caution said:
Thanks. I'm aware of the piezos going bad but I wasn't sure about the amp chips, either from use or power/thermal cycling, in a way that only stole from the overall quality of the audio and didn't inject junk like popping or motorboating. From what you're saying about discrete transistors and audio-range frequencies, probably not.
I'm getting the sense that in terms of replacing components, audio path electrolytics are, in terms of cost/labor versus possible improvements, about as far as one can reasonably go to bring a unit up to its better-sounding counterpart, unless you have the schematic and a working knowledge of audio sections of tuners and audio amplifier circuits in general, that would allow you to identify components with the highest chance of impacting audio quality through what you mentioned - either ESR, drift (especially printed components) or in the case of tuner transistors, degradation.
Other than the occasional polystyrene I've seen none of the older types of caps (Mica, paper, etc.) and instead the more stable materials like mylar, ceramic, polypropylene and tantalum, which I have to believe minimizes meaningful drift.
First regarding the caps, only electrolytics should be indiscriminately replaced. Ceramics do go bad but not with any amount of predictability. Tantalum are notoriously poor in terms of tolerance but they are usually only used in applications where tolerance is not critical, although there have been known cases where tantalums fail across the board in some models due to use of poor quality components to begin with but that is an idiosyncrasy of specific models and not a case in general.
As for monolithic amp modules, they frequently but don't always fail dead. In most cases, when they start going bad, they will begin to distort and heat up badly. But I would say that this isn't something I would recommend replacement in a refurb unless they are known bad. Genuine replacements are hard enough to find when there is a real need.
In general, your best bang for the buck is going to be non-selective recapping. Some would advocate only replacing those caps that tests bad, but what threshold are you going to use to determine "how bad" a cap is? Also, my general philosophy is that if you take the time to remove a cap from the circuit for testing, unless it is a large expensive one, why bother sticking the same cap back into it? Time is money and caps are cheap. Ok generally speaking that is, because audio grade electrolytics are many times more expensive than general standard grade caps but I would surmise that general grade is probably "good enough" for most boomboxes. If one cap goes bad, others are probably not far behind, after all, they are all the same age, all have been exposed to the same number of power on/off cycles, same environmental conditions (heat, corrosion, voltage surges, etc). As far as transistors go, I have in the past seen forums where people advocate general replacement of all signal transistors in the equipment they are restoring. I would not go there. Firstly, you are just asking for trouble if you replace the transistors in a tuner. I think the chances are equally good that a tuner will no longer work after replacement of the transistors compared to working better. In this case, identifying the bad transistor and finding a suitable replacement is challenging enough. The reason why is because of the ultra high prevalance of counterfeit stock in the marketplace. China has done a number on us all and flooded the market with literally tons of counterfeit products. Most do not spec properly and in an audio circuit, might somewhat function, but in a tuner where the transistor spec is ultra important, it's generally enough to render the tuner totally inoperative. If you don't believe me, look at the transistors in a tuner and see if you can find a replacement part. Oftentimes, you will find them on eBay only and they are all from Chinese sellers. They are 99% fake and won't work. Worse yet, they will introduce diagnostic nightmares to your job. The reason is they will test good with your tester, but not perform properly in circuit.
For example, let me share this story with you. A few months ago, I did the following repair. It is on a Harmon Kardon AVR receiver. It is not as old as most classic boomboxes, I would say about 25 years old. I purchased it new and used it for about 2 years, then passed it on to my brother in law. He used it for about 2 years before the display went out. Even though it still sounded fine, he couldn't see what he was doing so he took it out of service where it sat for 20 years. Recently, we rediscovered this and he offered it back to me for repair. First thing I did was repaired the bad flourescent display. Turns out it had a bad solder joint. Once that was repaired, I noticed that it distortion in the left channel, and there was a power on/off delayed growl. A quick test shows that the amp had unacceptably high DC balance in left channel. Right channel was about 15mv. Left channel was -200mv, way out of spec. My pre-conclusion was that there is a bias issue with that channel, and that aging/failed caps were to blame for the growl. Before ordering parts, I studied the circuit diagram and identified certain parts that might be bad. Among them was 2SC2290BL transistors in the preamp section. After I opened it up, I replaced all of the electrolytic caps and there was a lot, and a ton of boards. Like a dozen of them. I replaced all 5-channel's bias pots with multi-turn units. Voltage measurements indicated there was indeed problems, several failed or out of spec fusible resistors of unique values -- all obsolete. This required some fancy combination to come up with something that would work as the value was critical since they control bias voltage. Finally when I had the transistors out, I tested them and discovered that the original had beta value of 600hfe. The replacements I bought from eBay all were slightly smaller and all 40 tested beta at 200hfe. THIS IS COMPLETELY WRONG. I tested each and every one. Upon contacting the seller, he immediately refunded my payment and begged no negative feedback. Luckily, the transistor was still working properly so I reinstalled it and didn't need the off-spec replacement. However, if I had installed them without testing the beta, then I would be scratching my head since now, I will have a working but not properly spec'd transistor in the system which surely will not work right. How are you going to figure it out? Also, a poor transistor that is rated for 30v placed into a system with 60v could cause an avalanche cascading failure of the entire amplifier wiping out lots of very expensive and perhaps obsolete parts. Lesson learned here is that simple diode testing of the transistor will show it to be working, but it's not really is it? Hence the caution regarding counterfeit parts and the difficulty in procuring good originals.
BTW, after servicing, the receiver now works perfectly again, and should be good for another decade or more of use. Parts alone for the refurb costed more than you can buy this receiver for online. However, there is simply no way to buy one of these guys that isn't already having problems, or will soon have problems. After, all they are all the same age and the one having undergone the refurb is only lightly used -- that much I have personal knowledge of.
However.....................
