What is a powerful amp?

[hybridauth_Facebook_664715932]

Hi All,

what really is a powerful amp? does power equal to loudness of the amp or is it more of a overall thing?

for instance my amp is an evo 50a theres also evo 100a which is supposed to be twice as powerful so does it mean it is going to be twice as loud.

I don't really listen to music very loud, most of the @ -35db. Does it mean that buying a powerful amp on my next upgrade would be a waste of money?

thanks

 

[matt49]

Amp power has to be seen in the context of speaker sensitivity and impedance. A powerful amp is an amp that drives your speakers to the required volume without clipping.

Chasing more power for the sake of it is pretty pointless, especially when you consider that doubling amp power output yields a 3dB increase in volume.

 

[drummerman]

Amp power has to be seen in the context of speaker sensitivity and impedance. A powerful amp is an amp that drives your speakers to the required volume without clipping.

Chasing more power for the sake of it is pretty pointless, especially when you consider that doubling amp power output yields a 3dB increase in volume.

Probably as good an explanation as you are likely to get imo.

 

[davedotco]

Amp power has to be seen in the context of speaker sensitivity and impedance. A powerful amp is an amp that drives your speakers to the required volume without clipping.

Chasing more power for the sake of it is pretty pointless, especially when you consider that doubling amp power output yields a 3dB increase in volume.

Exactly.

It is also worth pointing out that a 3dB increase in volume (measured spl) is pretty small.

If you were playing a piece of music and turned it up just enough that you could perceive a small but definite increase in volume, that will be around 3dB.

Twice the volume (measured spl) requires 10 times the power.

Please note these are scientific measurements, not open to debate.

Perceived loudness is something else, it is the subjective and personal evaluation of measured spl. In tests people perceive 'twice as loud' very differently, some require a much larger increase than others.

A reasonable rule of thumb, from empirical data, is that the average person interprets 'twice as loud' to an increase in measured spl of around 8dB, however this will vary enomously with different sounds and volumes etc.

 

[matt49]

It is also worth pointing out that a 3dB increase in volume (measured spl) is pretty small.

If you were playing a piece of music and turned it up just enough that you could perceive a small but definite increase in volume, that will be around 3dB.

Twice the volume (measured spl) requires 10 times the power.

Please note these are scientific measurements, not open to debate.

Perceived loudness is something else, it is the subjective and personal evaluation of measured spl. In tests people perceive 'twice as loud' very differently, some require a much larger increase than others.

A reasonable rule of thumb, from empirical data, is that the average person interprets 'twice as loud' to an increase in measured spl of around 8dB, however this will vary enomously with different sounds and volumes etc.

Interesting.

The big and rarely answered question is how much amp power is actually needed to drive normal speakers to the required level without clipping. Let's assume you want to be able to turn the wick up now and again so that you can double the volume and (according to Dave's model) achieve an SPL increase of 8dB. If doubling of power only yields 3dB increase in SPL, then you're going to need a lot of extra oomph.

Imagine a real-world situation: you want loudish music in a medium-sized room from a pair of speakers with moderately high sensitivity (90dB) and a fairly benign impedance curve. I've read different views on how much amp power is needed. The highest plausible figure I've seen is 500wpc. That's a big amp, way bigger than what most people use.

Any thoughts?

 

[drummerman]

It is also worth pointing out that a 3dB increase in volume (measured spl) is pretty small.

If you were playing a piece of music and turned it up just enough that you could perceive a small but definite increase in volume, that will be around 3dB.

Twice the volume (measured spl) requires 10 times the power.

Please note these are scientific measurements, not open to debate.

Perceived loudness is something else, it is the subjective and personal evaluation of measured spl. In tests people perceive 'twice as loud' very differently, some require a much larger increase than others.

A reasonable rule of thumb, from empirical data, is that the average person interprets 'twice as loud' to an increase in measured spl of around 8dB, however this will vary enomously with different sounds and volumes etc.

Interesting.

The big and rarely answered question is how much amp power is actually needed to drive normal speakers to the required level without clipping. Let's assume you want to be able to turn the wick up now and again so that you can double the volume and (according to Dave's model) achieve an SPL increase of 8dB. If doubling of power only yields 3dB increase in SPL, then you're going to need a lot of extra oomph.

Imagine a real-world situation: you want loudish music in a medium-sized room from a pair of speakers with moderately high sensitivity (90dB) and a fairly benign impedance curve. I've read different views on how much amp power is needed. The highest plausible figure I've seen is 500wpc. That's a big amp, way bigger than what most people use.

Any thoughts?

