Circuit, version 1

As previously mentioned I based myself on this schematic (so this is where the basic theory is explained), but with several modifications:

 The dance pad does not need debouncing circuitry, so it has been removed.
 This was a bit trickier: the contact pads on the dance pad have a self capacity that is already quite high compared to the one of the human body, so the circuit only has a narrow detection margin. When the game requires it, the player needs to touch two pads at the same time... and in this case his ’capacity’ is discharged twice as fast by the circuit, and it becomes problematic. So a bit of circuitry has been added and now the pads operate not all at the same time as they did with the original circuit, but one after another. This is why the 4060 and the 74ls139 have been added. Don’t worry, this happens at 100KHz or so, it’s not a problem for stepmania ;-)

Here’s the schematic that has been working rock-solid for more than a year:

And this is what it looks like, soldered way too dense on a perfboard:

I use a 3.5MHz cristal. It should be connected to two capacitors of a few pf, but my circuit works fine without these.

The schematic and a routed version of the circuit is available here in Eagle format.

In its current form this circuit is to be plugged into a joystick port. Each of the outputs will be plugged in the buttons inputs of the port, which are active low. Declare the pad as a 2-axis 4-button joystick, and voila !

The 50k resistors are here to simulate the presence of the two axis, this is required for this ’joystick’ to be detected by Windows. But do not try to run the calibration procedure, this would be asking for trouble, because of these fixed values.

This circuit can also be used on the parallel port. PPJoy can be used to map the parallel port as a joystick. This is what I used to do on my laptop, which runs XP, but when I decided to play on my desktop, which runs win98 (yeah.. I know), I ran into trouble as PPJoy is not fully supported on this OS (and I could never get it to work). The parallel port has no +5V, so an external power supply might be needed. Setting several data lines to a high level might provide an acceptable +5V though.