Amplifiers - how big, how small? The truth.

[A.S.]

We constantly receive emails from anxious customers asking which amplifier they should use with their Harbeth speakers. Will XYZ amp to too powerful? Will ABC amp not be powerful enough?

Of all the questions here, this amplifier issue seems to be the one that causes the most (needless) concern. Amplifier power capacity is a simple technical matter perverted by downright falsehoods, mumbo jumbo, irresponsible gossip and no concept of what an amplifier actually does. I've tried to allay these fears in various postings that I've made here over the years (clearly without success) so here for the final time is my opinion in one posting. Note that I have not made any comment about the claimed sonic benefits of amplifiers - that is another matter which I will leave entirely alone for you to decide for yourselves. But it's unlikely that there is a perfect correlation between amplifier power and sound quality. So let's just look at amplifier power rating.

Typical questions that you ask us almost daily and our answers ....


You ask: "Can I use a really powerful amplifier with my speakers? For example, can I use a 250W amplifier with my Compact 7s?"

My answer: It doesn't matter if the amplifier is rated at 50W or 250W or even 2500W. What matters is how you use the available power.

You ask: "Really? Surely the big amplifier has 'too much power ' for the speakers?"

My answer: No. It is irrelevant how big the amplifier is. The amplifier could be 1,000,000W (one maga watt).

You ask: "But surely that is too much power for the speaker?"

My answer: Yes, that is a huge amount of power but ... who is in command of the power? You or the amplifier?

You ask: "Surely the amplifier forces power into the speaker so that if the amplifier is too big (too many watts) it will destroy the speakers?"

My answer: Who is in command of that power? The amplifier doesn't have a brain. It is dumb. It is your servant. It does what you command it to do.

You ask: "What do you mean I am in command?"

My answer: Is your hand on the volume control? If so, YOU are in command of the amplifier. YOU and you alone decide how much loudness, how much volume and how much power will be made available to the speaker according to YOUR setting of the volume control.

You ask: "So the volume control is like a bath tap, and if the tap is only open a little, only a small stream of power will flow. If the tap is fully open, a torrent of power will flow. Right?"

My answer: Absolutely correct. YOU decide upon the power level you wish to deliver to your speakers. The amplifier doesn't sneak power past the volume control into the speakers! The volume control is the power-control gate and YOU control the gate! The water from the reservoir does not gush into your bath flooding your house - it is controlled by the tap.

You ask: "So the amplifier does not push or force the power onto the speakers, the speaker is permitted by me to draw power according to the setting of the volume control? So actually it does not matter how big the amplifier is, what matters is how I command the power by adjusting the volume tap."

My answer: Absolutely correct! You are the boss! You are in command. If you have a big amplifier act responsibly. Do not play too loud!

[A.S.]

The other question we are frequently asked is about the minimum recommended power for an amplifier. This question is a little more complex that the 'too big?' question as I'll try and explain here. When I say smaller amplifier or bigger amplifier I don't mean the size of the case, I mean only the power output rating of the amp. The case size is irrelevant: the circuitry may only occupy a small part of the inside. You cannot judge the power rating of an amplifier by the size of the case! Marketing people are smart enough to disguise small amplifier circuits in big cases! You have to check the specification. It will say something like "Power rating .... 75W into 8 ohms continuously per channel".


You ask: "If I look inside a small amplifier or a big amplifier what difference would I expect to see?"

My answer: The circuit may be exactly the same. But the bigger amplifier has more of a reservoir of power compared with the small amplifier.

You ask: "What do you mean that the bigger amplifier has 'more of a reservoir of power'?

My answer: I mean that the bigger amplifier has a bigger transformer and bigger PSU capacitors so that it can draw more power from the mains and store that inside the circuit ready to deliver to the speakers according to how you adjust the volume control.

You ask: "So if the speakers are not connected, or if the volume control is at zero, no power is being drawn from the reservoir so that It doesn't matter if the reservoir is small or large. Right?

My answer: Correct. The only (technical) advantage of a big amplifier is when you are playing loud. At low volumes, there are no (technical) advantages to having a big power amplifier. If you are not actually drawing from the power reservoir then it doesn't matter how big the reservoir actually is. It's like the reservoir from which you draw water for your home. If you don't turn on your water taps then it's irrelevant how much water there is in the reservoir - it could even by completely dry.

