Amplifier Power Questions

[AverageJoe]

I guess I've always accepted the oft-mentioned statement "You need a more powerful amp to get these speakers to really open up", when looking at some of the high-end loudspeakers from Aerial, Salk, and Thiel. The companies themselves usually recommend pretty high Power Requirements in their specs: 100 to 600 watts, 50 to 350 watts, etc.

But when a salesman claims that "...buying these speakers is a waste of money unless you also get the matching Krell amp", I gotta wonder why (since this is the same guy that says I also need the "matching" $1200 cables)?

Crown has an interesting little calculator on their site (www.crownaudio.com/apps_htm/designtools/elect-pwr-req.htm)
for determining Amplifier Power Requirements when plugging in the following variables:

Listener Distance from source
Desired SPL Level at Listener Distance
Loudspeaker Sensitivity Rating
Amplifier Headroom

So... If I plug in my figures - about 15' (4.5M) distance, 85dB, sensitivity rating of 92dB (Pinnacle's), and 6dB headroom (out-of-a-hat figure), I get -
Required Amplifier Power = 16 Watts

If I change the sensitivity figure to 86dB for the Aerial 7B, I get -
Required Amplifier Power = 64 Watts

OK, this supports the "more power needed" theory (especially if the Distance and SPL numbers are increased - if I went to 18' I'd need 96 watts with the Aerials), but the Salk Veracity and Thiel CS3.7's have higher sensitivity figures, but still recommend more power (Thiel - 90dB - recommends 100 to 600 watts).

Dumb questions time (sorry it took so long) :

1. Are there other variables involved that the Crown formula doesn't address? It looks like Sensitivity is the only speaker variable driving the formula. Am I over-simplifying the power-to-SPL requirements of the Aerials? (For example, might power requirement differentials exist between the drivers...that is, do bass/larger drivers require exponentially more power/headroom than the tweeters to match spls?)

2. If I turn my 130 wpc receiver up enough to get 85dB out of the Aerials @ 15', is that the same as turning the 250 wpc Krell down to get 85dB? Obviously the Krell has a lot more headroom, but do I need it in my 15' room?

3. If 85dB in a 15' room (or 64 watts) is not enough to get the Aerials to perform properly ("open up"), do I need a bigger room / turn them up? How ‘bout the more sensitive Thiels? - I’ve heard the same comments regarding them requiring more power to perform well, but if 25 watts will get me 85dB, won’t my 130 wpc receiver still have plenty of headroom?

I've seen parts of this answered elsewhere, but sometimes I didn't completely understand the concepts - for example: I didn't really get what is meant when a speaker sounds "strained" due to lack of headroom. The fault wasn't in the reply, but in my own lack of experience. So, thanks in advance for any help in clearing my cobwebs.

 

[Pyrrho]

First of all, let me say that I love your quote.


To directly answer your questions:

1) Yes. They have omitted impedance. You need the amplifier to deal with the impedance of the speaker. And it needs to deal with the minimum impedance of the speaker, not just the "nominal" impedance.

2) Extra power capability that is never used is useless. What you need depends upon how loud you want it all to play, as well as the characteristics of the speaker and the interaction with the room. And you are right, the power output will depend upon how high up the volume control is turned (as well as other things, like input signal level). The ratings for amplifiers are about claimed ability; this does not mean it is putting out that much power all the time, or even ever.

3) I don't like the way your questions are formed, and therefore will not directly answer them.

Basically, you need an amplifier that can handle the impedance of the speakers, and will be able to drive them as loud as you ever want them to be driven (of course, the speakers might not be able to play as loud as you want, in which case you need different speakers).

You do not need an expensive amplifier in most cases. Salesmen typically want you to part with your money, and therefore try to sell you expensive things, whether you need them or not. If the speakers do not have terribly low impedance dips, or are not terribly inefficient, a standard "mid-fi" receiver will be adequate unless you are wanting literally deafening levels of sound. I have successfully used a receiver that retails for about $600 with speakers that retail for over $6000. The reason I can do that is that the speakers are of normal impedance and not overly inefficient. More expensive amplification would not give me better sound. (In point of fact, I replaced that receiver for one that retails for about $1600 because I wanted more features. However, it sounds the same.)

