Keeping the spirit of the “making stuff together” origins of the project, I thought I’d share some things that I’ve learned from the prototypes in a little review.
At the end, I’ll also describe a couple of the small changes that I’ll be doing to improve it for the final version.
Lets start with some better pictures though (sorry, white balance got a bit off for some pics):
Quality of the machining:There were some minor issues with one of the two prototypes (the violet one). It had some dents around the edges (looks like they clamped it too strong) and some scratches where visible.
What’s more, it was smoothed instead of bead-blasted, and the tapped holes were countersunk for some reason.
I’ve shown the vendor the pictures and they agreed that it’s unacceptable, and made no fuzz about it resolving the issue with me quickly and to my satisfaction.
You can see the dents bottom left and right edges in the pictures below - I would definitely not ship a case with these marks to you guys:
The silver/unanodized one has no marks, it has the desired surface finish, and overall has the quality that I would feel comfortable shipping.
Given their response and the quality of the silver one, I’m pretty confident that they are the right manufacturer.
I also am confident that such issues will not occur with the bigger order (and if they do, I’m confident they will be resolved easily).
One thing that I wasn’t sure on what to expect was how the seam would look between the two halves.
It’s not a seamless design, so the seam will always be visible, but if they are poorly machined they can look ugly.
For these prototypes, I’m pretty happy with how it looks overall - the seams are pretty uniform (which is the most important part imo).
The exception of course are the areas on the purple one where there are the dents.
For both, there are no overhangs or anything like that, the outer dimensions are machined nicely within the tolerances.
Although, since there is a little play in the clearance holes for the screws, one should pay a little attention that they are perfectly aligned when tightening the screws!
Look and feel:The purple one with the “smoothed” surface actually also feels really really nice to the touch.
However, the smooth surface also reflects the light differently than a matte surface would.
This is why the purple also looks a bit brighter in the pictures than it does in the renders.
I think the picture below gives a pretty decent idea of how it looks IRL:
The silver one, with the bead-blasted surface, feels a bit different to the touch.
One thing that I realized is that anodization also increases the hardness of the surface.
Since the silver one is only bead-blasted and not anodized, it feels a bit warmer, “grabbier” (but I wouldn’t call it “rough”/sandpapery).
Another observation here is that
we can expect the un-anodized version to be a little less scratch resistant than an anodized one (due to the lack of hardening).
That’s absolutely no problem if you plan to get it for adding your own treatment later (painting/powdercoating/cerakote), but you should be aware of this in case you plan to keep it as is.
Assembly:The assembly went exactly as I had expected, so I’m pretty happy with that.
It’s a little easier if you first mount the plate and then the PCB, compared to assembling plate+PCB+switches separately and mounting them in,
as you’ll have to thread those socket screws through the holes of the PCB, but it’s by no means complicated.
I only have prototypes of the hot-swap PCB here, but it holds so rigidly, I’d almost say that the PCB-to-top-case screw are also optional even for the hot-swap PCB.
I’m also really happy with all the tolerances & dimensions i.e. in the top case.
I have really no concern that any binding will occur between keycaps and case (something that I’ve had to do some hands-on fixes in the past with my plastic prototypes).
Blocker:The blocker fits just as intended and to me, it looks just as great as it already did in the renders.
I’ve tried using blockers before on the KBDfans 5° 60% case (to get the HHKB look) but those are really crap compared to how it looks here, because on the 5° case, the blockers attempt to sit flush with the top-surface (but they never really do just quite), plus they have these ridiculous chamfers.
I’ve noticed (and some people also commented on this already on the pictures I posted it above) that it looked a little better when it was a bit more sunken into the case (due to the silicone gasket I’ve evaluated inbetween plate and top case).
I’ll definitely update the design to just make them 0.5mm shallower so we get this look consistently.
Below is a shot of how it looks now, without the gasket:
LED Indicators:I’m really glad I made the change to these more subtle pinhole indicators. They absolutely don’t disturb the look at all even when not using the LEDs.
I usually have them on and configured to show a nice accent color complementary to the case.
I personally also really find using the LEDs to indicate some states of the keyboard quite useful (different layer indicators, indicator when putting the keyboard into reset, macro-recording mode, ...).
The default firmware that I’m writing (and that will be pre-flashed on each PCB I’m shipping through the group buy) will make full use of these LEDs in a way that’s more powerful than what you usually get from stock QMK firmwares.
(TODO: make a better pic with the LEDs on).Ergonomics:I’ve been using the BMEK for almost a year now as my daily driver and obviously still love it.
The only major change I can report on with the prototype in regard to ergonomics is the increased typing angle (used to be 3 degree, is now 6).
Trying both side by side, I’m equally happy with both. I guess other people have stronger feelings about typing angle than me
.
Sound & Experimental Dampening Inserts:Disclaimer: this stuff is still voodoo science to me, but I think I’ve taken a reasonably rigorous approach to evaluating this very subjective topic.As mentioned, I’ve prepared a number of foam and silicone inserts for different places on the keyboard, and of course we have the variable of using the keyboard with or without weight.
In total, I’ve prepared three plate + pcb assemblies.
