A few weeks ago, I had the privilege to chat on Skype with Jesse Vincent, the co-founder and CTO of Keyboardio.
Keyboardio has two permanent team members—Jesse and his wife Kaia. In 2012, the couple began a keyboard quest to design a better ergonomic keyboard. Since then, Keyboardio has morphed from a side project into a real company with an upcoming Kickstarter campaign. After Kickstarter, Jesse and Kaia plan to take their final design into full factory production.
That design, pending any final tweaks, will be the Model 01, which is the result of tireless prototyping. It’s a split, adjustable keyboard with an organically designed maple case, custom molded keycaps, and RGB backlighting. It currently features Matias ALPS switches, and because it uses an ATMega32U4 controller, it’ll be fully hackable for anyone comfortable with Arduino programming. For the rest of us, there will be a user-friendly GUI configurator.
There’s no MSRP or launch date yet—Jesse is very committed to under-promising and over-delivering. But as he told me, Keyboardio has been “5 weeks away from Kickstarter for months” and beta units are already in testers’ hands—so there’s no doubt we’re reaching the final stretch. As for price, Jesse expects the Model 01 to be “competitive with the Kinesis Advantage.”
KC: To start, can you tell me how you first got into mechanical keyboards?
Jesse: Sure. That’s an interesting question, because all the computers I used as a kid came with mechanical keyboards. The very first computer I ever typed on was a TRS-80 Model 3. I’m embarrassed to say I don’t remember what switches it had. But the keyboards I grew up on were Apple IIs, which, if I remember correctly, had mostly ALPS switches. I’m sure a GeekHacker will correct me on that.
I’ve always had bad wrists, so for me it hasn’t been as much about mechanical switches as much as comfort. I’ve been through pretty much every ergonomic keyboard out there. There was a period of about 8 years when I carried around a Microsoft Natural Elite, the sort of mainstream white curvy one that cost about $20. I’d buy them in 5-packs, and needed to, because they’d melt in the rain. Literally, the traces inside would be irreparably damaged if they got a little wet. It was an OK keyboard, it of course was rubber dome, and they way it worked was that it had 3 pieces of plastic film, traces on the top and bottom, and holes in the middle one. Pushing down the keys would hammer those pieces of film together and make the contact. Incredibly inexpensive to make.
Over time, I’ve tried a Kinesis Advantage, a DataHand, a SafeType, a Maltron, and even spent a little time with an IBM M15. I spent a while trying to type on a TouchStream by Fingerworks. It’s basically a pair of trackpads with some awesome firmware, so it can do application-specific macros. The touch targets for each key get smaller or larger depending on what you’re typing. I have friends whose careers depended on them for a decade. It never really did it for me, and that company vanished one day out of the blue. Nobody knew what happened with them until two years later, when it turned out they were bought by Apple to become the touch technology in the iPhone.
At last count, I think I have 40 or 50 ergonomic keyboards, and none of them ever really quite did it for me. Maybe three years ago I was sitting on the couch at my in-laws’ over Christmas break and somehow ended up researching keyboards. I figured there had to be something else I could carry around besides the Microsoft Natural. That was the week I joined GeekHack, but it was also the week I found the original Poker. It wasn’t available yet, but I knew I wanted one. I bought three or four keyboards based on that research. I bought a Happy Hacking, I bought something that had MX Browns… my hands are how I do my job, and it’s really important that my hands be comfortable while I’m typing. I played with a TypeMatrix sometime around then. It was OK, but it wasn’t right. For me, having those straight up-and-down columns meant I was still bending my wrists, and I’ve had wrist issues all my life.
Soon after that, I discovered Dox talking about the beginnings of the ErgoDox project, and it looked really cool. I figured, I want to build one of those. At that point they hosted the STL files for the enclosure but not the circuit boards. When I looked at the ErgoDox, I thought it had solved all the problems of the Kinesis, because what it basically did was take the Kinesis and make it flat. So I got Shapeways to print me the enclosure, and this took place over the course of about 6-9 months. I figured, if I get the enclosure printed, I’ll buy some switches and diodes. So I did, and I bought them. Then, I figured if I could wire up the left hand and get it to go, I’d take a look at the right hand, and it was sort of these micro-challenges to get myself one step further. I hadn’t soldered since I was a little kid, and knew almost nothing about hardware. But my previous startup had failed that year, and I swore I was going to take a year off and mess around. One of these things I decided, was, maybe I should make a keyboard. The ErgoDox didn’t work or feel the way I wanted it to. And that was the very early genesis, before I started doing full-on design myself.