You ask: "How much power is needed to play music at home?"

My answer: It depends upon how you set the volume control, the type of music, and how loud you like to listen, and how good your hearing is.

You ask: "OK, but typically, how much power is needed to play at a moderate listening level?"

My answer: Surprisingly little. To make a medium-loud sound on classical music (or jazz) you only need about 5W -10W or so.

You ask: "Really? Only about 10W? But why would I need an amplifier of more power than 10W then?"

My answer: Because a 10W will give you no reserve power because it has no power reservoir within the case from which you can draw immediate power to handle loud peaks. It will have small capacitors to store the power and/or a small transformer and/or it will waste much power as useless heat. So the available power reserve for the speakers will be very small.

You ask: "So it's a good idea to have some reserve capacity available on demand for those sudden peaks in the music, or where I chose to briefly play an exciting musical piece louder than normal?"

My answer: Yes it it always a good idea to have some reserve power. But how much? If 10W will fill a room with sound, then 50W or 100W should give plenty or reserve power. Anything more than about 100W is just a luxury - a power reserve that you will never dare command from the amplifier.

You ask: "Never dare? What do you mean?

My answer: I meant that even though you have a massive reserve available in a big amp, commanded by your operation of the volume control, you have to ask yourself can the speaker handle so much power?

You ask: "But I am always careful and responsible when adjusting the volume. I have neighbours and they will not tolerate too much volume from my hi-fi system. So I will never strain the speakers."

My answer: That's good. No matter how big the amplifier, if you are careful then you won't strain the speakers.

You ask: We're talking here about how little power I need. You said 10W is typically all that's necessary to make a reasonably loud sound. Is there any advantage to using a small amplifier? Maybe I should buy a small amp? Or a big one? I can't decide.

My answer: As I've said, let's pick a minimum power rating of (say) 25W. It could be 15W or 30W, something in that range would provide some reserve for louder than normal use. At the other end, perhaps about 150W represents all the power you would ever sensibly need before you risk damaging your hearing and if you are careless with setting the volume, your speakers too.

You ask: "So if I pick an amplifier in the 20W - 150W range I should be able to make great music at home?"

My answer: Yes; the lower powered amplifier will fill the room with sound but will not have much reserve power. The 150W amplifier will have a lot of reserve which you must command carefully. You must respect the fact that the speakers can't handle 150W continuously.

To be continued ...

[A.S.]

Last week I was carefully measuring a selection of amplifiers to give us an objective performance base line. The results were rather interesting.

First, even expensive amplifiers with a good pedigree and long warranty produce distortion. All amps do. The nature of that distortion varies somewhat from specimen to specimen. Surprisingly (to me) the individual distortion components down on the noise floor are granular in nature; that means those distortions are at many frequencies that are not immediately obvious to me as harmonically related to the test signal* and spread right across the audio band, not just the more familiar and easily identifiable second, third .... distortions of the input test signal.They may be there as well.

What has come to light though is - and this is just my initial comment, I may be wrong - that amp manufacturers seem to shelter behind a presentation of the distortion (and noise) components on paper. First, amplifier distortion (plus noise) seems to be quoted when the amp is at or near full power. As the background hiss (noise) is fairly constant (it seems) that this noise can then be quoted as a percentage of the full power, which on paper, makes it seem lower than it perhaps is to the ear with little or no signal passing through the amp and speakers.

Second: Maybe my maths is wrong, but let's take an typical high quality amplifier manufacturer's specification for distortion and noise - say, "0.05%". Looks extremely impressive doesn't it as it is such a small number. Now, as I understand it, 0.05% equates to five parts of distortion in ten thousand parts of signal or the distortion is 1/2000th of the total output. Again, impressively small number. But let's convert that to a log number and express the distortion in decibels as we do generally with audio things. I calculate that 0.05% distortion as -66dB relative to the full output. Now, that's not so impressive a number (on paper).

Please check my maths.

P.S. I subsequently found this on-line tool for number conversion which you may find useful.

* With a little personal experience with amp design in the previous HHB and Monitor 40Active modules, I suspect that these very low level distortions near the noise floor are related to circulating (earth) currents in the PCB's, wire looms and power supply. Extremely small differences in any of these will definitely change the distortion spectral display; whether or not they are audible I doubt at such a low contribution level to the overall signal.

[A.S.]

More on small amplifiers and power reserve here