Interfacing DOUT w/DIN in a switch matrix

[John_W._Couvillon]

TK, anybody

The output pin on DOUT is connected through a diode and key contact to the input pin on the

DIN. What value pull up resistor should be on the DIN PIN?

Is a pull up resistor needed on the Dout output pin?

Will the voltage drop to ground on the DOUT pin be lower if a ULN2803 driver chip is used at 5vdc inbetween the shift register and the output pin similar to the setup for driving magnets?

Thanks,

johnc

[CSC]

Heres SomeBody,

my Button Matrix without LED has to be connected over common Cathode at DOUTX4.

16x32 - 16 (Col) Channels at 2 Shift Register on DOUTX4 with original Resistors (220) over a Diode [Cathode - Anode] thru 32 Buttons at 4 Shift Register on DINX4 with original Resistors (10k).

I don´t know anything about the setup with ULN2803.

:clover:

Edited August 5, 2011 by CSC

[John_W._Couvillon]

CSC,

So you have the 220 ohm resistors in parallel with the diodes?

Do you have a schematic of your circuitry?

Thanks,

johnc

[jkinkennon]

John, I hadn't been following the entire chain of events leading to your related post about the DOUT to DIN connection. I wouldn't recommend using a ULN 2803 driver chip on the DOUT as that will leave all the matrix columns in a high impedance state when they are not being driven. I don't know that the drivers wouldn't work, but they are not needed. When the diodes are not biased off by a positive voltage on the matrix column they are subject to coupling unwanted noise pickup to the DIN inputs.

Edited July 14, 2012 by jkinkennon

[jkinkennon]

One more thought about your DIN and DOUT configs for a matrix. I'm running a similar matrix circuit for my Allen 603-D midification project. I just experienced "extra" notes when I connected my pedal matrix. The pedals have longer ribbon cables so I put the scope on the inputs (similar to your DIN config) and found that the combination of operating speed and increased cable capacitance was too much for the 10k pullups to deal with. Basically the rise time of the signal was slow enough that the voltage was at an indeterminate value when the next column was driven. So I'm swapping the 10k's for 4.7k and if that doesn't cure the problem then I'll slow the circuit a bit.