Lifespan of a quality solid state amplifier?

Filter capacitors have a half life of about 20 years. What that means is that in a 20 year period, about half of the caps will have failed. This is true of transistor and tube amps.

Sometimes when the cap fails the ESR goes up like Kijanki was talking about, you can also have the experience of the cap slowly loosing its capacitance. This may not result in the hum that is commonly associated with filter cap failure, but it will impose sonic degradation as the power supply is no longer properly bypassed. This can increase distortion and affect bandwidth, things that you may not notice over time but would certainly show up on the bench.

So in general, by the time a piece is 15 years old, there will be some degradation or outright failure, by the time 20 years has passed the chances of that are 50/50. This will be true even if the unit has been in continuous use, although such pieces in general do have the caps last longer, that's not the same as saying they actually still meet spec!

Herman, that is the correct use of 'half life' in this case. This is what they teach you in technical school. I have seen it borne out many times in the last 40 years.

English is an evolving language.

'Half life' works really well in this case since you have the statistical issues of a single cap and also that of several.

In the case of a single cap its half life is 20 years on average. IOW, you will find that it may well have dropped to half its capacity, with its ESR likely more than doubled. I have never seen caps over 40 years old that are safe to operate; but by that time they should be replaced, without question.

In the case of a bank of filter caps it will be found that in 20 years time half of them have reached the end of their useful service life, by leakage, by shorting, or simply so ineffective that the equipment using it no longer meets spec.

It seems that 'half life' works very well for this. However if sensibilities are offended (we're not in Alice's Wonderland, apparently) then MTBF is fine.

Herman, A few years ago I heard about a Chinese firm doing some corporate espionage. Apparently Panasonic had developed a more effective dielectric for their electrolytic caps. Fearing the discovery of their formula by competitors, they kept it in 2 halves in different locations. How I heard it was that the Chinese company had managed to get one of those halves and used it in their caps, which got used in a lot of computer boards that began failing about 6-8 years ago on this account. Is this anything like what your firm was experiencing?

Hi Herman, I too have taught in technical schools and my stint in the service industry goes back to 1974. That's how I financed my way through the Minnesota Institute of Technology. The use of 'half life' is my own and simply comes out of experience - as a service tech I have repaired thousands of amps, preamps, receivers, tuners, tape machines, etc. I'm sorry that the use of the term bothers you but it works so well and is so easy to explain that I will continue to use it. I'll try not to use it around you :)

I'm with Cathode on this one.

But- you won't want to listen to lofi all that much either...

Most electrolytic caps used in power supplies will show some signs of failure before they start to leak. Most such capacitors have some sort of plug on the terminal end that will show signs of bulging prior to outright failure of the part.

Its a judgement call as to how much of that you want to put up with, IMO if you want to be safe, any electrolytic that has a bulging plug should be replaced.

I've seen corrosion at the terminals open up computer grade caps that were only 5 years old.

It is true that heat will shorten the life of an electrolytic. But even if there is no heat, to continue to operate an electronic piece with 30-year old caps in it (even if it seems OK) is asking for trouble.