- Two with Gateron Milky top linears - completely stock/unlubed/unfilmed etc (the only switch I had enough of to deck out two whole keyboards - recordings 1-10),
- one with Lubed Sakurios (Silent Linear, recordings 12 & 13). One of the two Gateron assemblies is with the foam insert between plate and PCB (recordings 1-5).
- As a plus, I also have made some recordings of the Gateron Milky Top linears on my (previously daily driver) POM BMEK prototype (recording #11), to get an idea how the sound differs on a completely different material (but very similar case in terms of shape).
All assemblies are decked out with Cherry profile ABS keycaps. Here is a list of recordings of me typing on the different configurations (1 means yes, 0 means no):
case | switches | Bottom Foam | Bottom Weight | PCB Foam | Plate Gasket | RECORDING |
alu | gateron milky top | 1 | 1 | 1 | 1 | 1 |
alu | gateron milky top | 0 | 1 | 1 | 1 | 2 |
alu | gateron milky top | 1 | 0 | 1 | 1 | 3 |
alu | gateron milky top | 0 | 0 | 1 | 1 | 4 |
alu | gateron milky top | 0 | 0 | 1 | 0 | 5 |
alu | gateron milky top | 1 | 1 | 0 | 0 | 6 |
alu | gateron milky top | 0 | 1 | 0 | 0 | 7 |
alu | gateron milky top | 0 | 0 | 0 | 0 | 8 |
alu | gateron milky top | 1 | 0 | 0 | 0 | 9 |
alu | gateron milky top | 0 | 0 | 0 | 1 | 10 |
pom | gateron milky top | 0 | 0 | 0 | 0 | 11 |
alu | lubed sakurios | 0 | 0 | 0 | 0 | 12 |
alu | lubed sakurios | 0 | 1 | 0 | 0 | 13 |
You can make up your own mind, but I feel that all the dampening options I've evaluated had very minor effect.95% of the differences I hear are explained by the combination of switches and bottom weight. The next more noticeable effect comes from adding the bottom foam.
Since the bottom foam is also something that is trivially made in a DIY fashion, I'm inclined to not offer extra dampening options in the GB - but ofc I will publish the patterns/laser cut files for people to experiment with for themselves!
Other “Experimental” configurations (Flex!)I never particularly cared for “flex”, and I’ve been using BMEK with a thick integrated plate (which almost offer almost the least amount of flex conceivable) on the PC and POM prototypes and am just happy with it.
And clearly, using all of the 8 screws of the plate in this design with the alu plate also results in a stiff and crisp setup.
However, there is nothing holding you back just using fewer screws on the plate to increase the flex. I’ve tried two combinations, with just 4 screws (red marks below) and 2 screws (blue marks below):
First off, I was pretty surprised how well everything sat together even with just two screws.
Second, both setups allow for a reasonable amount of flex, although you really get the most with the two screw setup:
https://i.imgur.com/4qfif4b.mp4Of course, if people really want a flex setup, instead of (just) using fewer screws, one would add some flex cuts into the plate.
We’ve had some talk about that on the DIY channel on the discord in regards to this, which lead to this experimental plate design ripped of from
hiney's "Alice Three-Quarter" plate:
(note: I'm not sure if we can offer this as an option, but you can vote on this in the "KIT" IC Form)
However, I’m still glad that I played around with it this way, as I have discovered a first real “case-design-problem” this way: In the video above, I’m showing the flex without mounting the top case to the bottom case.
With the prototype, it’s not possible to flex the board when fully assembled, as the standoff for the center case mounting screw in the bottom case is bigger than the clearance hole on the PCB, thus holding back any downwards motion!
This is an easy thing to fix though for the final version, I’ve just made it a little slimmer, and additionally increase the hole in the PCB slightly, to also make sure that - if it is desired to be flexible - there can also be no rubbing.
Even though a flexible setup is not the “original design”, this fix is important to me to make, as it will allow more room for experimentation down the road.
Changes to the design post-prototypes:- Lower the blocker height just slightly
- Raise the upper surface of the top case slightly to cover keycaps & switches better.
- Fix center mounting hole standoff diameter in bottom case to allow for flexible plate configurations.
- Kinda unrelated, but I've decided to add a iso-split-left-shift to the layout options (so I've updated plate & PCB for this). I know it might be an edge case, and I've stated before that I don't like it... But I've contemplated a bit, and since it was just 20 minutes of work to add it, I thought I might as well do it.
- Since I really am happy with how the blocker looks, I'm looking into the costs & design of an extra polished brass blocker.
You might wonder if there are any risks attached to adding these changes. With the big order, I will get review units and a couple of days to fix things in case I messed up. So I feel pretty confident these changes won't break anything/add delay.
Updates to the timelineFirst and foremost, I'll post the final two IC Forms tomorrow for the Color options and Kit configurations.
I think with these pictures and information, you guys can fill those out much more informed.
Wwith the responses from these IC Forms, I should be able to get final quotes from the manufacturers, and I can finally decide on the price (I think everybody will be happy with what it's going to be!).
I'm also close to ordering the packaging material / boxes next week, as well as the final set of pre-production prototypes of the PCB (I want to have some samples of the solder PCB as well and validate that it works before producing the big batch).
Meanwhile, there is also some "legal stuff" to figure out, as this is technically a business operation.
I prefer to do things according to the books, and it looks like I need to apply for being a small business before I'm officially able to write invoices for the GB.