KC: Do you type using an “ergonomic” layout? Dvorak, Colemak, Workman?
I actually still type QWERTY. At various times I’ve learned Dvorak and Colemak, but neither has stuck. What’s weird about me is that I never learned to touch type until I started designing keyboards. When I was a kid, I learned to type before getting to any school with a typing class, and Mavis Beacon did not give me a hard time for using the wrong fingers. Now, Mavis Beacon is sort of the hallmark of a certain generation of nerds. That was the typing tutor you ran on your Apple II. So when I got to high school and enrolled in typing class, I got kicked out of class after about a week, because I “already knew how to type and was making trouble,” because I was helping other kids when they couldn’t figure it out.
So I never learned to touch type properly, and I have a non-traditional 7-finger typing style. I type “Y” with a different hand depending on the word. It’s kind of special, not that good. But as we’ve been designing the Model 01, we’ve had people ask us about other layouts and why we’re sticking with QWERTY. In the firmware, the keyboard already speaks Dvorak, Colemak, Workman, and a variant of the Maltron layout. But in terms of key legends, if we labeled it with something other than QWERTY, almost no one would buy it. We may offer a version with either blank legends or dots instead of legends so the LEDs shine through nicely, but that’s a little further down the road and depends on how many we sell.
KC: For readers who’re just learning about Keyboardio for the first time, can you summarize the project in your own words?
Jesse: So, I was taking time off explicitly to mess around with new things. I was assuming that one of the software projects I was playing with would be my next startup. I’ve done a bunch of things. I’ve spent a couple years as a project leader for the Perl programming language, one of my hobby projects was an email client for Android called K-9 that’s open source and has a terrifying couple million active users—
KC: You developed K-9?
Jesse: Yeah, K-9’s all my fault. I handed off project management when I started Keyboardio, but yes, that was me.
KC: Very cool! Anyway…
Jesse: Anyway, none of the keyboards I had were right. I’d been through pretty much everything, including a hand-wired ErgoDox, and nothing was right. I figured, “how hard could this be?”
For my first custom keyboard, I used OmniGraffle on the Mac, which is a Visio-esque drawing program. I figured out where I wanted the switches and cut the holes, and sent it off to the local laser cutter. They cut it out of a sheet of acrylic, and then I super glued switches in it and hand wired it. I needed something to get it off the ground, so I went to Michael’s Arts & Crafts and bought a picture frame. They didn’t understand why I didn’t want the glass. It was desperately wrong, but it worked. It was very similar to the Maltron Executive layout, if you’ve seen that, but flat. All the columns were straight, it was spread apart so my hands were about shoulder-width. It was all built around a Teensy, and it was running TMK, I think. It was OK but not right, so I started asking myself what I could do better.
I started playing with a couple things, including a symmetric 60% layout, which I cut out of plywood. That was the first thing the laser cutter shop let me do myself. For everything else I was using acrylic—you can find them on my Flickr—the Mark 2, 3, 4. I was playing with a columnar layout where the keys are offset a little but without too much splay, and thumb layouts I found in paperwork from a Japanese effort from the early ’90s called the TRON project. TRON was this alternate computing infrastructure from Japan, everything from pocket computing up to grid computing. They basically looked at everything we did in the West, and asked “what can we do better knowing what we know now?” So their keyboard was based on hand studies. I have photocopies of these papers from ’90s Japanese journals. They have little circles where a survey of 100 Japanese hands could reach to.
That was the first time I’d seen that thumb layout, with the two arcs, and it made sense to me. Your thumb doesn’t move according to a grid layout, like on the Kinesis or ErgoDox. Your thumb moves in an arc, and so that’s where we started playing with that arc, and we tweaked the spacing and angles over time.
