[Thermally Modified Acacia Wood Edition]
[Taiwanese Hinoki Epoxy Edition, fully finished. Requires some minor adjustments]
Discord LinkIC Form0. Table of Contents1. Short Intro
2. Why Wood?
3. Why Project gwŷdd?
4. Project Journey So Far
5. Material Talk
6. Bonus Material Config
7. Who Will Enjoy This Board?
8. Plate Options
9. Specifications
10. Layout Support
11. Gallery
12. What’s Included?
13. Pricing
14. Credits and Special Thanks
15. Current Progress (Updates)
16. Changelog
TL;DR For Simon
Wood is an obnoxiously difficult material to make a non-tray mount keyboard out of. It was a very silly idea but I hope I've managed to overcome the material limits and machining challenges to share with everyone this passion project of mine. Grown, designed and machined in Taiwan, the gwŷdd65 is a gasket mount wooden keyboard that represents 14 months of development (so far) and a board that I hope many other wood-lovers and keyboard hobbyists around the world will enjoy <31. Short IntroHi everyone! I’ve been a long-time enjoyer of mechanical keyboards but only really dove into custom keyboards roughly when the QK65 arrived on the scene. Since then, I’ve truly enjoyed my time as a part of the community, witnessing beautiful projects come to life as well as interacting with fellow keeb-enjoyers (though mainly just lurking on geekhack). I’ve lived and worked in a couple of countries around the world but currently reside in Taiwan :)
I’d also like to take the opportunity to mention a good friend,
Masje, who was incredibly instrumental in the progression of this project. He got involved pretty early on in the project while I was still in the exploratory and feasibility stage, providing a fantastic source of information, comradery and just generally being a great guy to have discussions with and bounce ideas off of. The design of the board was also largely handled by him after I received guidance from the materials side of things. I couldn’t have asked for a better collaborator for this project :D
2. Why Wood?More
At the end of 2021 my family and I decided to move out of a major city to live in a more rural town in Taiwan. The proximity to nature – in particular the surrounding forests – in combination with the storied legacy of woodworking in the small town we now reside in and lastly the rise of woodworking channels on YouTube, inspired me to bring more wooden products into my life. For me, a well-machined metal product conveys a sense of advanced, high-precision technology, whereas a well-made wooden product conveys a sense of quality and timelessness.
3. Why Project gwŷdd?More
I wanted to find some ready-made wooden keyboards that would fulfil my desire to have that earthy feel and allow me to feel closer to nature. Sadly, most of the wooden boards I found didn’t hit the mark for me. Most of the existing wooden keyboards were tray-mount, which provided a typing feel and experience that was too stiff for me (side note – later on in the project I totally understood why most mass-produced wooden boards are tray mount, it’s just way easier to machine and more stable than other mounting methods).
During my search, I came across a major issue that wooden keyboards seemed to face, which was the eventual warping, boating and twisting that would in some cases inevitably compromise the usability of the board.
With these factors in mind, I sought to make a wooden keyboard for myself that would –
1) Feel great to type on
2) Find some way to overcome the inherent “playful nature” of wood
3) Have fantastic acoustic qualities
In addition to these objectives, I wanted to be able to share such a creation with hobby enthusiasts and enjoyers around the world, so some factor of production at scale was necessary. This almost instantly ruled out making the board by hand, as the high cost of labour and time meant that a hand-made board could never really be done at scale to meet realistic delivery timelines. If this project makes it to GB, I expect shipping to occur in 3 months or less, unless there's some delay with the logistics of non-case components. That is to say, nailing production down and resolving any issues encountered has always been a primary focus of this project. It's easy to make a prototype but hard to make that scale into production.
4. Project Journey So FarMore
This project is approximately 14 months in the works now. It has been an incredibly rewarding, fulfilling and challenging experience so far. I essentially started off knowing almost nothing about woodworking, forestry or the relevant ecosystems. I was quite new to the hobby then too, so I had to read up and learn a lot about keyboard design choices and their outcomes. Thankfully, I’ve had previous experience in exploring fields which are unknown to me, so I was able to eventually find the right people who could help and educate me on the wood material side of things. The same went for the keyboard side of things, with many of the OGs in the hobby being very willing to share their experience and knowledge with me. For that, I am eternally grateful and thank you all for these informative sessions and input, you know who you are :)
The journey so far has not been smooth sailing and without hiccups, of course. The initial challenge was approaching woodworking establishments and trying to ask them for help on something like this, because most of them knew next to nothing about custom mechanical keyboards and the requirements. The average age of the business owners being about 50-60 years old didn’t really help either. Quite a few of them just looked at me with bewildered eyes: “you wanna make that thing *points at keyboard*, out of wood???”
