What could I expect to hear from a Class A amp?

Well designed class AB doesn't have crossover distortions because it is removed by negative feedback. There is no THD or IMD but instead hidden TIM distortion. TIM causes odd harmonics (sharp sound) and fatigue (brain fills moments of silence caused by momentary saturation).

Class A doesn't need feedback as deep because of operation on linear characteristic of transistors but also because it needs less of overall gain before output stage.

Unfortunately while class AB consumes, playing music, on average only few percent of maximum rated power the same time class A consumes full rated power.
Efficiency is extremely poor - in order of 12%.

My 2x200W class D amp consumes, during playback at full power about 35W while class A would consume 3200W (even idling).

Bearotti - class D is very good at low levels (lows, highs, definition, dynamics, clarity etc)

Once system becomes linear (by negative feedback or otherwise) it doesn't produce IMD - just look at IMD specifications of class AB amplifiers. Because amplifier has limited bandwidth it introduces delay to signal and negative feedback cannot compensate in proper phase (is too late) to fast changing (high slew rate) input signals. Because gain of amplifier before feedback is in thousands momentary lack of feedback response causes spikes in output signal causing unpleasant odd harmonics up to a point of saturation. Momentary saturation of output transistors causes charge to be trapped at their junctions for a moment. For this moment there is no sound. We don't hear it since our brain fills the gaps but it makes us tired.

It is possible to design class AB amp without (or with minimum TIM). Sane designer would set up gain before feedback to get with feedback only about 0.5% THD (or even 1%). Than he would measure bandwidth before negative feedback and would limit it (with feedback) in input stage to the same value. That would guarantee no TIM. Also he would use many local negative feedbacks avoiding deep global one. I don't want to bore you with detail of amp design (and I don't design audio) but my point is this: If you see a class AB amp with THD=0.001% stay away from it (something has to give).

There are amps that allow to adjust depth of feedback and people claim that the best sound is at the lowest feedback and highest distortion. Designing class AB for THD=0.001% is insane since speakers have few percent of THD. (Look at Stereophile article "Life without feedback")

Buy an amp for the sound and don't even look at specifications.

Class A is the best but at the price of electric bills.

Shadorne - I suspect that ferrofluid used in many tweeters changes viscosity with temperature.

"Bearotti, class AB amps usually operate in class A up to some power level"

Yes, but won't give class A performance. Much higher signal gain and deeper feedback were already applied in order to remove THD and IMD and you already paid the price. At this point it doesn't really matter if you listen loud or soft.

Class AB is a compromise for another reason. In order to linearize transistor that is completely OFF (class B) it would take much higher signal gain and deeper feedback than class AB, creating even more TIM problems and odd harmonics.

"On separate issue, would all that heat for 24/7 at full power take toll on the circuit board? Of course the counter arguments would be the on/off as you said, and the expansion/contraction fatigue cycle due to the hot/cold."

Spatine - the only perishable component inside of SS amp is electrolytic capacitor (other than ON/OFF switch). Rush current does nothing to electrolytic capacitor other than heating it slightly (very short time and very low ESR) in spite of large momentary current. Electrolytic caps die of dehydration. Each increase in temperature of 10 deg Celsius cuts their life in half (water evaporates from electrolyte). Eventually drier capacitor has higher ESR and starts heating up to the point of runaway. In critical application (lots of current) they get to thermal runaway and explode (they have fuse to relieve pressure). The other way to damage them is to leave them without voltage for years. Aluminum oxide layer that serves as isolator is eaten by electrolyte but presence of voltage builds this layer back.

Fairy tales about rush current (turning ON) damaging equipment comes from two sources. First is the bulb or tube that breaks on power ON since cold filament has many times lower resistance. Second is the fact that if you do anything around your amp like for instance moving speakers while amp is OFF and create short it might damage amp when you turn it ON and you will blame turning ON (rush current).
Sometimes also mechanical component tends to break at start-up like for instance computer's hard disk.

Circuit boards, resistors, film and ceramic caps, transformers etc last forever even in extended temperatures. As for semiconductors - as long as they operate at sensible junction temperature (let say under 125 deg C) they also last forever. There are electronic circuits with semiconductors being turn ON/OFF hundreds times a second and they last for over 25 years. Simple example of it is seven segment multiplexed display.

As for sound being optimal after a week of continuous operation - it's possible. I don't have class A amp, don't have experience with them and perhaps less than "golden ears".

People who calculate reliability are probably ready to kill me for saying all this but statistical analysis they use is something else. When I walk with my dog we have statistically 3 legs each. Also statistically tattoos are causing motorcycle accidents.

"It seems I have to continue putting up with listen to music late at night after giving the system most of the evening to warm up."

Spatine - Can you hear it improving over time. Would you know if amplifier was ON for one or three hours? If you cannot tell then don't listen to somebody's recommendations.

You made excellent point about power outage and surge. I use surge protector but still turn off the switch while not at home and unplug my gear during thunderstorm or long absence. Surge protector works for small spikes but is useless for direct lightning hit. My insurance agent told me that stereo equipment is always insured independently of value (One doesn't need special insurance) but it's a little hard to believe.

My amp is class D and uses about 10W idling but I still turn it OFF - I don't see (should I say "hear") reason not to. In case of class A amps electric bills should be one more argument - unless one can really hear big difference over time and is willing to pay for it.

How about blind test - you come home from work and find your gear ON. You have to guess how many hours it was ON. This would perhaps tell us what is real and what is placebo effect.

Spatine - why do you call it waiting period? Your gear most likely plays at 99% of quality and you can still enjoy music.

On my system it takes short time probably half an hour and it might be related to tweeters warming up.

If I can toss monkey wrench here: Is your time of 4-5 hours same at different hours of the day? I suspect that minuscule improvement might be caused by drop in radio interference at about 6PM since radio stations have to lower their power (FCC rules). I hope that Bearotti doesn't mind that we sidetracked his thread a little.

Al - read also this http://stereophile.com/news/10065/

Class A is probably the best SS. The real question is what is the best amp for the money one has assigned to spend. For me it was class D - maybe not at the level of top class A amps but way better than my previous class AB amp.