It was probably right around prototype 3 or 4 when I would start taking the keyboard to a cafe and people would start asking me, “Hey, where can I buy one of those?” And it got to the point where it was a little hard to get work done in cafes because people kept interrupting me to ask about the keyboard.
We switched from laser cutting acrylic to 3-D printing when our 3-D printer finally showed up, which let me play more with non-flat surfaces. I made a lot of mistakes that one shouldn’t make during keyboard design. Sharp points under your wrists—definite no-no. The 3-D printer also let us play with having the thumb keys and finger keys at different angles. It let us play with iterating on key layout a lot faster. One of the weird things was that as we’d bring 3-D printed keyboards out in public, people got really excited about the fact that we had a 3-D printer, not about the fact that we had a keyboard.
And then, the 3-D printer caught fire, and that was the end of 3-D printing. So we went back to laser cutting acrylic prototypes. At this point, I figured we might sell about 50 of them if we did a Kickstarter, because there was definitely a little bit of interest.
KC: What time frame was this?
Jesse: A year and a half ago, maybe, a little bit before the “building a keyboard” blog post (part 2). So there was a big snowstorm in Massachusetts that delayed an acrylic order from McMaster Carr, but I didn’t want to miss my slot at the laser cutting shop, so I went to a local lumber supplier and bought a bunch of plywood. And so, I cut a keyboard out of plywood. As it turned out, the wood looked nicer, showed fewer imperfections, and felt nicer than the acrylic, and that was the point we realized wood might be the right material. So I posted these blog posts about how to build your own keyboard, and at the very last minute added “sign-up to our mailing list” as a link at the top of the blog post. It took me about 10 hours to write the first one, and went to bed. When I woke up, it was sitting at the top of Hacker News. There were 30,000 clickthroughs to the blog post, and 1,000 people had signed up to our mailing list. And a week later, when I wrote Part 2, it happened again. And that was when we realized there might be more than 50 people that would buy them. At that point it was still a hobby project, it was me and my wife Kaia, and we’d done a run of 6 or 8 identical wooden keyboards that were very cartoon-shaped. They had circuit boards and Arduinos instead of Teensies, but looking back at them, they weren’t that well done.
So, we sent out a mail to our mailing list, and said, “we have two test units, would anybody be interested in being a beta tester? Tell us why we should pick you.” We sent it out around 5 p.m. on a Friday, which is not the right time to mail your mailing list, but by Sunday morning we had over 100 replies from applicants, people who’d written weird Arthurian fanfic, people who’d written pages of prose, one guy even recorded a music video.
KC: Did he get a keyboard?
Jesse: He got a keyboard. And this was right around when we got into Highway1, which is a hardware incubator in San Francisco. We applied on a lark, but ended up moving out to California. We found out we got in on day 2 of a two-week vacation to South America, and when we got in they told us that we had 3 weeks to move to California. So we got home, had basically a week to move, and didn’t come back home for 8 months.
We actually had to incorporate the company from a friend’s couch in Buenos Aires, so that we could give the incubator the little bit of equity they take in exchange for cash and other inducements. So, the corporate birthday was one year and one week ago. And since then we’ve been through Highway1, where we learned an awful lot about manufacturing, and iterated through a complete product development cycle. That’s where you saw the aluminum keyboard we launched at Highway1 demo day, before realizing it just wasn’t quite right. It didn’t feel like we wanted. It was too heavy, it didn’t feel quite as warm and friendly as we wanted it to. We made some interesting design choices about key switches and keycaps at the time, and based on a bunch of research we went for an 18mm key spacing, which everyone says is more right than 19mm, but then we put it in front of actual people and found they couldn’t get their fingers onto the home row. And so, the Model 01 production went back to 19mm.
The keycaps we got help designing at Highway1 were shaped completely wrong. They got my verbal specs and totally satisfied them, but that was the first time we had fully custom keycaps and we’ve learned a lot since then.
