More power for moderate listening levels?

My amp puts out 500 watts into my 4 Ohm speakers. Sounds great. I added a T+A power supply that adds 1800 watts. I rarely play music over 84 decibels, generally in the 64 to 78 range at my listening chair.  With the extra 1800 watts I found listening at low levels to be more enjoyable & where the biggest difference was. You can actually feel the music at low volume levels, incredible dynamics also. 

 I added a T+A power supply that adds 1800 watts.

Power is V x A.
Increasing 4Ω power from 500 to 1800 requires changing the voltage from 45 to 85. Current would increase from 11 to 22A. It very unlikely the output devices were that over spec'd and the heat almost certainly wasn't.

FanBoy prattle.

@brianbiehs

If I got this right and the Harbeth is 6 ohms and 86 dB, its actual efficiency is around 84dB. That isn't quite criminally low efficiency, but it is quite low and in most rooms you're going to need about 600 watts to make it play. This is why the amp is being overloaded (which BTW also puts the tweeter in the speaker at risk; that's the classic method to damage a tweeter).


600 watts is a lot of power and that sort of requirement makes it a lot harder and more expensive to find an amp that sounds as good as one that makes only 100 watts. Further, there's really no reason to make a speaker that inefficient. It isn't bandwidth, it isn't speed, it isn't resolution. Plus when speakers are that inefficient, you get thermal compression caused by heating in the voice coils.


If you are driving that speaker with 100 watts (which is about what the Pass will do at clipping) it will make a certain sound pressure. But if the speaker were 94 dB instead of 84, you would need only 10(!) watts to make the same sound pressure. Now my speakers at home are 98dB and just to avoid doing the math I'm going to say for this argument that they are only 97dB... so 3db more than 94dB means that I can make the same sound pressure as you can with only 5 watts. I can tell you that in my listening room which is slightly on the small side that even with my speakers that isn't nearly enough. BTW a 3dB difference is a doubling of amplifier power, a 10dB difference is 10x the amplifier power. Decibels (dB) are on a logarithmic scale since that's how our ears hear.


I'm running an amp that makes about 45 watts/channel and its impossible to clip it in my room. To put that in perspective, that would be like having an amp that has 900 watts/channel in your system. In simple terms you're a country mile off when it comes to matching these two products.


The simple solution here is that if you like that amp and don't feel like going on the hunt for a 600-900w/channel **integrated** amplifier, change out the speakers for something more efficient! The combo might sound great at low levels, but if you want to hit anything that might be considered 'satisfying' its not going to work as you well know.

@atmasphere 
Here's Stereophile's test of the Harbeth SLH5 plus.
https://www.stereophile.com/content/harbeth-super-hl5plus-loudspeaker-measurements
It drops to 6 ohms at one point, and just below 6 ohms at another point.  But for much of the range it is over 8, and for some of the range significantly over 8.  With the degree of variation of impedance that one can see in the plot, how does one assign a nominal single value in ohms to describe the impedance of this speaker?

how does one assign a nominal single value in ohms to describe the impedance of this speaker?

Everyone does it different I imagine. 'Nominally' this looks like an 8 ohm speaker- if so that throws off my efficiency calculation by about 1.5dB. The impedance curve looks fairly benign, which is good. Its really the efficiency/sensitivity (which since its nominally 8 ohms are about the same thing) which is the issue here. Its assumed from the plot that the amplifier won't make a lot of power driving the impedance peaks- since the amp is a voltage source and the speaker is intended for that.