Roland D-5: every 8th key doesn't "fire"

tombone · Joined: Aug 01, 2009 Posts: 3 Location: San Francisco

I've got a Roland D-5 from around 20 years ago. It still works normally -- but, every 8th key starting from the low Ab (Ab, E, C, Ab, E, C, Ab) does not "fire". I assume there's a failed line in the key detection circuitry, but not being familiar with how it's designed I'm not sure. Does anyone have any ideas on what may be wrong and how (or where) I might get it repaired (economically)? I'm in the Bay Area.

Thanks for any advice.
-Tom

DrJustice

The keys are most likely scanned in an 8 x 8 matrix. So you're probably right in that a line in the detection circuitry is failing. As always, the first things to suspect is a bad connection. I'd open it up, then follow the circuit from one of the keys that fail and check that the connections are good all the way to the main circuit board.

DJ
--

tombone · Joined: Aug 01, 2009 Posts: 3 Location: San Francisco

I opened it up -- though haven't disassembled it completely -- and checked the voltage of the lines on the single cable connecting the keys to the main board.

The cable has 24 lines which enter the main board in two groups, 1 with 14 lines, the other with 10. It appears that the lines on the group of 14 remain at 5 volt regardless of whether or not a key is pressed. This is also the case with two lines in the group of 10. The other 8 (in the group of 10) change depending on key presses. 5 volts is definitely high, but I also see 200 mv, and maybe a lesser value (I haven't checked closely).

I did find when pressing the keys which don't "fire" that the 5 volts is not produced on any of those 8 lines.

I'm going to dissemble it further to try to get better access to the keys board, but thought I'd report back and see if you have any thoughts.

tombone · Joined: Aug 01, 2009 Posts: 3 Location: San Francisco

Took a long break from this.

I took at look under the keys and read up on matrix switching.

It looks like each key has two switches with two inputs and one return. My problem keys start at low Ab, and include every 8th key after that. As would make sense, these are all on the same return line. Oddly, the lowest C (the first on the keyboard) is also on that return line, but it works.

The keys that don't work never seem to get a +5 on the return line (as seen with a voltmeter, maybe an oscilloscope would show something more).

What could be going wrong here? It seems like the input lines (which correspond to the 8 blocks of keys) are being scanned correctly, as not a single block is entirely out. Connection of the first return line can't be the problem since the lowest C is returning a signal. Seems as if there's a break on the first return line somewhere between the lowest C and the next Ab -- though I don't see anything. Any other explanation?