First of all, true "full Class A" amps generate a massive amount of heat.
A full-on Class A amplifier is about 25% efficient at full power...
But, more specifically, they consume about 4 TIMES their rated output power, REGARDLESS OF THE LEVEL THEY'RE RUNNING AT.
So, for example, a 100 watt/channel stereo Class A amplifier consumes about 800 watts at all times, with all of the excess ending up as heat.
(So it makes almost 800 watts of heat - whether running at full power or just idling.)
At full power a Class A/B amp, like the XPA and XPR, is about 70% efficient.
And, at idle, you're looking at "idle current" (since there is no output the efficiency is technically 0).
A Class H amp, like the XPR, has similar efficiency at full power, and at idle, but is somewhat more efficient at medium power levels.
So, at full output, they will generate quite a bit of heat, but that will reduce at lower Volume levels.
Depending on the bias setting an XPR will dissipate a few hundred watts at idle and will run somewhat warm.
(And, because there is a lot of mass, it will heat up slowly, but still needs air flow to maintain some level of cooling.)
The bias on the XPR amps can drift over time...
The bias adjustment pots also occasionally fail or drift over time (and so need to be replaced)...
These are both things that a highly qualified technician could do...
However, the XPR amps are somewhat difficult to work on, so we DO NOT recommend this as something an "end user" should attempt.
(One side running a little hotter than the other is not a significant issue.)
In general, any amplifier that doesn't have fans uses "convection cooling".
Basically, as the amp warms up, the hot air rises, and goes out the top, pulling more cool air in from the bottom or other vents.
(This works the same way even if the heat sinks are on the outside.)
Therefore the "cooling" doesn't actually occur until the warm air is warm enough to rise.
Also note that, because "warm air rising" doesn't do so with a lot of force, it's important to make sure that vents and such remain unobstructed.
(Otherwise warm air may be trapped inside.)
When you add fans to an amplifier you need to pay some attention to the way air is INTENDED to flow.
For example, if you're putting that amp in an enclosure, having one or two fans at the top "to move the hot air out and make room for cool air" may be sufficient.
But you really would NOT want low powered fans pushing air in at the top - because they would be preventing the hot air from leaving that way.
And, if you puts fans on individual amplifiers, especially ones with heat sinks inside, you want to "encourage air to flow the way it was intended to".
(That usually means pulling air out the top, and making sure vents remain unblocked, so cool air can enter through the bottom and/or sides where the vents are.)
It also seems obvious but...
How hot an amp gets will depend A LOT on how much air flow it has.
Even a big amplifier, that doesn't generate much "bulk heat", will run VERY hot if you shut it in a cabinet where that heat cannot get out.
Likewise, even a relatively small or slow fan, that effectively "moves the hot air away", can get rid of a LOT of heat.
Also, on a final note....
That 25% efficiency is a solid limit for Class A amps (a pure Class A amp
CANNOT have higher efficiency; although many "almost Class A amps" do).
Likewise, 70% is really a sort of top limit on the efficiency of Class B and Class A/B amps (and that includes Class G and Class H - which have better efficiency at medium power levels).
(A "high bias" Class A/B amplifier will, by necessity, fall somewhere in-between.)
(A really apt analogy is when you set a car to "idle higher" so "it runs smoother and warms up quicker"... but at the expense of fuel economy... and of burning a lot of fuel at idle.)
However, Class D amps, including those that use the well known Hypex and ICE modules, operate on entirely different principles, and CAN have efficiency at or above 90%.
They can also have high efficiency, and very low idle power consumption, at very low power levels.
Note, though, that Class D amps do still generate some heat, so you should be careful to avoid blocking whatever heat sinks or vents they do have.
(Even a very efficient Class D amp, which uses its cabinet as a heat sink, may still overheat if you were to seal it up in a cabinet with no air flow at all.)
I'm surprised Keith hasn't gotten involved here.
EMO should know how hot any of there amps run.....and especially anything with high-bias or what they pass-off as Class A.......