Apple Extended Keyboard Matrix Problem

[davidg5678]

I am repairing an Apple Extended Keyboard (not II) which has a number of key-switches that do not work. I was able to repair the majority using contact cleaner and reflowing solder joints, but I noticed several keys still did not work even after I reflowed and cleaned them.

In order to troubleshoot, I confirmed that the switches themselves work correctly by using a multimeter. I then decided to try bridging the contacts of the switches using metal tweezers, but the keys still didn't register as pressed.

The nonfunctional keys are J, K, L, N, 9, F7, and F8. On a whim, I looked up a diagram of the keyboard's matrix and found that all of my broken keys are on the same column. I now know that I have a non-functional column on the matrix, but I do not know enough about how these keyboards work to proceed with the repair.

Does anyone know what I need to do to repair a broken section of the key matrix?

Thanks!

 

[IlikeTech]

I would check for continuity between a key in that column and one in that row with a multimeter.  That might rule out having a bad input on the keyboard microcontroller.

 

[davidg5678]

After a week of troubleshooting, I was eventually able to repair my keyboard! This information should apply to people repairing Apple Extended Keyboard IIs as well because as best I can tell, they are electrically the same. Here are the steps I took to get everything functioning:

When I first noticed that some of my keyboard's keys did not work, I sprayed electrical contact cleaner into the switches, this step alone fixed several keys on the keyboard. Not every key was functional at this point, so I then reflowed each dead key's solder joint. Several other keys started working after this step.

I was not able to fix all of my keys after these two steps, so I then checked a diagram of the key matrix. Sure enough, the only keys that were still broken happened to all be part of the same column. Using a multimeter, I checked for continuity between every key in the column. They were all connected perfectly to each other. I already knew that the inputs on the main controller chip were working because every row of keys worked.



After a lot of research, I found that the columns on these keyboards are hooked up to a hexadecimal to binary decoder chip that then attaches to the main controller. I think Apple did this so that they did not need to use a controller that had a ridiculous number of input pins. There was no continuity at all between the keyboard's eighth column and the decoder chip. Strangely, after a quick check, I found no continuity from any column to the decoder chip. Initially, this did not make sense as most columns on the keyboard had to be connected, considering they worked just fine. Using a flashlight, I discovered that there is a series of diodes between each column and the decoder chip, located on the other inaccessible part of the PCB. Using my multimeter's diode tester, I found that every diode aside from the one on the eighth column was functional. Because the diodes are completely inaccessible without desoldering literally every key-switch, I decided to try bodging another diode onto the bottom of the PCB. This did the trick, and now my keyboard works once again!