Merry crimbo to AcidFire and everyone else eagerly waiting on this divine piece of kit.
Merry Christmas from Australia! There must be geeks all over the world dying waiting for this keyboard. I know I am!!
I'm mainly anxious to get hacking on the software.
Merry Christmas to everyone! Apologies for not getting on to reply over my break, was quite busy with family this past week.
I'm still working on the software side, and I have to admit I'm struggling a bit with the ARM side of things. While I have a firm handle on the electronics, getting the firmware up and running has been a bit more troublesome. I may enlist the aid of a few of the people who applied as beta testers who have experience programming for the Cortex M0/M Family or ARM in general to get the first version out the door. If anyone is interested in helping me out with this, shoot me a PM.
Show Image
Installing diodes made it super simple to connect my columns. The best part is, it takes minimal effort to remove the hot glue & solder, so when I do finally get PCBs made it won't take me long to recover my keys for the new board.
How viable is it to make a board like this with no PCB? Seems really awesome and cheap :thumb:
Well, If you're making one board for yourself it is the way to go, but if you're going to make lots, not so.
Pretty much this. This was a bit time intensive, but was fun to work on over an afternoon. I will say though, that unless your doing a different layout, the PCBs are still the way to go in terms of time efficiency, especially since the boards won't cost a lot and I do plan to make completely bare boards available for those who want to truely build from scratch.
I'm curious, what controllers do the rgb leds that you're using have?
I'm currently using standard LEDs that are connected to NXP PCA9685 12-bit 16 Channel i2c LED controllers. These allow the LEDs to produce 4096 color per channel, instead of 256, or 68,719,476,736 colors as opposed to 16,777,216 (8-bit.)
For the single color backlights, the LEDs are connected to NXP PCA9674 8 Channel & PCA9675 16 Channel 8-bit i2c LED controllers.
Both of these chips have a couple of distinct advantages, chief among them being an internal PWM clock, meaning they don't need to be constantly updated like the TLC5940. Basically that means they're set & forget, which is very processor friendly.
I try to look at the situation in a positive light.
Because this design looks so good, it has actually put a HALT to my keyboard spending: I'm reluctant to spend more money because I know something better (this project) will come to fruition, eventually!
So the situation is actually saving me money.
Of course, once released, I can't guarantee that I won't go broke spending money on all these yummy modules...
No pressure! I've heard this from a number of people and its very encouraging to know that there are people eagerly awaiting what I'm working on. With what I've seen so far with my numbers, I don't think it's going to break the bank either.
AcidFire, will there be room in the layout for a few F-keys? I got an ErgoDox kit recently and ended up using most of the thumb-clusters for F-keys (F1 through F6), but this is less than ideal due to being split between two hands. I don't need all 12, but room for F1-F4 in one block or F1-F8 in one/two blocks would be awesome!
In any case, really looking forward to what you come up with, especially the compact (travel) variant!
Currently, there are two different solutions to this request that are part of the current revision.
First, the thumb cluster can be set up with all single keys, giving you a ton of options at your thumbs. There will also be an optional top row of keys that can be added that by default would act as F keys. Because of the minimal number of parts required outside of the key switches, I'm hoping to keep this add on to $25 or less. As with the pricing of everything on this project, most of that is the cost of the key switches, and if I can get the pricing I want, I'm looking at a price of roughly $16-18, but that wouldn't include caps.
Depending on how the sizing works out there will most likely be space available in the cases for the desktop sizes, however the travel variants would probably need a different casing. I'll be using the kickstarter campaign to judge how popular a travel variant with F keys would be.
SHUT UP AND TAKE MY MONEY!! Awesome project.
The addition of assignable scroll wheels on both sides would make this project my perfect keyboard. If you don't do it, there's a good chance that I'll try to add them to mine when I buy it. I've always wanted a keyboard with a scrollwheel. When I'm surfing, the arrows are too slow, pgup/down is too jumpy, so it's back over to the mouse for the scrollwheel all the time.
Acidfire, I just hope that you're not spreading yourself too thin with all these features and options. Please don't be offended by this, I'm just very keen to see this come to fruition - KISSASS Keep it simple stupid and ship something! I totally understand that you don't want to limit yourself by underdesigning the initial offering though.
Keep up the good work! It looks awesome.
I'm currently looking at three accessories/addons that sit at the thumbs.
1) Scroll Wheels - The code/electronics for this are actually quite simple, mechanically however it's a bit more of a pain as the wheels need to be sourced, mounting worked out, etc which all needs to be set up within an enclosure that would allow an end user to add it to their set up.
2) Jog Wheels - A horizontal version of the scroll wheels with a high precision, the problem mostly comes with how much space a wheel needs. I do however see this + the RGB backlight being quite useful to editors.
3) Mouse/Trackball/Trackpoint - This is the most requested option that currently isn't being included in the base project, mainly due to cost & sourcing (mouse actions will be supported in software however.) I think this may also provide a viable alternative to the scroll wheel, since it should be possible to map the mouse to the scroll functions.
As for KISSASS, this is something that has been guiding the project from the start. The only major revisions/feature changes that have been made are those core to the functionality of the project. I have a long list of additions and features that can be added due to the open nature of the firmware & modularity of the hardware.
---------------------------------
Ongoing, I realize that I haven't released a video explaining how the new key scanning/auto config works. Unfortunately the green LEDs I had originally purchased for testing have a terrible viewing angle and as the previous video showed, tend to flare badly. I have some nice diffused ones included in our next parts order at which point I'll be not only posting a video of how it works, but another of what the backlight is capable of.
I've also nearly completed the new versions of the PCBs, including the seperate control PCBs for the keys & backlight. This also includes the new thumb clusters which can be customized in a number of different configurations. This does present a problem with doing RGB backlight plates unfortunately, so most likely the plates that will be released will be voted on which layouts to do first.
As I mentioned previously I've managed to reduce the cost of the RGB addon significantly, and am currently in the design stage of the PCB layout for these boards. While they most likely won't go in with the next batch of board designs, I intend to get them in within the next 30 days so that they can be made a part of the kickstarter campaign.
On the case design side, one of the guys in our office has offered to teach me the basics of both Inventor & Solidworks, which should hopefully result in a workable case design in the next couple of weeks. Once I have that done I expect to be able to get quotes done and have a final price for the kit, not to mention being able to 3D print the design for active testing.
I know a number of you are anxious to help me beta test and we're definitely close to that mark. As previously mentioned, testers will only be paying for the cost of materials for their kits. While most of what would be included should be considered production ready parts, the final case will not be, partially since the final design will be based on their feedback. Instead, 3D printed and hand finished (sanded) cases will be sent for testing, with the final case being made available when ready at cost.