KDM20-P Problems, please help!

[SaltySeller]

HI all, having issues with the outputs from our KDM20-P controller on our MG series 1. We had a load of buttons that weren't working, they've been swapped out but since then 4 of our buttons no longer "work". The buttons themselves test fine and they've since been swapped out for known working buttons with no luck.

I'm after a bit more information on the two signal cables (TX+RX) and what the receiver in the main cabinet is expecting from/ sending up them from certain button presses. We've got the following 5 cables:
*+24V
*Gnd
*Em Stop
*TX
*RX

The four buttons that don't generate signals out of the controller are:

*Set surface
*Set Home
*Set Max depth
*Bar code (1)

Any help would be much appreciated!

[basswakr]

HI all, having issues with the outputs from our KDM20-P controller on our MG series 1. We had a load of buttons that weren't working, they've been swapped out but since then 4 of our buttons no longer "work". The buttons themselves test fine and they've since been swapped out for known working buttons with no luck.

I'm after a bit more information on the two signal cables (TX+RX) and what the receiver in the main cabinet is expecting from/ sending up them from certain button presses. We've got the following 5 cables:
*+24V
*Gnd
*Em Stop
*TX
*RX

The four buttons that don't generate signals out of the controller are:

*Set surface
*Set Home
*Set Max depth
*Bar code (1)

Any help would be much appreciated!

have you tried a new cable to the kdm20. i had button issues & it was the cable. just an idea to try .

[arsenic87]

You mention 5 wires. Are these the 5 wires that come from the controller?
From pictures of the KDM, it seems that these are the only 5 wires.

That means that the KDM itself is a "computer", that transmits what's going on over the TX/RX lines. (The E-stop is usually kept separate as an "active low" wire, so that any problems with it, will result in an E-Stop).
From what I've experienced on MultiCam machines, all TX/RX signals use the RS-232 (normal serial protocol). If you happen to have a BIT fancy oscilloscope, you can connect it to the TX line and read the signals it sends in bytes. This would be the "easiest" way to determine if the problem is KDM side, or Machine side.

For me, it sounds like this is a KDM side problem, and that even tough you've replaced the keys with confirmed working ones, the key-matrix lead to these keys is what's broken.
This particular KDM is probably a 4x8 matrix or something. (usually in division of eights, since a byte is 8 bits). In practice, this means that one lead of the key-matrix goes to 4 keys, and from each of the 4 keys to a unique lead of the 8 others. Just google "keyboard matrix" and you'll get an idea of the wire-layout inside the buttons.

So I'd recommend following the PCB trace (or plastic matrix) from the button, to where-ever it goes. Most likely, there will be one trace that goes to all of the 4 buttons, and this trace is either broken on the way to the chip it goes to, OR the chip-pin itself is fried. If you check continuity from the chip-pin, all the way to the buttons, then the problem most likely is the chip itself (thus requiring replacing that chip). If it's just the trace, you can possibly manage to solder a small physical cable to replace the PCB trace.
If it's a physical tear in a plastic keyboard matrix, I've managed to repair such a problem by using aluminum or copper foil to bridge the gap. Or even just using a graphite pencil just drawing heavily across the gap to create enough conductivity.

Hope this can help!