After Highway1, we’ve been working toward something we can take to Kickstarter, and we’ve believed we’re 4 to 5 weeks away from Kickstarter for the last 5 months. Part of this is because it’s really important to do as much as we can before we go to crowdfunding. We’d much rather be very late to ask for people’s money rather than be late to deliver once we have the money. We’ve definitely got a whole lot of “Shut up and take my money,” as well as a little bit of “Why are you so late, you guys clearly don’t know what you’re doing.” And the answer is no, no first time hardware manufacturer knows what they’re doing, and we’re taking our time so we can do it well and be confident that when we quote a price and promise a delivery date, we can earn enough money to keep eating and keep our promises. Because in reality, as a small keyboard startup, we probably only get one shot at this. If we screw this up, there will not be a Keyboardio Model 02. We are very much building for a long haul, and we’d like to be a viable business that makes money by selling products to customers at a profit, and that the customers then want to rebuy, ideally for other computers or people, and not 6 months later.
KC: I wish every group on Kickstarter and Indiegogo had that attitude. Unfortunately, that’s not the case.
Jesse: A lot of it is that we’re lucky that both Kaia and I have been in business before, and we had a sense of how little we knew. Though, if I had a sense of how actually very little I knew when we first got started, I never would have gotten started. I had to learn how to solder, how to do baby 3-D CAD, how to 3-D print, how to laser cut. I’ve had to learn embedded C and C++, I’ve had to learn the basics of circuit board design. I’ve had to learn more real CAD modeling as I took ownership of the keycap design. I’ve had to learn a whole bunch about mechanical design just to work with the people who are helping; CNC milling for prototyping. It’s… real different than software. The way I like to describe software is that your MVP, your minimum viable product, is an auto-updater. But in hardware, your minimum viable product has to be a real, working product. If I ship a product that’s missing a couple pins on the microcontroller, I can’t just send you a software update that magically makes it better.
KC: That’s a huge selection of skills you’ve had to learn. And you mentioned you’ve been through Highway1, a hardware incubator. So what exactly does that entail? Did they give you access to machinery or experts?
Jesse: So, Highway1 has engineers on staff, both electrical and mechanical, who can advise us and step in when it’s a little over our heads. They’re owned by a company called PCH International, which is a giant contract manufacturer. It’s an Irish company operating out of Shenzhen. So we had access to PCH’s internal resources, including world-class engineers, plastics people, sourcing folks, etc. They took us to Shenzhen for two weeks. We got to take factory tours and see how things are really made. They helped us walk through the design process to make sure we had everything we needed. They helped us with pitch coaching.
A lot of the other teams at Highway1 are raising millions of dollars. In comparison, we have some money, it’s not all out of pocket, but not that much. Highway1 exposed us to a lot of people with complementary skills who aren’t competitors, so I’ve been able to lean on some of the other Highway1 teams when I needed advanced electrical engineering help, and in return I’ve been able to help with some software things. But Highway1 made us a real company. Until we got there, it was a hobby project and we weren’t intending to be a real company.
It’s a workspace. They have 3-D printers, an actual workshop, an electronics workshop. There are tools you get free access to, on-site experts, weekly reviews, they bring in mentors and guest lecturers a couple times a week to talk about something. Like, “here’s how we did really well/badly at crowd funding,” “here are things you need to be worried about related to intellectual property,” and so on.
KC: And they give you some cash up front?
Jesse: In our batch, they gave us $20k up front, and I think current teams are getting $50k up front. It’s not a lot of money, but it’s a lot cheaper to do hardware today than it used to be. We’ve been able to do things on a shoestring that would have been impossible ten years ago. But in exchange for all of that, they get a slice of equity. I think the number they’re quoting now is 4-6%, or 3-7%, depending on the team, how far along they are, how good their negotiating skills are, that kind of thing.
KC: And you said you’ve been through it?
Jesse: It’s a 4-month program. We were in Batch 2, and they’re now doing Batch 4. They’re still helpful and friendly when we need things. They claim it’s a family and it does seem to be that way which is nice. The current batch is actually going to be the subject of a new SyFy reality series, “The Bazillion Dollar Club.” I’m really glad we weren’t in that. I can’t imagine trying to do a startup and also having a camera crew.
KC: Yeah. So having been through that experience, how was it?