Of course, many are seduced by various nonsensical claims, and therefore insist on expensive amplification. But level matched double blind tests show that much of the claims in high end audio are pure BS.

 

[Seth=L]

I love that you started this thread, maybe it will turn some heads. It's nice that Crown came up with a calculator for this. It's worth mentioning that Crown's amplifiers are typically fine with lower impedance loads (they will likely nearly double their output power for it), the impedance factor doesn't fit into the equation, but you can figure it for yourself with ease. People are very surprised or disbelief when you tell them they hardly use over 50 watts total when listening to music even at higher levels. You hear a lot of audio noobs talking about how many watts their stereos have "MINES GOT 800 WATTS MAN!!!" but that means next to nothing.

Stable performing amplifiers with lots of headroom are ideal.

 

[AverageJoe]

Thanks Pyrrho, for the reply.

3) I don't like the way your questions are formed, and therefore will not directly answer them.

Basically, you need an amplifier that can handle the impedance of the speakers, and will be able to drive them as loud as you ever want them to be driven (of course, the speakers might not be able to play as loud as you want, in which case you need different speakers).

Fair enough. Actually, #3 was sorta' a loaded question designed to confirm (or not) an assumption that if my amplifier volume is turned up about 1/4 to get the Aerials to a level at which I like to listen, a more powerfull amp with volume turned to 1/8 to get the same SPL's hasn't gained me anything in quality , all other things being equal.

I'm not really sure if that statement is any better...

 

[PENG]

I think Seth is right, Crown amplifiers typically can handle low impedance speakers, so may be that's why they do not factor impedance into their calculations because it does not matter much for them.

 

[AverageJoe]

...People are very surprised or disbelief when you tell them they hardly use over 50 watts total when listening to music even at higher levels...

Yeah, I was surprised at the result I got when pluging in the figures to get 85dB from the Pinnacles - 16 watts. And with some gain from the room and interaction from multiple speakers, I guess it's even a bit better than that.

Sure changes fast with less sensitive speakers, more SPLs, and bigger rooms, though.

 

[fredk]

Yep, thats a great calculator. It really opens your eyes to the range of power requirements a speaker might have.

Another thing to consider is the dynamic range of the music you listen to. Jazz and Classical recordings tend to have greater dynamic range with 25db above average not being unusual for a good classical recording.

There is another Crown page that explains power requirements quite nicely.

You can also add around 5db to your headroom for room acoustics.

So, if you listen to pop quietly on very efficient speakers, you may never need more than say 50wpc. On the other hand, if you like to listen to classical at concert hall levels (100db spl) on less efficient speakers, you could be a candidate for 300wpc.

Me? I learned that I will probably never need more than 100wpc for my listening.

Fred

 

[PENG]

Yeah, I was surprised at the result I got when pluging in the figures to get 85dB from the Pinnacles - 16 watts. And with some gain from the room and interaction from multiple speakers, I guess it's even a bit better than that.

Sure changes fast with less sensitive speakers, more SPLs, and bigger rooms, though.

If you only need 16W, get this one.
http://www.onahighernote.com/luxman/?c=8&id=31

You get 20 WPC, will double down, no question about its SQ potential. I read a review on this thing, the reviewer did not feel wanting more power.

 

[PENG]

Another thing to consider is the dynamic range of the music you listen to. Jazz and Classical recordings tend to have greater dynamic range with 25db above average not being unusual for a good classical recording.

Fred

I mostly listen to classical music. It is true that they tend to have greater dynamic range but I have not experienced anything like 25 dB above average unless you are comparing the peaks to the quietest passages that are always silent.

 

[Seth=L]

Sure changes fast with less sensitive speakers, more SPLs, and bigger rooms, though.

Woops, another factor they missed. They missed the room size itself. The distance from your ears is important, but not the whole picture. Imagine what a concert speakers would sound like in your home at the same distance from your ears as they where in the front row at the concert. I say imagine because if you tried it, you would be deaf.

I get a chuckle from lines like this...

Linky

Even the power cord is a cut above the normal give-away and actually sounds good. No wonder Luxman achieves a superior sound.