Yes, I remember Michaelson (MF) preaching the importance of multi hundred watt amplifiers and yet ... they make 50w Class A monsters.

Bearing in mind that most of us probably use between half a watt and at most perhaps ten, it would be reasonable to assume that even if dynamic peaks reach ten times the average, a hundred watts would be more than enough for most situations with reasonably sensitive speakers.

I'd go as far and say that even half of that is enough for most folks.

I had a twenty watt/channel tube amp and even with that I rarely wanted for more. That was using IB speakers.

 

[CnoEvil]

35W is all I need into a Sensitivity of 90dB (but an impedance that drops well below 4 Ohms).

 

[matt49]

Yes, I remember Michaelson (MF) preaching the importance of multi hundred watt amplifiers and yet ... they make 50w Class A monsters.

MF used to make low-powered Class A amps, but no longer. Their current flagship integrated amp is the Nu-Vista 800 at 300wpc. Or you could opt for their top-of-the-range monobloc power amp, the M8-700m, at 700wpc.

I had a twenty watt/channel tube amp and even with that I rarely wanted for more. That was using IB speakers.

Rules are different for tube amps as they soft clip. It sounds nicer than tranny hard clipping, but it's still clipping.

35W is all I need into a Sensitivity of 90dB (but an impedance that drops well below 4 Ohms).

You may be right. I honestly don't know the answer to my question. But I'd like to know what you'd think of your speakers driven by a 500w amp.

 

[mitchellhatpeg]

I would say that a 500 watt amp to drive normal speakers would be a total waste of power.

I run Spendor SA1s, with only 85 dB per watt efficency with a 25 watt amp, and it will go louder than comfortable for my ears with no distortion. The amp is a Temple Audio Bantam Gold which, unlike other amps, is 90% efficient. This means you can effectively use the whole of the volume range without distortion, instead of 1/3 to 1/2 as with most other amps.

If you want louder, add a sub woofer ( or two)

 

[davedotco]

It is also worth pointing out that a 3dB increase in volume (measured spl) is pretty small.

If you were playing a piece of music and turned it up just enough that you could perceive a small but definite increase in volume, that will be around 3dB.

Twice the volume (measured spl) requires 10 times the power.

Please note these are scientific measurements, not open to debate.

Perceived loudness is something else, it is the subjective and personal evaluation of measured spl. In tests people perceive 'twice as loud' very differently, some require a much larger increase than others.

A reasonable rule of thumb, from empirical data, is that the average person interprets 'twice as loud' to an increase in measured spl of around 8dB, however this will vary enomously with different sounds and volumes etc.

Interesting.

The big and rarely answered question is how much amp power is actually needed to drive normal speakers to the required level without clipping. Let's assume you want to be able to turn the wick up now and again so that you can double the volume and (according to Dave's model) achieve an SPL increase of 8dB. If doubling of power only yields 3dB increase in SPL, then you're going to need a lot of extra oomph.

Imagine a real-world situation: you want loudish music in a medium-sized room from a pair of speakers with moderately high sensitivity (90dB) and a fairly benign impedance curve. I've read different views on how much amp power is needed. The highest plausible figure I've seen is 500wpc. That's a big amp, way bigger than what most people use.

Any thoughts?

Yes, I remember Michaelson (MF) preaching the importance of multi hundred watt amplifiers and yet ... they make 50w Class A monsters.

Bearing in mind that most of us probably use between half a watt and at most perhaps ten, it would be reasonable to assume that even if dynamic peaks reach ten times the average, a hundred watts would be more than enough for most situations with reasonably sensitive speakers.

I'd go as far and say that even half of that is enough for most folks.

I had a twenty watt/channel tube amp and even with that I rarely wanted for more. That was using IB speakers.

This is a fairly complex subject, even matt appears unclear on the difference between perceived loudness and measured spl.

In reality, the power thing is pretty simple. In a normal domestic room, playing music at socially acceptable levels does not require huge power. Good recordings can have a dynamic range around 40 dB, ie musical peaks are 20dB above average levels, ie 100 times the power. As has been pointed out, average systems in average rooms tend to use less than 1 watt, so 50 to 100 watts (peak) will be enough most of the time.

The problem is when things move away from the average, less sensitive speakers, a larger room, more people, higher background noise levels, a penchant for turning up the bass, all these factors require power, lots of it. People underestimate just how much extra power is needed when these factors come into play, lets look at an example.

Two people in a decent sized room playing a 50 watt, average sensitivity system quite loud, two speakers, 87dB watt (say), listening at 3-4 meters. In these circumstances a fraction of a watt is all that is needed, the 50 watt amp (70 watts peak) is ample.

