Page 1 of 1

Measured Capacitance of a Model F Switch / Making the 3178 switch work

Posted: 24 Aug 2017, 03:40
by orihalcon
Hello All,

I'm refurbishing a 3178 Model F and I would like to hook up the blue switch into the matrix in the position of an unused key so that it can be used as a function switch.

Anyone know the best way to emulate a flipper electrically? Would think it would be a small capacitor, but not sure how small, or if anyone has ever measured the capacitance of a Model F switch. Shouldn't be too hard to do with the right tools, just don't think I have them.

I also thought about using a big resistor as 4 mega ohms seemed to work, but when installed it seems pretty affected by static. the switch itself has a grounded case, so I thought running a ground to that would fix it, but seems not to. Perhaps I just haven't played around with it enough.

Any thoughts, or anyone out there with the tools to measure the capacitance of a switch assembly?

Getting this to work reliably would be a good way to add all kinds of external switches to do things, so it definitely has applications beyond the 3178 blue switch :)

Posted: 24 Aug 2017, 04:56
by XMIT
I wouldn't bother. The blue switch is already a switch, just read it directly. If you're using an xwhatsit or whatever I'd see if the MCU on the board has an extra pin that you could dedicate to this, and hack the firmware. Making a switch emulate a capacitive pad so that you can use your capsense controller, is putting the cart before the horse!

Posted: 24 Aug 2017, 16:09
by orihalcon
True, but when you put it into the matrix directly, it activates a lot more than just the one key. It will light up several nearby which I think is due to the voltage being way over the set threshold and is also sensed by neighboring pins of whichever chip does the initial discrimination.

On the 3178, it probably doesn't matter as there are entire areas that are unused by the matrix, so if extras are set off, it wouldn't affect anything. However, if you where to want an external switch where most of the matrix is used (like an F122), erroneous nearby sensing would be a problem.

Posted: 24 Aug 2017, 16:33
by Halvar
Why would you want to put it into the matrix? I think what XMIT meant was to keep it outside the matrix and just dedicate its own GPIO pin to the blue switch.

Posted: 24 Aug 2017, 17:40
by orihalcon
Wouldn't putting it onto the matrix be simplest with the least modification to the existing hardware/software? Just run one wire of the switch to one column and one row that is at the position of an unused switch on the matrix? Mapping then is as simple as regular mapping for any switch :)

Posted: 24 Aug 2017, 20:56
by seebart
Please share some pictures of this project when you have time!