 

[Pyrrho]

I love that you started this thread, maybe it will turn some heads. It's nice that Crown came up with a calculator for this. It's worth mentioning that Crown's amplifiers are typically fine with lower impedance loads (they will likely nearly double their output power for it), the impedance factor doesn't fit into the equation, but you can figure it for yourself with ease. People are very surprised or disbelief when you tell them they hardly use over 50 watts total when listening to music even at higher levels. You hear a lot of audio noobs talking about how many watts their stereos have "MINES GOT 800 WATTS MAN!!!" but that means next to nothing.

Stable performing amplifiers with lots of headroom are ideal.

Indeed, with Crown amplifiers, typically they deal with low impedances quite well. That means that their calculator is useful for selecting one of their amplifiers, which is, of course, what they intend with it. But if one is going to use it for general information, then it is lacking with respect to the impedance issue. Since the original question was not simply, "which Crown amplifier should I buy", I thought it worth mentioning that the calculator is not entirely satisfactory for selecting a power amplifier of just any brand.

 

[Pyrrho]

Thanks Pyrrho, for the reply.



Fair enough. Actually, #3 was sorta' a loaded question designed to confirm (or not) an assumption that if my amplifier volume is turned up about 1/4 to get the Aerials to a level at which I like to listen, a more powerfull amp with volume turned to 1/8 to get the same SPL's hasn't gained me anything in quality , all other things being equal.

I'm not really sure if that statement is any better...

The problem with what you are saying is that volume controls are not all the same. But given what I think you mean, you are correct. Using less of the capability of a more powerful amplifier will have the same end result.

More power than you actually use will not give you better quality sound. It is only what you actually use that counts. Of course, having more power capability than needed does no harm, but it really does not help unless you actually use it.

If we consider two amplifiers, one more powerful than the other, let us say one can put out 100 w rms, and the other 200 w rms, but otherwise, the two are exactly alike, the only time it will make any audible difference is when exceeding 100 w.

 

[mtrycrafts]

Yeah, I was surprised at the result I got when pluging in the figures to get 85dB from the Pinnacles - 16 watts. And with some gain from the room and interaction from multiple speakers, I guess it's even a bit better than that.

Sure changes fast with less sensitive speakers, more SPLs, and bigger rooms, though.

Bigger rooms doesn't matter unless your listening distance changes also; that is why that measurement is needed. Most of the sound is the direct sound wave, hence the distance you sit at.

 

[Tomorrow]

Thanks for the perfect thread to show off my ample ignorance, AJ.

I think I understand that as input frequency changes, speaker impedence values change. I would guess that in a typical 3-way system, it is the individual drivers that determine those values. My question then to Pyrrho, Mtry and others, is how is it that speakers maintain the balance of low-mid-high frequency drivers' relative spl levels? In other words, how does the speaker 'sound' or presented (high/mid/low) frequency 'balance' remain consistent at different amplitudes?

Also, how can one determine (from the limited or inaccurate published specs) whether a mid-level receiver can 'adequately' drive a speaker spec'ed in the neighborhood of the Aerial 7B which sports relatively low minimum impedence?

Thanks for the wisdom, guys.

 

[TLS Guy]

Thanks for the perfect thread to show off my ample ignorance, AJ.

I think I understand that as input frequency changes, speaker impedence values change. I would guess that in a typical 3-way system, it is the individual drivers that determine those values. My question then to Pyrrho, Mtry and others, is how is it that speakers maintain the balance of low-mid-high frequency drivers' relative spl levels? In other words, how does the speaker 'sound' or presented (high/mid/low) frequency 'balance' remain consistent at different amplitudes?

Also, how can one determine (from the limited or inaccurate published specs) whether a mid-level receiver can 'adequately' drive a speaker spec'ed in the neighborhood of the Aerial 7B which sports relatively low minimum impedence?

Thanks for the wisdom, guys.

The impedance at any frequency is the sum of resistive, and reactive loads, both inductive and capacitative.

Now a speaker voice coil is an inductor, so impedance will rise with frequency. However if voltage remains constant and the device, speaker, is linear spl will be constant with frequency, although power consumed with rising frequency will be less. Now the crossover components especially the inductors and capacitors change the situation drastically.

Now if inductive elements are dominant voltage will lead current. In a capacitative dominant circuit current will lead voltage. So we have a complex situation where the current draw demanded of the amp at a given frequency can not be known from the impedance alone at that frequency. You have to know the phase angle at the frequency also. Now most speaker manufacturers do not publish an impedance curve and even less often a phase response curve as well.

