Would It Be Possible To...

[Lethal Squirrel]

...Wire the Led's in the switches in such a way that they light up when the key is pressed?

[pitashen]

I don't think it is impossible as long as you got the know how and skills. A separate circuit board to regulate/control the lights would probably needed at the least.

[F u r u y á]

Now we'll be only missing olfactible and gustable feedbacks

[Lethal Squirrel]

xD exactly what I was thinking.

I wonder how it would be possible to taste each keystroke

[F u r u y á]

good question, maybe chewing gum that releases flavor at the keystroke?

typing at 150wpm would be an explosion of flavors

[mich]

Connect backlight LEDs (maybe in series with some resistors) where NKRO diodes are supposed to be connected and you are done.

[dorkvader]

Connect backlight LEDs (maybe in series with some resistors) where NKRO diodes are supposed to be connected and you are done.

This is it. I was going to say to wire them in series with the switch, so that pressing the switch (and completing the circuit) turns on the diode, but this is effectively the same.

Problems include:
not enough current
putting two switches on at the same time will halve the brightness (depending on switch location, etc)
Voltage drop/impedance matching (I suspect this won't be an issue if you just replace the NKRO diodes)

[rodtang]

Depends on what swtiches you are thinking about doing it to.

[dorkvader]

Depends on what swtiches you are thinking about doing it to.

What do you mean? I would imagine it's more a matter of what board layout and controller you have than the switches. I guess it wouldn't work for capacitive ones, but it should be the same concept for any NO SPST keyswitches.

You could even do it with NC ones, though it'd be harder.

Here's an idea. The switch turns a transistor "on" and that causes the LED to light up. LED is driven by the +5V from the USB hub. You'd want a constant current power regulator for it, or something to deal with the current drop from multiple LED's, but I still see that as the biggest issue for now.

[rodtang]

If you for example have cherry mx switches you could do some fancy wiring to the switch.

[Findecanor]

How about using a separate matrix just for the LEDs and let the µC strobe them (turn on/off really fast) instead of letting them be on all the time.
No extra component per switch. Some more components per line in the matrix and more pins required on the µC.
That would allow you to do lots of fancy light tricks as well. Not straightforward to program the µC, though.

Now we'll be only missing olfactible and gustable feedbacks

Wasn't it Apple who had claimed that they had made a lickable interface, already? :-þ

[LEGITBALLIN]

you wouldnt be ablt to see the key since your finger will be on it

[dewcorps]

What about the possibility of taking a ducky, razer, etc with pre-existing backlight, and modifying the driver to have the key light up on press, and then fade to black over the period of two seconds or so?

[dorkvader]

What about the possibility of taking a ducky, razer, etc with pre-existing backlight, and modifying the driver to have the key light up on press, and then fade to black over the period of two seconds or so?

Sure, once you have the light on keypress capability, you can just add a capacitor. You can tune a resistor in parallel with it to get the time right.

Hm, I think I'd rather have it where hitting a key "add's" it to the lights on, and you can have a bigger capacitor right at the powersupply, then it'll fade continuously until a key is pressed again. You can then connect a potentiometer to it and control the delay.

Or, once you have the delay right, some surface-mount capacitor and resistor you can have keys fade individually. It'd be pretty awesome.

[Lethal Squirrel]

guys, let me explain a bit more what I am trying to accomplish.

Basically. I want to make it so when the key is pressed the light turns on, then when the key is released the light turns off. not a toggle, but a momentary light I guess you could call it (i don't know the electrical terms well)

I'm not sure if you guys thought that's what I meant or not, but that is what I plan on trying.
I was going to get the translucent keys from the group buy. the ones that look like normal black keys when there is no light shining through them. i was going to attempt this mod on my black widow.

i just wanted to know if it was technically possible without addition circuit boards, without me tearing up the keyboard first.

[nar]

I've been wanting a keyboard like that for my fighting games ever since I saw the Beatmania IIDX arcade cabinets that have keys that do that.

I know there a ton of tutorials on how to do this for arcade sticks, but those are just either wiring the leds to the switch contacts or to a separate controller.
In fact, the easiest way would probably be to wire all the leds to a controller and have it independent from the actual keyboard and use software to tell the controller what key was pressed. Or possibly you could hook that controller up to one of those DIY PS/2 USB converters and have everything done in firmware. Either way, you'd have to wire every single key's LED.

If a keyboard had lighting that was customizable at a key level, I assume you could just hack the driver and implement everything in software. Not sure there is any keyboard like that though.

[funkymeeba]

This is a pretty neat idea. A friend and I have some ideas on how we could go about doing something like this. I might buy myself a numpad with whatever MX switches on it to play around with. If I end up doing this and good progress is made, I'll be sure to share it.

Although now that I think about it, using a separate PCB and making it software controlled could create some interesting possibilities. It's something I'd be interested in looking into.

[dorkvader]

Possible but not as simple as wiring in series with the switch contacts.  In fact it'll be a real PITA.

Why not? Is the voltage drop across the diode an issue?

[mich]

I've just measured some old BTC keyboard controller I had laying around. Unfortunately ripster is right - the scanning pull-ups are way to week to drive a LED, they only pass about 1μΑ of current. That is, at least in this particular keyboard, but in others probably too.

[dorkvader]

I've just measured some old BTC keyboard controller I had laying around. Unfortunately ripster is right - the scanning pull-ups are way to week to drive a LED, they only pass about 1μΑ of current. That is, at least in this particular keyboard, but in others probably too.

Ah, that makes sense.

@ripster, I just put the multimeter in amps mode, and bridged a keyswitch. I'm am empiricist, so naturally I don't believe you. Luckily, I have the equiptment to test it.

Also, I doubt 1 microamp is enough to flip a transistor accurately enough, so you probably can't really do that ether.
---
Right, so I'm seeing one of those large teensy's. You just have the LED's wired into the Teensy, and have it flip them off and on when the right key is pressed. You'd even be able to do duty cycle, fading, etc.
Drill small holes in the PCB where the LED should go, then solder jumper wires to the extra Teensy pins.
---
Hmm, can you have a teensy interpreter, that "intercept's" the signal the normal keyboard is putting out, reads which key is pressed, and lights the proper LED based on that information?