Cambridge 220-II: other stuff I fixed
The organ had a number of minor problems when I first got my hands on it. The following is a description of the work I did to fix most of the smaller problems.
The technical manual (still sitting in the organ bench!) was incredibly helpful for troubleshooting problems in the console. In addition to the schematics, it had a checkout procedure to exercise every circuit in the console in a systematic way. This sort of troubleshooting is fun -- in modern electronics all of the logic would be accomplished in a single chip so these sorts of problems would never come up. (And if a hardware problem did occur, it would be in some impossible-to-replace surface-mount chip anyway.) It was fun going through the circuits, trying to deduce what single failed component could be the common cause for the observed problems.
The most obvious problem in the console was that key 54 on the Great keyboard didn't produce any sound. (Each keyboard contains 61 keys, numbered 1 through 61 from low to high.) The dead key moved with different transposer settings, so the key itself was OK. Key 54 on the Swell keyboard was fine, so the problem wasn't in the transposer either. (This organ uses a 'time-share' system that puts the signals from Swell key 54 and Great key 54 through the same path in the transposer.) After leaving the transposer, the keying signal goes to the coupler diode gates, then the Great pulse-to-DC converters, and then to the Great stop diode gates. Since the problem was with a particular key, rather than an oscillator, the problem had to be before the diode gates. Since the Swell and Pedal coupler diode gates weren't affected, the problem had to be something that impacted only the Great key 54's pulse-to-DC converter circuit. In the end, I found a diode (D197 on Rodgers drawing 1443) on the 8'-to-8' Great coupler diode gate card for key 54 that conducted in both directions. That let the capacitor in that key's pulse-to-DC converter circuit discharge back through the coupler diode gate, so it could never activate. The fix was simply to replace the diode.
There were other dead keys that suggested a problem with an oscillator: On the Swell keyboard, keys 49 and 61 for the Fife 1' stop, key 61 for the Piccolo 2' stop, and key 57 for the Tierce 1 3/5' stop all were silent. What all of these have in common is that they produce the same note, using oscillator number 85. However, other notes that use oscillator 85 (such as key 61 for the Blockflote 2' on the Great keyboard) were fine, so the oscillator itself was not at fault. Looking deeper, the three dead keys all activate the same signal keyer (for the Swell unit flute). I examined that circuit and found a crack in the circuit board that ran through that keyer. I bridged the broken traces with bits of wire and that solved the problem.
There was also a problem in the transposer: switching to the three-semitones-flat setting made the organ go silent. Everything worked fine after switching back to a different setting. The way the transposer is wired, pressing one of the transposer setting buttons simultaneously activates the chosen setting while signaling all the other settings to turn off. The button itself had to be fine, since the other settings were deactivated when it was pressed (leaving the organ silent). There had to be some problem with the circuitry for the three-semitones-flat setting that prevented it from turning on. Each transposer setting has one SCR associated with it, the setting is on when its SCR is active. For the three-semitones-flat setting, the SCR (Q87 on Rodgers drawing 1404) was burnt out. The SCR was part number 2N5060, which I wasn't able to find in stock anywhere. I replaced it with a 2N5061 instead (identical electrical characteristics except for a higher voltage rating), which fixed the problem.
The switch under the Swell keyboard that activates the 'tutti' setting didn't work. There is a toe switch that does the same thing and that one worked fine, so the 'tutti' circuit itself was OK. I examined the keyboard switch and found that a wire had come loose. I soldered it back on and that fixed it.
I replaced the expression indicator light bulbs and driver entirely after going through a few rounds of replacing one burnt-out light bulb only to have another one fail. The new circuit is described in another article. This fix worked pretty well, except the amber LEDs I used to replace the lights now look green, which is annoying.