Now this conundrum creates a vector triangle, in which the hypotenuse is the apparent power in VA (S), the base is the true power in watts (P), and the height the reactive power in VAR (Q). Now the angle between the base and the hypotenuse is the phase angle. Now since in these circuits, any speaker, the current and voltage are not in phase, there are complex situations were the current can be much higher than you would imagine. That is why some speakers are difficult loads, in fact most. You can't even assume a speaker will not require a high current draw at a frequency at which its impedance does not seem excessively low. You have to take into account the vector triangle, and that's hard to visualize. It gets complicated, and that is why I have glossed over the issue in previous posts and have been called to task by some members for doing so.

Anyhow the apparent power in VA (S), is easily measured by measuring voltage and current at a given frequency. The true power in watts (P) is the S X cosine of the phase angle.

If you are now more confused than ever, that's why I have avoided this issue of vectors in posts before. But since you ask, the manufacturers specs are not adequate for knowing the actual power draws in relationship to frequency, still less to determine the actual current requirements demanded of the amplifier.

So you can be lucky or not, but generally I advise higher priced amps able to provide high current when required so as to avoid disappointment.

 

[fmw]

To put things into perspective, the 90db sensitivity rating of the speakers means that the speakers will produce a sound level of 90db at a 1 meter listening distance with 1 watt of input power. That's 1 watt, not 100 watts. 90 db is loud. Each 3 db increase in sound pressure level requires a doubling of amplifier power. So 93 db would need 2 watts, 96 db would need 4 watts, 99 db (REALLY loud) would eat up 8 watts of input power. Moving away from the speakers to a normal listening distance certainly reduces the SPL heard but not a lot. You can check out the Crown amp recommendations to see how that varies. Headroom refers to the amount of power needed to handle loud peaks in the music. It is possible for the peaks to require 10 times the power but that is rarely seen (check out Tchaikovsky's 1812 overture for one that might for the few seconds that the cannons fire.) In movie sound tracks, the loud peaks are normally LFE and are handled by a powered subwoofer. In pop and rock music, there really aren't any meaningful peaks at all.

Now let's look at the difference between a 100 wpc amp and a 200 wpc amp. The difference is 3 db of sound pressure level. Yes, that is audible but it isn't exactly overwhelming.

In some cases it makes sense to get a strong amp in order to handle very low impedance speakers but rarely does it make any sense to make speakers "open up" which is about as nonsensical as anything you will read in the high end audio world.

 

[Tomorrow]

The impedance at any frequency is the sum of resistive, and reactive loads, both inductive and capacitative.

Now a speaker voice coil is an inductor, so impedance will rise with frequency. However if voltage remains constant and the device, speaker, is linear spl will be constant with frequency, although power consumed with rising frequency will be less. Now the crossover components especially the inductors and capacitors change the situation drastically.

Now if inductive elements are dominant voltage will lead current. In a capacitative dominant circuit current will lead voltage. So we have a complex situation where the current draw demanded of the amp at a given frequency can not be known from the impedance alone at that frequency. You have to know the phase angle at the frequency also. Now most speaker manufacturers do not publish an impedance curve and even less often a phase response curve as well.

Now this conundrum creates a vector triangle, in which the hypotenuse is the apparent power in VA (S), the base is the true power in watts (P), and the height the reactive power in VAR (Q). Now the angle between the base and the hypotenuse is the phase angle. Now since in these circuits, any speaker, the current and voltage are not in phase, there are complex situations were the current can be much higher than you would imagine. That is why some speakers are difficult loads, in fact most. You can't even assume a speaker will not require a high current draw at a frequency at which its impedance does not seem excessively low. You have to take into account the vector triangle, and that's hard to visualize. It gets complicated, and that is why I have glossed over the issue in previous posts and have been called to task by some members for doing so.

Anyhow the apparent power in VA (S), is easily measured by measuring voltage and current at a given frequency. The true power in watts (P) is the S X cosine of the phase angle.

If you are now more confused than ever, that's why I have avoided this issue of vectors in posts before. But since you ask, the manufacturers specs are not adequate for knowing the actual power draws in relationship to frequency, still less to determine the actual current requirements demanded of the amplifier.

So you can be lucky or not, but generally I advise higher priced amps able to provide high current when required so as to avoid disappointment.

