An easy fix for a TRS-80 Colour Computer 1 keyboard problem

You may have heard of the term "vintage computer rot". This refers to a irritating phenomenon where computers which have been stored for a length of time seem to develop faults. A fully working machine can be put away, only to fail on retrieval!

I experienced this recently with my Tandy Colour Computer (CoCo) 1 seen opposite.

In March 2009, I purchased a TRS-80 Colour Computer Deluxe Joystick off e-bay from a Canadian buyer. Believe it or not, the joystick took an incredible three months to get to New Zealand?! I had written it off as being lost! Anyway, it finally arrived so I dragged out the CoCo to check it out. The machine had been in storage for about seven months.

The joystick worked just fine. What didn't work though was the keyboard! To my dismay keys I,O,H,J,K,L,N,M simply did not respond. They had been ok when I packed the machine away so somehow, as had happened with my IBM PC , Apple II clone and Osborne 1, a fault had developed between that time and now.

I didn't have a technical manual for this machine. However, I knew from my Commodore 64 experience that keys were often addressed in a matrix. If I could find a diagram of that matrix I could ascertain if the affected keys corresponded to a single pin or line. If not, perhaps it was just dirty contacts under the keys. However if there was a pattern, then perhaps a pin connection was loose/dirty or (worse-case scenario) an IC was damaged.

Googling around soon uncovered a keyboard matrix diagram. Sure enough, the effected keys were all on one pin, pin 2. Dirty contacts under individual keys then were unlikely. Still, I was optimistic. Maybe a pin in the keyboard lead had just come loose or was a little dirty.

The naked Tandy Colour Computer 1

Time to undress the CoCo. Removing the top was easy; it was simply held on by a few screws in the bottom. Once opened, I could see how the keyboard was connected. A plastic connector was simply pushed onto 16 pins on the main board. There didn't seem to be anything loose but I disconnected the keyboard for a closer inspection.

Keyboard pins on the main board (left) and the plug on the base of keyboard (right)

The first thing I did once the keyboard was detached was to check if there was a closed circuit between the wires going into the ribbon cable from the keyboard, to where they came out of the plastic connector via the sockets on line 2.

There was a closed circuit, so no problems there.

The pins were a little dull and had dark specks on them. One possibility I considered was that a light layer of oxidation may have built up over time, just enough to stop the circuit on pin 2 working. To check this theory I cleaned the pins with fine sandpaper. After this was done, I plugged the keyboard back in and switched on. I figured that if the problem was still there after this test, it must be a component on the board itself.

Thankfully, after testing the keys I found the problem had gone! They all worked!! It seems my oxidation theory may have been sound.

TRS-80 Colour Computer 1 (plus joystick) now fully operational

The moral of the story is that sometimes (in fact, some would say often) vintage computer fixes are just a case of simple cleaning. Oxidation on contacts develop over time, whether the machines are going or not.

It always pays to check the simple possibilities first.


6th June, 2009

comments powered by Disqus