Getting a manufacturer that was able to machine the wood was another struggle and has accounted for most of the time spent on the project since February 2023. It’s not easy finding a manufacturer that’s willing to entertain you for a low MoQ project and also try something completely new. Due to the gasket mount requirement, we were in fact pushing some boundaries in terms of material tolerance and design choices as neither the wood supplier or CNC workshops were sure that it was even possible. Through iterative prototyping and design refinements, though, we were finally able to get to where we are today.
5. Material TalkMore
I had originally wanted to use Taiwanese Hinoki, a wood that holds a legendary status in Taiwan as being one of our national treasures. However, after carefully considering many aspects including sustainability, legitimate sourcing of the wood, variance in wood quality and overall stability and performance as a keyboard, I ultimately had to pivot to a more stable material, Thermally Modified Acacia Confusa, otherwise known as Formosa Acacia or Formosan Koa.
Acacia Confusa has an appearance similar to the coveted Hawaiian Koa and there are rumours that some of the trees in the eastern parts of Taiwan (around Taitung) are actually long-lost relatives of the Hawaiian Koa, however these claims haven’t been proven yet and simply make for a good story. In any case, expect to see some great wood grain and chatoyance, where the wood has below-surface gem-like reflective qualities that change depending on the angle of the light striking it.
There are many resources detailing what the Thermal Modification process is, but in a nutshell the wood is treated with heat (180C and above) in a vacuumed chamber to remove moisture and carbohydrates (residual sugars) from the wood. This also collapses some of the cell structures and thus makes the wood more dense, shrinking the wood by approximately 10%. Uneven moisture in the wood is what contributes to the majority of the warping, boating and “playful nature” of wood. It is something that most woodworkers have to find ways to mitigate if they want to craft a wooden product that stands the test of time. In addition to this, carbohydrates left over in the wood are what attracts living organisms, such as fungi, bacteria and insects. Thermally Modified Wood overcomes both these issues by removing the moisture and carbohydrates.
So what does this mean for our keyboard? We should expect way less movement in the wood after the board has been produced, increased resilience against temperature and humidity fluctuations as well as any wood-consuming organisms. There are some scientific studies and data available on this
over here if you like to read that sort of thing.
More
I’ll be doing a very small run of epoxy + Taiwanese Hinoki material cases, between 5-10 units. The costs (both monetary and time-wise) to produce just the epoxy + wood material are exorbitant and so I do not foresee these ever being feasible for a large-scale run.
This particular material choice fulfils my original goal of using Taiwanese Hinoki, however it is reinforced by the strength of epoxy. It also uses pieces of wood that by themselves could no longer be utilised for larger furniture or products, so it is a fantastic way to make use of wood that would otherwise have been discarded.
This version’s PCB will also have some bottom-facing LEDs that could be utilised to create some ambience or lighting effects. We’ll have an idea of how good this looks once we receive our prototypes for this version.
7. Who Will Enjoy This Board?More
I will, obviously, but I imagine anyone else who has a love for string instruments or wooden products for their warm to the touch and earthy qualities will appreciate this board. We’re still exploring the acoustic qualities of the board but we’re hoping for a warmer tone when compared to metallic boards.
8. Plate Options
More
One of the goals of the project was to achieve a pleasant bouncy typing experience with a wooden board, so we’ve opted for a leaf spring plate to work together with the gasket mount. The plate material for this will be FR4.
For those who prefer a firmer typing experience and a different sound profile, we’ll have a PC/POM/ALU (tbd) version, which may have a thickness of up to 3mm, but this requires further testing before we settle on the final choice for the second configuration option.
9. Specifications- Case Angle: 9
- Material: Thermally Modified Acacia Confusa
- Mount type: gasket strip.
- Weight: Approx. 510g unbuilt, 875g built, wrist rest 336g
- Effective Front Height: 22.5mm
- Screws: 8x M3 (4x 30mm, 4x 14mm), 2x M2 (5mm)
- PCB configs: Soldered (6.25u + 7u, ANSI + ISO), Hotswap (6.25u, ANSI)
- Feet: 8x Geonworks W1-AT Silicone Feet (black)
- Customised Hard Carrying Case that fits the keyboard and wrist rest will be included.
