Amplifier Capacitance

Okay, naive question of the day: except for cost and perhaps size limitations inside the casing, why don't manufacturers go bonzo large on capacitance? For instance, I'm thinking of replacing the caps in this damaged Hafler DH220 I've got and, while I probably won't find any that fit, I started to wonder why limited myself to two 18-19k cans? Why not 50k, 80k, even 100k if I could fit it? (I doubt any of those would, but you get the drift.).

much can be written about this but I'll try to provide a compact answer ;-0
large amounts of power supply does not come for free. you'll need a high(er) current transformer to supply the current to charge up those huge caps in rapid time so that the power supply voltage does not sag when a big slug of charge is demanded by the power amp electronics. Large power supply caps are huge reservoirs of charge, true, but they take time to recharge. Often this recharge time is too slow to keep pace w/ the music hence you'll often find smaller caps (like 10,000uF) in parallel w/ the huge power supply caps. these smaller caps are there to provide the charge for sharp transients thereby taking the load off the large power supply caps. The smaller caps are more nimble in terms of recharging hence they work in complimentary w/ the larger power supply caps.
So, note the transformer current delivery capability before you go "bonzo" on increasing the caps. if you did, you'll saturate the transformer core & bring in distortion into your power supply damning your Hafler sonics. Over time you'll also damage the power transformer.
Also, the larger the power supply cap, the larger the power-on current surge. My amp has a lot of power supply capacitance & it takes 37A in-rush current upon power on!! :-o. The amp designer has to consider this when designing the power supply wiring & choosing the AC fuse(s).

Capacitance seems like once of those things that should really make a big diffence in amp performance, no? In fact, shouldn't it particularly help offset a somewhat weaker power supply as well?

yes, power supply capacitance can make a big difference in the power amp's power supply ONLY if the power supply is designed to take advantage of the larger capacitance amount.
Huge amounts of capacitance cannot hide the flaws in a weaker power supply. In fact, large capacitance will quickly bring a weaker power supply to its knees as it'll expose its weakness quickly.
hope that this sheds some light on the matter....

excellent reply Kijanki.

Aewhistory, if you read my reply & Kijanki's you should have all your questions answered. :-)

This is why an amplifier can 'motorboat' (repeated thump) if a filter capacitor fails in the supply- the timing constant has become so high that the amplifier exhibits low frequency instability.

I recently bought an NAD 2600 with a disclosed possible flaw--a "thump" when it turns on and off. I got it for about $100, so it think it was a reasonable price. Amp plays great, but that on/off sound isn't so nice. So I had posted about this to confirm this was an issue and not just "the way this amp is" and someone confirmed this was a flaw. So my question is: is this the 'motorboat' thump to which you are referring Atmasphere, or is this a different problem?

the turn on/off thump in your NAD amplifier is a totally different issue compared to what Atmasphere is talking about.
The thump that you are hearing during on/off is the inrush current charging the capacitors. this (huge) inrush current will create a spike on the supply rails of the power amp which in turn will create a voltage spike on the power amp output terminals (speaker binding posts). This in turn will create a thump in your speakers. The issue with the NAD is that it never had any output protection circuitry to avoid 'thumping' the speakers (if the NAD had output protection circuitry that was malfunctioning, that amp would not play any music). This is indeed a design flaw. Back in the 1980s, I & some extended family members did own integrated amps that always thumped the speakers. I guess back then output protection circuits were not always put into cheaper amplifiers - just the way it was back then!
The thump that Atmasphere is talking about is when a power supply cap fails (bad part/age/heat, etc) & the power supply is now faulty. Then, the power amp can motorboat/thump which is a sign of low freq instability.
Two different issues but quite similar symptoms.
The thing to note about the motorboating is that the amp will thump every time it's playing low freq content.
The turn on/off thump occurs just once at turn on & once at turn off - no thumping while playing program material.

If the amplifier has a -3 db point that is a frequency lower than that of the timing constant of the power supply, then the amplifier can modulate the supply, which results in IM distortion amongst other things.

Ralph, let me see if understand what you wrote above. I'm getting confused because w.r.t. the amplifier you are using frequency & w.r.t. the power supply you are using timing constant. Freq & timing constant are reciprocals of each other.

So, what you are saying is that for motorboating to occur the time constant of the amplifier has to be higher ("If the amplifier has a -3 db point that is a frequency lower than...") than the time constant of the power supply. Correct?
So, basically, the bandwidth of the power supply should always be lower than the bandwidth of the power amp. Correct?

but once we draw less than max current, power supply voltage modulation at low frequencies becomes compensated by amplifier (since it is regulated).

Kijanki, I don't understand this statement. Power amp power supplies are not regulated (I don't see a feedback loop around the power supply). The power amp itself has minimal feedback around itself for sonic purposes (as you already know) & this feedback is for the music signal & not for any vairations of the power supply. The amp is counting on the power supply to be essentially DC. So, any power supply modulation will modulate the output voltage signal (AM-AM modulation/distortion).
I'm afraid that I'm not seeing how the power supply modulation at lower/bass freq is compensated by the power amp. Thanks.

Amplifiers are line regulated. It means that amplifier supplied from 40V and set to produce 5V output voltage will still produce 5V output with supply lowered to 35V or increased to 45V.

Kijanki, I can believe this BUT I do not think that this is due to "line regulation".
If the power supply rail is 50V (or 45V) & the power amp is outputting a 5V signal, there is plenty of headroom for the output transistors (& the rest of the power amp stages) such that if the power supply rail goes up/down by 5V, the output voltage will not change. I believe that this is the reason that the power amp will continue to output the voltage its being asked to.
If one keeps on lowering the power supply rail, one will come to a voltage where the BJTs will saturate/the MOSFETs will triode & the output voltage will drop as the devices go into the linear region.
I still believe that all power amp stages are very sensitive to the cleanliness of the power amp power supply rail.
thanks.