Jesse: It was incredibly useful. If we hadn’t been through it, we would have been on Kickstarter, thrilled we’d sold 120 keyboards. They would have cost several hundred dollars, made of plywood, and we would’ve made 10 bucks a keyboard or something. Now, we’re much better set up to make something that’s a much better product that we can be much prouder of, and it’ll be much closer to our vision, and executed much better.
KC: Switching gears a little bit… I know there’s probably no such thing as a normal day for you, but to the extent possible, can you walk me through an average workday at Keyboardio?
Jesse: There’s absolutely nothing typical about any of my workdays, and that’s one of the things I like about tiny startups. Our hours are flexible, which is nice, my wife is my cofounder, which means for better or worse we are somewhat working all the time. Pretty much the first thing that happens when I wake up, is I roll over and check my mail to see if there’s something urgent I need to deal with. Typically we get up between 10 and 11, which has a lot to do with our bedtime being in between 2 and 3 in the morning.
Morning is usually email triage, pushing various conversations forward, responding to inquiries that have come in overnight, checking in with vendors or contractors. One of the things we do everyday is, we sit down and make what we call a “today list.” This is a thing I’ve been doing for a lot of years, it’s a one-day to-do list. Write up a list of things you need to do today, and the list expires 18 hours later. It helps keep us focused on what’s actually top of mind. A lot of times in the early afternoon, we’ll go work from a cafe. Getting out of the house everyday is important. In the afternoons, I’ve been spending a lot of time working on our firmware, or electrical CAD recently. As it gets down toward evening, these days I’ve been spending lots of time doing bits of soldering here and there, as we’re getting all the parts together for our 20-unit beta run. We have all the mechanical for the run, all the LED boards, what we don’t have is the actual boards with the microcontroller, diodes, and key switches. The first batch of those boards were astonishingly bad, 2 months late with a 40% failure rate. Right now, getting these boards replaced is one of our highest priorities, because it’s the thing that is driving when we can get our units to folks we want to have in our Kickstarter videos. Until we can get them their units, that’s our big blocker.
Other things… we’ve been working through our product documentation, packets we can hand to manufacturers that tell them everything they need to make our products. We’ve been exploring both working with individual vendors for each part, as well as vendors that’ll do an entire keyboard, soup to nuts.
KC: What’s your MSRP?
Jesse: We are being careful not saying anything in public at this point, but we think it’ll be competitive with a Kinesis Advantage and we think you’ll get a lot more for your money.
KC: In an earlier article on keyboard ergonomics, I did a literature review of academic keyboard research I found on Google Scholar. It wasn’t much. Maybe everything is hidden behind pay walls. Did you do a lit review? What did you find, and how did you apply it to your work?
Jesse: A lot of it is hidden behind pay walls. I spent a good amount of time doing a lit review, and I also have the advantage of people sending me their caches of articles. At some point, we’ll make our keyboard document archive public, I just need to verify everything from a legal standpoint.
I’ve read through a lot of historical patents. The earliest patent I’ve managed to find for a keyboard with a split columnar layout is a mechanical typewriter from the 1890s. Walking it back, people will say we stole from the ErgoDox, but no, we stole from the exact same things the ErgoDox was stealing from. And “steal” is the wrong word. The ErgoDox appears to be based pretty directly on Kinesis keyboards, which in turn is based on Maltron keyboards, which is based on various earlier things including the Tron keyboard I mentioned at the beginning of our discussion. And so on.
A lot of the research is contradictory, so it’s a bit tough to identify specific applications of the reading to the Model 01. It’s important that we don’t do things that are actively bad for people. The keyboard is comfortable. One thing Kaia pointed out was that after a couple months of using the prototypes, I’d stopped taking Advil every day and wearing wrist braces. So that’s great anecdotal evidence, but obviously the plural of anecdote is not data.
There is a lot of straight-up contradictory research, and a lot of it is testing somebody’s specific design. One of the most useful things I heard is that most tests work on an hourly or weekly basis, but the metastudies suggest you actually need to be testing over a period of months.
KC: What specific design cues did you take from existing ergonomic keyboards? Did you look at the ErgoDox, and say, “we’ve got to have this feature,” and “this feature from Kinesis?” Are there any specific influences or things you actively tried to avoid?
