geekhack
geekhack Projects => Making Stuff Together! => Topic started by: JPG on Fri, 15 November 2013, 12:28:58
-
Lately, I have been curious about why people always use the old controller on the model F and plug a converter at the end of it. So I read about the model F switch, learned about how it buckles and mostly that instead of forming a closed/open circuit, it changes the capacitance by moving a part between 2 sensors.
But, I want to know why no one has tried to use a modern controller wired directly (with a matrix or not) to the capacitance sensors to determine if each switch is actuated or not.
It seems to me that if we could just add the part that would detect capacitance, then all the code from the modern controllers could be used as if it was some cherry switches.
Is there a flaw in my reasoning? Is there a big challenge in detecting the capacitance from a sensor?
-
I thought I remembered a discussion on this from a while back, and unfortunately, all of the attachments are gone, but here it is anyway... (http://geekhack.org/index.php?topic=21459.0)
-
I'd think it would be possible since it was done by xwhatsit for the Beam Spring 5251 and 3278
http://deskthority.net/workshop-f7/beamspring-usb-controller-t6044.html
-
Thx for the links, very interesting.
-
We have been working on an F replacement controller for a while and will announce it properly once it's done.
-
The thing is that a converter is dead-easy thanks to Soarer's firmware: buy a Teensy 2, solder 4 wires to a connector, upload the firmware and plug it in.
Why do it the hard way when it can be so easy (hardware fanatics not withstanding)?
-
The thing is that a converter is dead-easy thanks to Soarer's firmware: buy a Teensy 2, solder 4 wires to a connector, upload the firmware and plug it in.
Why do it the hard way when it can be so easy (hardware fanatics not withstanding)?
Then you could make a new pcb with the layout of your choice, with function layers, leds, whatever you want. Simply put, you get back full control on the features you want to implement in your keyboard while being able to use the mythic F switches.
-
Then you could make a new pcb with the layout of your choice,
Wow!
You have the means to do that?
If Model M SSKs are pulling $400+, a Model F SSK should be worth A LOT more!
-
Then you could make a new pcb with the layout of your choice,
Wow!
You have the means to do that?
If Model M SSKs are pulling $400+, a Model F SSK should be worth A LOT more!
I have absolutely no means of doing this. Well I don't really know how hard it would be. But I know that if we can't detect capacitance with a controller, then we can't do anything at all. But if we manage to detect a switch actuation, then we only have to figure the other challenges!
I don't even have a model F btw (I am considering getting a cheap one). But one thing I know is that I would not be using it with such a big layout. I went from fullsize to TKL and I start finding my TKL a little big.
-
The thing is that a converter is dead-easy thanks to Soarer's firmware: buy a Teensy 2, solder 4 wires to a connector, upload the firmware and plug it in.
Why do it the hard way when it can be so easy (hardware fanatics not withstanding)?
Because there are a lot of Fs that cannot be properly converted due to not sending break keycodes on some keys (usually shifts and other modifiers), just make. Meaning, it registers the press but not the release. Converters can't help with this, and one has to replace the controller or the micro entirely.
Then you could make a new pcb with the layout of your choice, with function layers, leds, whatever you want. Simply put, you get back full control on the features you want to implement in your keyboard while being able to use the mythic F switches.
Well yes, in theory. Except that building a capacitive matrix is more difficult than a regular switched matrix (such as for Cherry or Alps) due to shielding requirements, the switches require a plate, and one would have to create new barrels and flippies. Perhaps the springs as well, though some can be sourced from Unicomp. Reusing existing barrels is an option, but there's a finite supply of Fs, and it's not a viable long-term approach, not to mention that it requires making curved front and back plates since the barrels themselves are curved.
So it's doable, but requires significant investment into recreating the required parts. On the upside, said parts are relatively simple compared to fully integrated switches.
-
We have been working on an F replacement controller for a while and will announce it properly once it's done.
if you need a test PCB, i'm getting pretty good at making prototypes.