Indeed. Depending on to what I am listening I need between -22dB and -2dB approx attenuation for a similar sound output level from my DAC. I should point out that the DAC has been set to give 0.53V output for 0dB input else with my amps I'd have the speakers on fire.
One thing that I haven't seen mentioned above is the phase angle response of speakers. This is important because under certain conditions a (say) 50W amp will deliver less into the speakers and use the rest to cook the output devices. This is another reason why a beefy amp performs better. Its not the power per say nor to go louder but the ability to handle difficult loads under any condition. What struck me about my amps was the dynamics and timbre of the instruments that made the performance sound real. Also using an spl meter these more powerful amps sounded louder but the measured average spl was lower. I guess that this is due to the ability to cleanly handle the dynamics of instrumental waveforms (when present) over the whole audio frequency spectrum.
Cheers,
DV
The conditions you name are the worst for the transistors yes but it is in the safe operating area (normally abbreviated to SOAR) that is the biggest risk under these conditions and as output power goes up the problem gets
worse as we need more Volts available... hence we end up with loads of pairs of output transistors, not so much to
give all this power but to not blow up when
trying to! If you look at a typical high power audio output transistor data sheet you will see impressive figures like max current 16A, max voltage 200V, max power dissipation 200W, unfortunately the gotcha is that on the Volts and Amps this does not mean at the same time! In the more honest data sheets you will also find a SOAR graph and it's pretty shocking.... at 100V the max safe current is usually around 1 Amp! Above this we go into Secondary Breakdown Region of the SOAR graph and will likely get a short circuit at a random point on the die where it happens to be the weakest. As power goes up in the 300W (and far above in some powerhouse amps) range you have to have loads of output devices for them to not blow up under the conditions you mentioned! Also of course to avoid it going "power into 8R = 300W, 4R = 500W, 3R = BANG!" the same applies...
As a general rule the more output transistors you have relative to the maximum power, the less protection (or any in some cases!) is needed.
Now of course another alternative to having it go bang is protection circuitry... it's REALLY bad for the sound when protection kicks in... and although intended to come in at a certain point it's a "sliding scale" and the more it needs protecting, the harder the protection kicks in. It can start to effect things well before obvious distortion is audible! This is one of those things that are really important for sound but hardly gets mentioned!!
The main implication is that for SS amps of lets say 40-50W and above (the higher the power into 8R the worse it gets), that have only one pair of output devices, then the way they sound at especially higher volumes and into more awkward loads can often have more to do with the design of the protection circuitry and less to such things as power supplies than many would guess
As exhibit A and B we have the Quad 405 Mk1 and 303 as extreme examples but it can happen more subtly in many other amps.
Of course if you have 8R speakers that are not "the load from hell" and don't play it really loud it's not going to even be an issue.... but if you have 4R speakers or really awkward ones and certain SS amps sound crap at higher levels then there's a very good chance this is whats going on.