deskthority

Here's what likely happened: from fresh install, build mode switched to "Debug" (supposed to be specified in workspace!) - effectively screwing up ALL build settings, include paths among them.
Fix: download the project anew, look at the toolbar. If the combo box on the right says "Debug" - switch to "Release". Then compile as described in README.

I'll create a windows VM and will try to reproduce this.

Here is the completed conversion of the IBM 3277 beamer while using this controller. Thanks again to DMA for developing this firmware and helping me out with setting it up!

photos-f62/ibm-3277-mini-beamspring-key ... ml#p400493

hansichen wrote: ↑Here is the completed conversion of the IBM 3277 beamer while using this controller. Thanks again to DMA for developing this firmware and helping me out with setting it up!

photos-f62/ibm-3277-mini-beamspring-key ... ml#p400493

Oh, it just occured to me that you used emdude's PCB. In which case the shim for the connector could've been just a piece of paper folded couple of times. emdude's PCB have both sides routed. Original PCB, on the other hand, has rows on one side and columns on the other and no contact between sides.

Oooh, I'm not wireless crazy, but I know there are some people who are!

I'll just leave this here: https://www.instagram.com/p/BgXrNSpBzLa

Wow, that's awesome! Imagine a F62 or F77 with bluetooth I'm really looking forward to the documentation and the first results about battery life

suddenly have this idea to make 20x20mm castellated module (1.27mm pitch) and just soldering it to PCB with whatever form factor is required. Or even to the sense card.

I have been playing around with a one-piece board-to-board connector from Samtec. The low profile version is just 3mm tall so I am hoping it will work sandwiched between the sense card and top plate. I plan on using the 15x2 version solder side to the controller (compression side on the sense card). USB connector and LEDs on the small controller side too, so there will be no soldering on the sense card at all.
http://suddendocs.samtec.com/catalog_en ... i06_10.pdf

For the sense card, I am looking to go with 4 layer board with rows and columns on the top two layers, ground flood on the 3rd layer and traces to controller connector on the bottom. With this it is possible to have a "frameless" switch matrix. The prepreg between the first two layers is much thinner than FR4 on a 2-layer board, so bonus there.

wcass wrote: ↑I have been playing around with a one-piece board-to-board connector from Samtec.

I don't understand how it's fixed on the sense card.
Also http://suddendocs.samtec.com/catalog_english/sei.pdf - 1.65mm height, up to 30 pins. 30 pins is enough for everyone.

wcass wrote: ↑it is possible to have a "frameless" switch matrix.

It is possible not to have frames around the pads with 2 layer board, I checked. It will have some crosstalk, not too much of it. As for "much thinner" - vias work pretty well for that (also experimentally checked) - like 5x more swing, IIRC.

I don't know if 4-layer with it's big capacitance on sense lines will work. May be better, may be worse, and testing that is pretty expensive. If some secret santa's experiments are representative - should be better, but the sense card is the most expensive component, doubling the price of it doesn't seem like a good idea to me.

The connector works very much like pogo pins; soldered on one side and "compression contacts" on the other. Think of the contacts to the sim in a cell phone. You just need something holding the two PCB together (like a top and bottom plate). Alignment is assisted by locator studs on both PCBs.

And by "frameless" i mean the PCB does not extend beyond the the size of the switches. All previous capacitive buckling spring keyboards have needed a few mm of space outside of the key area (top for the columns, one side for the rows) to route traces to the controller because routing under the keys would intersect the capacitive pads. Four layer PCBs are not that much more expensive. I am planning on ordering some soon and would be happy to send you one. I wanted to wait until i had the thing built before starting a thread, but ...

wcass wrote: I wanted to wait until i had the thing built before starting a thread, but ...

I'm looking forward to the build thread!

Feature parity! So I kept my part of the promise, finally.

Macros! Congratulations and thanks for all your hard work. I hope to have a need for a few CommonSense controllers one of these days.

Techno Trousers wrote: ↑Macros! Congratulations and thanks for all your hard work. I hope to have a need for a few CommonSense controllers one of these days.

I think there is a world market for about five CommonSense controllers.

https://www.instagram.com/p/Bg73h9lBkhN/
Testing of the proximity sensor. Made out of outer plates of F122 and a piece of wire.
Detects me hand from 5cm away. CS doesn't seem to mind the new neighbors.
Doesn't react to metallic objects.
Great way to save the battery.

That's really cool! So a Bluetooth keyboard might not even need an on/off switch?

Techno Trousers wrote: ↑That's really cool! So a Bluetooth keyboard might not even need an on/off switch?

If there will be one.

But really - there must be on/off switch, because BLE will eat the battery maintaining connection/advertising (and you don't want your keyboard to turning off every time you take your hands off it) - also proximity sensor eats battery too. So if you're going away for a week..

Also so far I'm unable to reproduce that behavior. It worked with example firmware, but no longer does.

https://www.instagram.com/p/BhLAcDchf-k/
First BLE model F in the world.
It's typing gibberish because it's an XTant layout and profile soldered to F122 PCB.

But it has 18 KRO. Over BLE. Which is, I'm reasonably confident, is also first in the world.

Very cool.

JP! wrote: ↑Very cool.

Bluetooth Beamdesk!

Awesome! I'm starting to think perhaps I am interested in a Blootoof capacitive BS keyboard after all.

DMA wrote: ↑https://www.instagram.com/p/BhLAcDchf-k/
First BLE model F in the world.
It's typing gibberish because it's an XTant layout and profile soldered to F122 PCB.

But it has 18 KRO. Over BLE. Which is, I'm reasonably confident, is also first in the world.

Awesome to see the progress! I guess something like 6kro is perfectly fine for most people but it's interesting to see what you can do with the firmware.

hansichen wrote: ↑ Awesome to see the progress! I guess something like 6kro is perfectly fine for most people but it's interesting to see what you can do with the firmware.

Anything above 10 is stupid - you only have 10 fingers. But, for some reason, people demand "Full NKRO".

What if someone wants to use the faceroll technique?!

I like to hold down two keys with each finger when I play games. /s

Techno Trousers wrote: ↑I like to hold down two keys with each finger when I play games. /s

Why 2 when you can hold 3?

Srsly tho - going above 18 keys will require BLE 4.2 and also the longer your packets - the more chance they'll be garbled while transmitted. Also will use more power.

DMA wrote: ↑https://www.instagram.com/p/BhLAcDchf-k/
First BLE model F in the world.
It's typing gibberish because it's an XTant layout and profile soldered to F122 PCB.

But it has 18 KRO. Over BLE. Which is, I'm reasonably confident, is also first in the world.

outstanding!

18x13.5mm castellated module for soldering to breakout PCB.

Nice. Smaller than 1U. It could fit in that space between Esc and F1.
I will see if I can draft a matrix up.
What layout would you like to see? 104, SSK, or something else?