10. Layout Support
(Note, this is for solder only. Hotswap will only have one configuration available, which is ANSI + 6.25u bottom row and stepped capslock)
11. GalleryRendersMore
Physical Prototypes (v4.2) Taiwanese Hinoki Epoxy Edition, raw and requires finishingMore
Physical Prototypes (v4.1) and PCB (v2.0)More
I don't have an exploded layout render :( this is the closest you'll get:
Please don't newspaper me Simon
Previous Prototypes (v4.0)More
Previous Prototypes (v3.1)More
We initially messed around with some ideas of "sound vents/ports" at the rear of the case to allow for sound to be projected out of the board. This type of acoustic adventure would require a lot more work to make feasible so we had to forego it for the scope of this project. In future? I imagine the shape and positioning of the sound vents could, in theory, make some sort of acoustic difference. Perhaps shaping the cavity to build resonance like how musical instruments do, could be a thing? I'd love to just get a sound map of the vibrations inside the case to understand more about this.
The dorky-looking proportions are due to the initial guidance provided by the woodworkers (5cm additional bezels on the left and right sides to counteract potential warping and also edge grain weakness).
(to be uploaded later)
12. What's Included?I wanted to make sure that the final package included everything that I, as a consumer, would want in it. It was fairly challenging assembling all the components that I wanted as a first-time maker but I eventually found my way. Here it is:
- Keyboard case (bottom base + top frame)
- Wrist rest
- M3 and M2 screws required for assembly + some additional (in case you lose any)
- PCB - 1.6mm thickness, QMK/VIAL compatible. Default is solder for maximum layout support, hotswap requires add-on purchase.
- USB-C daughterboard
- Plate (choice of FR4 leafspring OR one of [PC/POM/ALU] to be confirmed. Add-on purchases should be available if you like more than one plate to mess around with)
- 12 adhesive gasket strips (60A or 70A, tbc) for each plate purchased (will probably include some extras)
- Geonworks W1-AT Silicone Feet (black), again will include some extras
- 30ml Natural Wax Polish (for any future maintenance you’d like to perform on the wood, it's a pretty easy process)
- EVA Hard Carrying Case, customised to hold not only the case but the wrist rest too.
13. Pricing$380-$420 MoQ 50, soft cap around 200 units to ensure a realistic delivery timeline (3 months) and also to allow for any post-GB follow-up that may be required.
Taiwanese Hinoki Epoxy version likely to be $650+ due to the greatly increased base cost of the materials involved. 5-10 units, most likely a raffle.
14. Credits and Special ThanksMasje (collaborator and good friend)
blacksimon (keeb education, feedback, a good keyboard friend (usually))
gok (maker experience, feedback, super nice guy)
baul (maker experience, feedback, the ultimate sweetheart)
corbett (feedback and good dude)
just.hymn (renders and awesome person)
To you, the reader, if you've made it this far: thank you very much! This project has been an obsession of mine for over a year now and I'm so grateful it's reached the stage that it has. If this project interests you in any way, please do consider dropping by and saying hi on the discord server, or filling out the IC Form. Cheers :D15. Current Progress (Updates)The prototype photos you see are v4.1. We're moving on to v4.20 :thumb: which we believe should be the final version. This will feature an updated wrist rest slope as the 9 degree slope of the v4.1 was too steep and somewhat uncomfortable to use. Final tweaking and keycaps clearance will also be done in this final iteration.
Blacksimon IC review16. Changelog28 Aug: Feedback heard loud and clear. Will be adding split backspace support to the soldered version.13 Sep: Should be getting our next round of prototypes within the week. The first attempt at the Hinoki Epoxy board went well but sadly experienced some chipping off of epoxy during the screw hole drilling step (RIP $$$). We have one more set of the material to hopefully get it right and establish a proof of concept for that particular material.
Also, updated soldered layout image.14 Nov: Original wrist rest angle was aesthetically pleasing but fell short functionally. Adjusted wrist rest slope and greatly improved usability. Internal fitment woes have finally been resolved (there goes a few more prototypes T_T) and keycap corner interference issues have been address. Fitment is looking great, prototype looks finalised and ready for production, hooray :DDiscord LinkIC Form
