DOUT to DIN with Switch matrix

[John_W._Couvillon]

Midio128, ver. 3, LPC17, using three matrices for keyboard encoding.

In connecting the DOUT through the keyboard contact and diode to the DIN, the series 220ohm resistor usually on the DOUT has been eliminated. with this arrangement, I seem to be losing DIN 74HC165 chips. Pressing a key brings the correct note, plus others. Changing the DIN chips corrects the problem. Am I just paranoid, or is there an inherent problem with the circuit arrangement?

Johnc

[jkinkennon]

Midio128, ver. 3, LPC17, using three matrices for keyboard encoding.

In connecting the DOUT through the keyboard contact and diode to the DIN, the series 220ohm resistor usually on the DOUT has been eliminated. with this arrangement, I seem to be losing DIN 74HC165 chips. Pressing a key brings the correct note, plus others. Changing the DIN chips corrects the problem. Am I just paranoid, or is there an inherent problem with the circuit arrangement?

Johnc

Hi Johnc, There's nothing inherently wrong, in fact it's better to not have the 220 ohm resistor in the circuit. If you've lost more than one IC then something is going wrong for sure. I'd first check that the 10k pullups are correctly installed on the DINs and are working, that is the 74HC165 inputs are staying at about 5v when no keys are pressed. If that checks out I'd be thinking of watching the signals on an o-scope to be sure there are clean "square" transitions and that inputs aren't spending too much time at intemediate values between 0 and +5v. Long cables with high capacitance might be an issue. Certainly you want to be sure the +5v is hooked up on the DIN and DOUT and that each chip has the little rf bypass cap installed {the .01 or .001 capacitor).

[John_W._Couvillon]

Thanks for the input. The DIN is an older one and I don't think that it has bypas caps. I'll check it out.