trying to trigger multiplexed buttons with 0-5v signal

[Triode]

I'm a robotic engineer, but I am oddly limited in this case as I can't design the system or modify it much. For an expo they want to hook up real industrial style joysticks to a gaming joystick board for a 3D simulator of our machines, they have already designed the software around the gaming joysticks and assumed that dropping in these other ones would be easy, now they are out of time and gave it to me. They want me to quickly patch the industrial joysticks into the gaming joystick board. I have the analog axes figured out, but the buttons are an issue:

The buttons connect to the main IC by 6 wires, I can't look at it's code, but it seems to be using matrix multiplexing. One wire branches to 4 diodes going to 4 buttons, another has 2 diodes going to 2 buttons, the other 4 wires are returns from those buttons. They are switching at high frequency. The wires from the joystick buttons are 5V normally and go to 0V when pressed.

I've tried just putting NPN transistors across the button contacts and connect the signal to the base (with a resistor) and it seemed to be working, at least on the 4 buttons that share a lead, when I add the transistor with either of the 2 buttons on the other lead, or both for that matter, they all stop responding. I don't know why that would happen, but I need to get it working by tomorrow evening. I'll be working on it and I'll post an update if I find anything new. I have basic transistors and resistors on hand, and a few support ICs like inverting buffers.

 

[ChrisP58]

Can you post schematics of what you have, and what you are trying to change it to? It is difficult to understand exactly what you need from a text description.

 

[4pyros]

Try using a PNP transisters on the other two buttons.

 

[dr pepper]

The software demux might be using both hi and lo logic levels to scan the inputs.

Cant you just unsolder the original switches and solder 2 wires to each of the industrial joysticks switches, or am I missing something.

You might need to put a resistor in series with the switch contact, I'd guess 1k ohm, as if the chip apllies lo to 1 side of the contact and hi to the other you might deck the 5v line resetting the chip.

 

[Triode]

I ended up doing what dr pepper suggested. I originally didn't want to because the wires have to go a few meters and I didn't want a huge bundle of them, but the electronics we're just too finely tuned to go any other way; a little more or less resistance or a transistor with different switching time changed the behavior completely. But I got it done by the deadline and it works so I guess that's that.