(my non-technical eyes glazing over, head hurting )...Many thanks, TLS Guy. I'm gonna read this 10 or 12 times and try to understand. I can see why you've not wanted to tackle this before.

 

[TLS Guy]

To put things into perspective, the 90db sensitivity rating of the speakers means that the speakers will produce a sound level of 90db at a 1 meter listening distance with 1 watt of input power. That's 1 watt, not 100 watts. 90 db is loud. Each 3 db increase in sound pressure level requires a doubling of amplifier power. So 93 db would need 2 watts, 96 db would need 4 watts, 99 db (REALLY loud) would eat up 8 watts of input power. Moving away from the speakers to a normal listening distance certainly reduces the SPL heard but not a lot. You can check out the Crown amp recommendations to see how that varies. Headroom refers to the amount of power needed to handle loud peaks in the music. It is possible for the peaks to require 10 times the power but that is rarely seen (check out Tchaikovsky's 1812 overture for one that might for the few seconds that the cannons fire.) In movie sound tracks, the loud peaks are normally LFE and are handled by a powered subwoofer. In pop and rock music, there really aren't any meaningful peaks at all.

Now let's look at the difference between a 100 wpc amp and a 200 wpc amp. The difference is 3 db of sound pressure level. Yes, that is audible but it isn't exactly overwhelming.

In some cases it makes sense to get a strong amp in order to handle very low impedance speakers but rarely does it make any sense to make speakers "open up" which is about as nonsensical as anything you will read in the high end audio world.

I agree with all of that. However speakers are often not presenting the load to the amp you think they are. In order not to have disappointment your best rough guide is to select an amp that will double it power when driving a 4 ohm load as compared to the power at 8 ohms. That is your best guide that the amp has adequate current reserve. The problem is in the issue of combined resistive and reactive loads, all moving coil speakers. We have this concept of apparent power and true power. The current required for the apparent power is real. The true power is less because some of the energy is stored and given back. I know this is a difficult issue and have avoided it before. But having doodled with loudspeakers for so long, I know current reserve is important to consider in selecting amplifiers.

 

[TLS Guy]

(my non-technical eyes glazing over, head hurting )...Many thanks, TLS Guy. I'm gonna read this 10 or 12 times and try to understand. I can see why you've not wanted to tackle this before.

You have my sympathy, but you did ask the question!

 

[AverageJoe]

The impedance at any frequency is the sum of resistive, and reactive loads, both inductive and capacitative.

Now a speaker voice coil is an inductor, so impedance will rise with frequency. However if voltage remains constant and the device, speaker, is linear spl will be constant with frequency, although power consumed with rising frequency will be less. Now the crossover components especially the inductors and capacitors change the situation drastically.

Now if inductive elements are dominant voltage will lead current. In a capacitative dominant circuit current will lead voltage. So we have a complex situation where the current draw demanded of the amp at a given frequency can not be known from the impedance alone at that frequency. You have to know the phase angle at the frequency also. Now most speaker manufacturers do not publish an impedance curve and even less often a phase response curve as well.

Now this conundrum creates a vector triangle, in which the hypotenuse is the apparent power in VA (S), the base is the true power in watts (P), and the height the reactive power in VAR (Q). Now the angle between the base and the hypotenuse is the phase angle. Now since in these circuits, any speaker, the current and voltage are not in phase, there are complex situations were the current can be much higher than you would imagine. That is why some speakers are difficult loads, in fact most. You can't even assume a speaker will not require a high current draw at a frequency at which its impedance does not seem excessively low. You have to take into account the vector triangle, and that's hard to visualize. It gets complicated, and that is why I have glossed over the issue in previous posts and have been called to task by some members for doing so.

Anyhow the apparent power in VA (S), is easily measured by measuring voltage and current at a given frequency. The true power in watts (P) is the S X cosine of the phase angle.

This reminds me of my Director of Engineering years ago. I was bench-sweeping broadband RF amps for installation in the field, and he was calculating a response fix and explaining the formulas and circuit changes as he went, until I did my Dr. McCoy: "I'm a Tech, not an engineer!" (It got to be a running joke with us - Ohms law is limited to a few statutes with me). So he took a grease pencil and marked on the scope where the response would improve. That I understood.

Anyway, thanks for the "best rough guide" comment you posted later.