You are working on an organ and a bunch of stops and couplers scattered throughout the console aren't working. The 4th DK in the first column, the 1st DK in the 3rd column, the 3rd DK in the 4th column and so on. It seems random. Pull out your Technical Manual and compare the stops and couplers that aren't working to the Stop Rail Wiring Table. You may discover that it isn't random after all, there is a pattern. In this case the 4th tab in every byte (8 tabs) wasn't working.
5) Is there a pattern?
Looking for patterns or commonalities, when a group of things aren't working right, is quite often the key to solving the problem. You will quite probably have to compare the console to the technical documents to do this. In the example above, the pattern wasn't readily recognizable, because although the stops were wired in groups of 8, they aren't positioned on the console in groups of 8.
Don't just look for things like the 4th bit of every byte, also look for something I call 'boxes'. I will sometimes see this with pistons or stops, especially in matrix wiring. This is when 4 things that are involved with the problem, when you look at how they are wired, form a box. For example: there are 3 stops that whenever they are on, a coupler will always come on. Looking at the position and divisions for these tabs on the console don't tell you much, but when you look at the wiring the 4 stops in questions are wired to
F-3 / P-1
F-5 / P-1
F-3 / P-4
F-5 / P-4
Now you can see the box pattern and you know to check to see if a diode is backwards or has gone bad.
Patterns and commonalities can also occur in the chamber. In a Classic system we are able to share driver boards, so you could have a couple different chests and an expression engine on the same board. So, the organist tells you that 4 of the stops in the Swell, one stop in the Pedal and part of a Great stop isn't working, all of these could very easily be on the same driver board. You have a Swell Main with 4 stops, and a 32 note unit which is used by the Pedal and Great.
This concludes my 5 part series on diagnosing technical problems. Too bad I don't get paid by the word, then I could finally replace that cheap keyboard I had back in school.