So what happens next, a few friends arrive, there bodies absorb power so the system is turned up a bit, 3-6dB maybe, everyone is getting into it so push the bass up a notch or two, another 3-6dB power required, a touch more on the volume, it's fun, another 3-6dB. A couple more people, a bit of chatter so you turn the amp up to compensate, another few dBs.

Before you even get to full party mode, you are already asking the system to deliver 15-20dB more than originally, that could be as much as 100 (20dB) times more. The amplifier that was originally being asked to produce a few handfuls of watts on peaks is now asked to produce much more, 10-15 watt peaks that were entirely adequate initially are raised by up to 20dB, ie 1000-1500 watt peak.

Thats right, peak power in excess of 1000 watts, easily done.

 

[ChrisIRL]

It has been suggested to me that my Rega Elex-r, rated at 72.W into 8ohm couldn't possibly be adequate for the SCM11s I use which are 85db sensitivity and are a mostly steady 8ohm resistance. Surely for moderate, say 70 to 80db listening, sitting 12 feet from the speakers in an average sized room the above comments would suggest it is plenty?

 

[MajorFubar]

It is. Lucky if you're using 10W RMS at most.

 

[davedotco]

It has been suggested to me that my Rega Elex-r, rated at 72.W into 8ohm couldn't possibly be adequate for the SCM11s I use which are 85db sensitivity and are a mostly steady 8ohm resistance. Surely for moderate, say 70 to 80db listening, sitting 12 feet from the speakers in an average sized room the above comments would suggest it is plenty?

It is.

Read my post above to understand when it is not.

 

[Vladimir]

1200W RMS continuous in 2 ohms <0.1% THD.

 

[ChrisIRL]

Anyone hazard a guess as to why ATC themselves recommend a minimum 75W amp then? It appears to be misleading people who might otherwise be interested.

 

[ChrisIRL]

Based also on the above that a 75wpc amp is adequate for the conditions I've described, another opinion you regularly get is that a more powerful amp will better "control" a speaker, even at the same volume. Is this nonsense? i.e. can a 150w amp provide anything that a 75w can't under the conditions I've described above. I'm not talking about damping factor or anything, both amps the exact same in all regards except watts.

 

[MajorFubar]

Is this nonsense?

It is in my opinion. I've read explanations where people have used analogies of car engines, where a 100bhp engine shifting 1,500kg of car at 70mph is struggling more than a 300bhp engine shifting the same car at the same speed because one engine's sweating its nuts off at the near-limit of its power and the other has barely jogged past its walking pace. But I really don't think it's an applicable analogy IMO, because achieving 70-80db in an average home environment doesn't even need a '100bhp' amp.

 

[CnoEvil]

Acalex, Roby and Neuphonix all directly compared the M600i with the 35i, and all preferred the 35i....so it's not necessarily all about ultimate power.

I've heard my speakers with Linn Klimax Solos, Coda CSI, Moon i7 and Rega Osiris...and still wouldn't swap my amp for any of them.

 

[hybridauth_Facebook_100004211376004]

Does the power supply to the amp itself also determine how well it can drive some sensitive speakers?

 

[Andrewjvt]

With most things written on this post from all the others but in my case changing from the same make amp 150w to the 250w amp made an overall improvement to the sound from the top end and mostly the bass

 

[davedotco]

Does the power supply to the amp itself also determine how well it can drive some sensitive speakers?

An amps power supply can be pivotal in the way an amplifier behaves. Lets take a typical modern amplifier designed to produce a maximum continuous voltage of 20volts. The power supply is almost irrelevant here, with 200+volts of mains power 20volts (strictly +/- 20volts) is easy.

The power delivered by such an amplifier is easily calculated, Vsquared/R, so for an 8ohm load 50watts, a 4ohm load 100watts, 2 ohms 200watts etc.

As we know, few amplifiers can actually manage this and the limitations of the power supply is the reason. Look at the current required in each case, for 8ohm it is 5amps, 4 ohm 10amps, 2 ohm 20 amps, that is a lot of current that requires big capacitors to supply it and a big transformer to keep the capacitors topped up, that is expensive, very expensive.

In a recent thread someone was comparing the Marantz PM6005 with the Creek 50a, most posters favouring the more expensive Creek. Although specified at pretty much the same power, the Marantz has a 150 watt power supply, the Creek 350 watt. Which do you think is going to be more effective in use?

 

[drummerman]

It is also worth pointing out that a 3dB increase in volume (measured spl) is pretty small.

If you were playing a piece of music and turned it up just enough that you could perceive a small but definite increase in volume, that will be around 3dB.

