having a choice of keycaps or a service to produce keycaps with custom markings on demand would make it perfect.
My understanding is that all modules including standard ones are going to implement the same open firmware we are free to customize.
For my needs, I’d be happy to have small (half-numpad) modules with some buttons to map according to my liking. Once I can have a dedicated Home+End+PgUp+PgDn+… group - that would make me so happy.
I don’t need a blank module, just the key switches themselves. The module I’d like to produce would be significantly different than just swapping key caps.
I think, at this level, it is a matter of finding a manufacturer in China who can produce a small batch of completely custom keyboard (membrane layers, rubber domes, stamped metal frame, switches used in mass production…) or cannibalizing an existing keyboard and figuring out how to reuse the switches in a custom layout.
A sandwich of layers found in a mass-produced keyboard might be too fiddly to make into an individual module of equal height and find a market for the price…
We are exploring a way to make one-key modules that use our existing keycap and scissor parts to make it more feasible to develop key-based Input Modules.
That sounds like an excellent solution.
I’m guessing/hoping that they will reserve enough room for these modules to be mounted to a rigid base plate of some sort, so I can still arrange them at my desired angles and wire them all together.
Thanks
Ohoy! I’m late to the party. I ended up here after thinking the exact same thing about Cherry ULP switches. I’m a bit unclear on the dimensions offered on the Input Module GitHub page though. I’ve seen the number 3.7mm getting thrown around. Is that the total available area between the bottom of an input module and a screen when the laptop is folded?
Yes. From the top of the keycap surface, to the bottom of the input cover / keyboard assembly.
It comes from the Framework-16 announcement
Really, almost anything can be created into an Input Module. The only limit is your imagination, and the 3.7mm height constraints.
Given that they’re not talking about parts or components but about whole input modules, I would take 3.7mm as the total available height of a complete input module.
Yeah… oof… That’s what I thought. 3.7mm is a reaaaaally tough constraint to work with. I don’t even think a good chunk of traditional scissor switches would fit in that tight of a footprint.
I think the one-key scissor module the @nrp brought up is going to be mandatory if Framework wants to see creative keyboard layouts as part of the ecosystem. The tooling required to get a custom layout from a laptop keybed manufacturer is likely prohibitive to most makers.
Aren’t you giving up on Cherry ULP switches a little quickly? We have a whole 0.2mm to work with!
I’ll eat my shorts if someone can fit a durable PCB and keycap in 0.2mm. 
I don’t know why Framework would launch a “power user” focussed laptop - especially mindful of the value they put on the Maker community - that does not accommodate the benchmark brand discrete keyboard switch, that being Cherry. This strikes me as a grave error.
Is the die cast or is is the design still subject to modification prior to full production?
They certainly could have designed a laptop around cherry switches, but I think Framework recognise that the market for 4cm thick laptops is not very large.
When referring to Cherry I meant their ULP switch linked to upthread (3.5 mm deep mechanism). I apologise for not clarifying, having just assumed readers would be aware.
Whilst nobody wants a 4 cm deep computer, Framework already have an ultra-portable, and this 16" model does not have to compete with the 13" for portability or aesthetics. In fact there are gamer or prosumer laptops with these discrete switches already, so other firms think that the better typing experience justifies any additional bulk. The implication is that power users will give up some degree of portability for a better computing experience.
Whilst not perhaps the “grave error” of my previous post, a keyboard depth that precludes discrete switches seems like a glaring omission from a firm that otherwise is keen to engage makers. The keyboard of the last model was widely discussed, and thus desire by some for discrete switches from the likes of Kailh and Cherry would be known about to the design team. That they now hope to offer “one key modules” to allow hand-built keyboards suggests a desire to support the maker community after all, but who wants to use a proprietary solution which is probably inferior to that from firms who specialise in key switches, especially when subject to their self-imposed depth constraint?
I close out by commenting that a keyboard module depth that forecloses on commercially available discrete switches seems a regrettable decision to seek marginal reductions of thickness, an oversight, or at worse might be a deliberate choice to lock out the makers or aftermarket parts community from this area. I assume not the latter, in which case perhaps it is possible to make detail design changes still to rectify this?
I think it could be possible to change hinges and increase the thickness of the laptop this way 
There is also one detail regarding this: The closed laptop screen will be suspended in the air and not supported by anything. So it would need to somehow not break and be supported by something.