Originally Posted by Trepidati0n
Hate to say this, but no. Heat has little to deal with why capacitors fail in audio amplifiers.
You should read this:http://www.lintronicstech.com/index%20pdf/reliability_of_capacitors_general.pdf
A rule of thumb that "us" engineers (yes, I'm a senior audio circuitry design engineer for Shure and calculate this frequently) is that lifetime of a capacitor doubles for every 10C reduction in temperature from its maximum rating - which you will even see mentioned in this paper. In practice, "we engineers" use more advanced techniques for calculating MTBF because we take into account several other factors and we want to make sure our products meet their specifications for the life of our products - which is typically longer than 15 years for Shure products.
Granted I don't design power amplifiers in a professional scenario, but I've never seen MTBF dominated by silicon in any of my designs. And even for the hobbyist amplifiers I've designed, the caps are always the limiting factor. If you go with the best transistor available for any given application, I would be extremely surprised to see it fail - especially in a well thought out topology.
Btw, your comments about the refresh rate of a capacitor, although true, is not the complete story. Your discharge rate and bandwidth of your power supply are far larger factors for dealing with power supply related sonic artifacts. You better have a crap ton of rail capacitance if the bandwidth of your power supply isn't much greater than your audio signal...
At the end of the day, I believe Class D should be the best amplifier topology when looking at the total system design....directly couple a good Class D amplifier to your loudspeaker voice coil and there's all sorts of tricks you could play that aren't as easy to do in the Class A realm.