Twice the volume (measured spl) requires 10 times the power.

Please note these are scientific measurements, not open to debate.

Perceived loudness is something else, it is the subjective and personal evaluation of measured spl. In tests people perceive 'twice as loud' very differently, some require a much larger increase than others.

A reasonable rule of thumb, from empirical data, is that the average person interprets 'twice as loud' to an increase in measured spl of around 8dB, however this will vary enomously with different sounds and volumes etc.

Interesting.

The big and rarely answered question is how much amp power is actually needed to drive normal speakers to the required level without clipping. Let's assume you want to be able to turn the wick up now and again so that you can double the volume and (according to Dave's model) achieve an SPL increase of 8dB. If doubling of power only yields 3dB increase in SPL, then you're going to need a lot of extra oomph.

Imagine a real-world situation: you want loudish music in a medium-sized room from a pair of speakers with moderately high sensitivity (90dB) and a fairly benign impedance curve. I've read different views on how much amp power is needed. The highest plausible figure I've seen is 500wpc. That's a big amp, way bigger than what most people use.

Any thoughts?

Yes, I remember Michaelson (MF) preaching the importance of multi hundred watt amplifiers and yet ... they make 50w Class A monsters.

Bearing in mind that most of us probably use between half a watt and at most perhaps ten, it would be reasonable to assume that even if dynamic peaks reach ten times the average, a hundred watts would be more than enough for most situations with reasonably sensitive speakers.

I'd go as far and say that even half of that is enough for most folks.

I had a twenty watt/channel tube amp and even with that I rarely wanted for more. That was using IB speakers.

This is a fairly complex subject, even matt appears unclear on the difference between perceived loudness and measured spl.

In reality, the power thing is pretty simple. In a normal domestic room, playing music at socially acceptable levels does not require huge power. Good recordings can have a dynamic range around 40 dB, ie musical peaks are 20dB above average levels, ie 100 times the power. As has been pointed out, average systems in average rooms tend to use less than 1 watt, so 50 to 100 watts (peak) will be enough most of the time.

The problem is when things move away from the average, less sensitive speakers, a larger room, more people, higher background noise levels, a penchant for turning up the bass, all these factors require power, lots of it. People underestimate just how much extra power is needed when these factors come into play, lets look at an example.

Two people in a decent sized room playing a 50 watt, average sensitivity system quite loud, two speakers, 87dB watt (say), listening at 3-4 meters. In these circumstances a fraction of a watt is all that is needed, the 50 watt amp (70 watts peak) is ample.

So what happens next, a few friends arrive, there bodies absorb power so the system is turned up a bit, 3-6dB maybe, everyone is getting into it so push the bass up a notch or two, another 3-6dB power required, a touch more on the volume, it's fun, another 3-6dB. A couple more people, a bit of chatter so you turn the amp up to compensate, another few dBs.

Before you even get to full party mode, you are already asking the system to deliver 15-20dB more than originally, that could be as much as 100 (20dB) times more. The amplifier that was originally being asked to produce a few handfuls of watts on peaks is now asked to produce much more, 10-15 watt peaks that were entirely adequate initially are raised by up to 20dB, ie 1000-1500 watt peak.

Thats right, peak power in excess of 1000 watts, easily done.

Uuhh?

 

[davedotco]

So what happens next, a few friends arrive, there bodies absorb power so the system is turned up a bit, 3-6dB maybe, everyone is getting into it so push the bass up a notch or two, another 3-6dB power required, a touch more on the volume, it's fun, another 3-6dB. A couple more people, a bit of chatter so you turn the amp up to compensate, another few dBs.[/b]

Before you even get to full party mode, you are already asking the system to deliver 15-20dB more than originally, that could be as much as 100 (20dB) times more. The amplifier that was originally being asked to produce a few handfuls of watts on peaks is now asked to produce much more, 10-15 watt peaks that were entirely adequate initially are raised by up to 20dB, ie 1000-1500 watt peak.

Thats right, peak power in excess of 1000 watts, easily done.

Uuhh?

[/quote]

Just a simple example of how little power is needed for normal listening and how much power is needed in more demanding situations.

 

[matt49]

Just a simple example of how little power is needed for normal listening and how much power is needed in more demanding situations.

Indeed. Again context is crucial.

 

[NS496]

Keep in mind the 'phase changes' of the reactive speaker load, which increase output transistor dissipation and can drive the transformer to its knees if not robust enough. Depends on speaker impedance and phase angles. A good amplifier will have enough capacity in it's transistors to handle these challenges presented by a 'difficult' speaker.