Neville Ward is the head guitar tech and owner of ZenWorks Engineering, he has been fixing amps longer than you've been holding a plectrum, and writes about his experiences in his Amped blog - Diaries of a Guitar Tech...
I have just completed the restoration of a British guitar combo dating from about 1964, manufactured by Bird. When it arrived it was working, just, but sounded thin, weak and nastily distorted.
This was one of those jobs that gets bigger and more complex the longer you work on it. As well as that, there were numerous times that I was left thinking "Who thought that would be a good idea?" This happens more often with modern, cheap machines; things that have been engineered down to a certain price, rather than being engineered up to a certain quality. As a result, many machines that I see are difficult to service for simple reasons like being designed to be easy to assemble (usually in a Korean sweatshop) but difficult to dismantle, or using unique and rapidly obsolete parts, instead of industry standard parts.
But sometimes when things get difficult I realise it is simply due to laziness or sloppy thinking on the part of the designer; I can see tiny changes that would have made the machine easier to use as well as easier to service, or changes that would have added nothing to the price but much to the flexibility or longevity. Of course, most manufacturers don't want their products to last a long time; they want them to fail the day after the guarantee runs out in the hope that the customer will stump up the cash for this year's model. Even so, after years in this trade, I have become very good at spotting "failures of thinking".
Taking the amp apart was a unique experience for me as it involved the use of a socket set: something I usually use for servicing the car. With the chassis removed I could see the full catalogue of horrors: large resistors had been added to the rectifier making the supply voltage droop under load, mains wires were running right next to high impedance signal cables, the PCB didn't have enough holes and many of them had the leads of two components in them. Honestly, what does it cost to drill a few more holes? Clearly these were all failures of thinking, and I expected there would be more. I was right: the output transformer had extra primary tappings for Ultra Linear Mode (usually only seen in hi-fi amps) and the secondary impedance didn't match the loudspeakers.
Leaving aside my reservations, I set to work. Replacing two capacitors was all it took to remove the distortion, but there was still plenty of hum, not much volume, and the reverb, tremolo and bass controls were not working. As you might expect with an amp of this age, most of the capacitors had deteriorated to the point of near failure. I had expected that. Replacing them brought many small improvements, but the lack of bass control was a puzzle. Trying it out with a guitar, the amp would only just reach clipping with everything up full. It also wasn't very bright.
Consulting the circuit diagram, and referring back to the amp in front of me, it was clear that several changes had been made to the circuit after it went into production. These were mostly in the form of added components that didn't sit on the PCB but were hanging off various wires and tags. There were several associated with the bass control. Removing them restored the action of the bass control, but at the expense of some loss of gain. So had Bird responded to complaints about a lack of gain in early models by tweaking the design to give as much gain as possible? Removing feedback increases the available gain, but since the bass control was in the feedback line, it was left with nothing to do. So you are left with an amp with a useless knob. Who's going to buy that? Another failure of thinking.