Jesse: The biggest thing we were inspired by was the layout of the TRON keyboard. What we took from the TRON was different splay angles for different fingers. It was also important to us that it be split and adjustable, so one of the things we took from the IBM M15, which is probably my favorite keyboard of all time, was the adjustable ball joint. We started with that, but then we learned how hard they are to engineer and how often they fail, so that went by the wayside. The feet on the M15 are something we aspire to. They’re a little over-engineered, but we want to aspire to something that gives you that much adjustability. It is important to us that the two halves be physically connectable so you can have it in your lap, or shoulder width apart. That’s something the IBM does that no other ergonomic keyboards tend to be good at. Most of them are either fixed adjustable or split. Even the new Matias Ergo Pro is split. You can’t actually use it in your lap.
A lot of this was starting from something that works and then re-engineering it from first principles. As an example, for keycaps, we started off using commodity Signature Plastics DSA keycaps. Pretty soon, we realized we weren’t getting used to them, and profiled keycaps were better. It turns out that profiled keycaps are a huge improvement for typability. It dramatically drops error rates, makes it much easier for your fingers to hone, makes it easier to type, and so from there we started looking at custom keys. We knew we would have a couple custom keys, like the palm key, but we hadn’t initially been planning to do a whole custom keyset. It quickly became clear that to get the shapes and feel we wanted, we were going to have to do custom molds. We started from, how do your fingers move, what can you reach, what can you reach without pressing down any unintended keys? Initially we were going to go with square keycaps like everybody else, but now we’ve been working with Autodesk and I’ve probably spent a couple hundred hours designing keycaps so far. And they feel nice. People like them. But to do this, we’re going to have to get custom injection molds made. It’s something pretty much every manufacturer goes through for any plastic parts, but most products have 3, 4, or 5 plastic parts, while we’re going to have 64 unique parts. That’s tens of thousands of dollars in tooling. But once we’ve made the investment, it really brings down the cost per keycap, and lets us choose the plastic, the coating, how the laser engraving works.
You know how some ABS keycaps can get really shiny? As we’re been talking to senior plastics people about how a lot of people strongly prefer PBT because it doesn’t get shiny, they get this funny look, and as I explain more they say, “Oh no, what they prefer are keys that don’t use the cheap UV coating.” It’s all about the coating on the plastic, and there are orders of magnitude in difference in terms of price. Per keyboard, it’s a difference between $0.03 and $0.30. ABS might not be the right plastic for a lot of reasons, but “it gets shiny” is not one of those reasons.
KC: We’re running out of time, but one thing I’d like to do for my readers is recap the current state of the project. You’ve got a ton of information in your blog, but it can be a bit hard to go back and figure out what’s current.
Jesse: Part of that is, we’re a small team and putting a lot of our effort into maintaining a full public status of the exact state of things takes away from the time to actually do the engineering and management to get closer to Kickstarter. So the line in the sand, is, when we get to Kickstarter, we’ll take the time to fully document what you’re getting from them.
We’ve also learned an awful lot about building hardware and making commitments to people—thankfully, before we’ve taken their money. Because early on, we started from zero. I was a software guy, and Kaia has a business in banking and consulting. I’ve run companies before, she’s done the business side before, but neither of us had ever done a hardware project before. So early on, we had no idea how long this would take or how hard it would be, and that’s probably good because we never would’ve gotten started if we had. And, there have definitely been times where we made public commitments about what we’re doing, and then it turned out we couldn’t do them, and I feel awful when that happens, even though I know it’s par for the course. But especially, when it’s a product people want, I don’t want them to get their heart set on a feature it turns out we can’t deliver. But I’m happy to talk through the current state of things with you, so you can get a full picture of the state of things without having to go back through blog posts.
KC: Understood, and sounds great.
