Taping into buttons

Nigel Goodwin said:

There are various issues because it's not truly isolated, where a relay contact is (up to a pretty silly voltage).

Click to expand...

Hi,

Sure, and usually the same power supply that supplies the main circuitry powers the CMOS analog switches.
Also, kind of amazingly, for an NxM multiplex matrix only N+M analog switches are needed, and for SPST functionality we can get four switches in one package. My 64 squares project took 8+8=16 switches which means it took 4 IC packages, and they are the usual DIP package type unless you want to go SMD. 16 relays would be kind of large and would never have fit inside the chess computer case.

Relays are often a go-to solution which does not take much effort, except the power consumption can be a little higher because of the relay coil.
These days they make some nice relay boards we can buy on Amazon and the like. One of those might work in this application. I think they even have screw terminals so no soldering required

Some type of opto-isolator may work? A transistor or darlington type possibly, if the polarity is correct?
Or such as an H11F1 which has a polarity-independent FET output?

That would keep the add-on size to a minimum, much smaller than normal relays.

I suspect the matrix method may be due to the manufacturer having other products with a larger number of buttons and the same software being adapted to this application. I very often use software that I have written for something else as a starting point for a new project.

Les.

It's also interesting that if we have a square matrix of N buttons we only need sqrt(N) switches in order to 'push' every button electronically.
For an NxM matrix we only need N+M switches to be able to 'push' every button electronically.