Jesse: The current status overall is that we have a complete and working design for the keyboards we’re going to Kickstarter with (the Model 01). That means we have the top wood enclosure, the aluminum key plates, the aluminum bottom of the enclosure, and aluminum center bars that connect the two halves either flat or with a 10-degree tent. We have the interconnect system. The keycaps we’re using are custom designed. For the 20 sample units, we’re CNC milling them, because it’s prohibitively expensive to injection mold them before everything is set in stone. It’s way more expensive per set, but much cheaper for only 20. We have feet, but we’re not happy with them yet. These 20 will go out with feet that aren’t good enough, but it’ll be corrected before Kickstarter. For PCBs, we have one with an LED under each key, capable of 16.8 million colors per LED. They are the gold standard for hobbyist RGB LEDs. We don’t have an Arduino on the board, but we have the same chip it uses, a ATMega32U4. We’ve made a choice to use the default Arduino bootloader and write our default firmware in Arduino C. So if you’re not afraid of Arduino, you can completely customize your keyboard’s firmware. We decided to do that instead of using an open source hobbyist firmware due to accessibility. If you know what you’re doing, you can flash TMK on this, you can probably flash HaaTa’s KLL, but we’re explicitly focusing on making it easy for folks who are hackers and programmers but not embedded people. We have the designs for the left and right PCBs, of course the first batch came back bad, but we’re getting them fixed.
Every single thing we’ve done will get another rev before manufacturing, it’s a process called design-for-manufacturing. There are things we can do to make it work better, be more robust, be cheaper and higher quality. There are some things we can do to make it more aesthetically pleasing.
We’ve committed to Matias switches for now. These boards will actually only slot Matias switches. Part of this is that it’s very easy to release new boards with different switch profiles, but also that the plate is only designed for ALPS because we want to have a plate with no weird gaps, so nothing can spill into your keyboard. We spent a lot of time thinking through our options on switches, and have tested more switches than I’d like to think about, really. I’ve been to the Kaihua (Kailh) factory in Shenzhen, sat down and talked with both sales and engineering about their switch quality, and to see if we could get them to make a switch that feels exactly like a Cherry Brown. They’re not interested, which is completely their prerogative. We’ve had some nice discussions with Cherry, but they’re too hard to source. If you’re going through a Cherry reseller, there’s an 18-month backlog, and if you’re not, you’re getting gouged on price. The prices I see for Cherry key switches are almost never less than double what they should be.
I was initially skeptical of Matias switches, but they’re nice. They’re quiet. You can use them in a meeting without people glaring at you. They’re available in quantity, and they actually glide smoother than the equivalent Cherry switch. They’re a little more wobbly, but the way Edgar Matias explains it, that’s to let you hit them off-center without any binding. So there’s a little bit of a perception issue that they might be cheap because they’re not completely sturdy, but the Matias switches are inexpensive, available, have a nice key feel, easier to 3-D print keycaps for when prototyping. The other thing is that they’re clear, which means you can mount an LED underneath.
KC: I think your keyboard will be the first retail product to have backlit Matias switches, unless KBParadise’s V80 MTS does it first.
Jesse: Well, I don’t want to spill the beans on anybody, but I’m aware of some other efforts. Right now, we actually have a 2 PCB sandwich to do this. It works, but we can save a lot of money, make the keyboard more robust, and make the LEDs brighter if we can get it all onto one PCB.
KC: Another issue that comes to mind for me is keycap interchangeability, but I guess since you’re designing your own keycaps and not expecting users to swap them, it’s fine.
Jesse: Yep, and I know that there are more and more people working on ALPS caps anyway.
KC: Sounds like you have your hands full right now. 5 weeks to Kickstarter, right?
Jesse: Yeah, it’s been that way for 5 months. We’re no longer promising deadlines now, but at some point we’ll start again, because we’re doing a coast-to-coast road show. We sent out a survey and asked people, where are you, should we visit, is there a group or hacker space nearby, what’s fun to do, where should we eat, so now we have a heat map of the country with places people want us to visit. Initially we thought that was going to be in late January or early February, but I’m glad it’s been delayed, because there are a lot of people up north.
KC: It’s been a cold winter.
KC: Got it. Well, very cool, and I know you have to get going so I won’t keep you any longer. Any final thoughts?
Jesse: I think we pretty much covered it!
KC: Great. Jesse, thank you very much for your time and I can’t wait for the Kickstarter campaign to go up.
On behalf of KeyChatter and its readers, I’d like to thank Jesse for his time. Check out Keyboardio’s official site, and make sure to register for their mailing list so you’re in the loop when the Model 01 goes up for sale.