Two Basic Amplifier Questions

[Sokel]

Looked at the link but missed the part where the type of load was specified.

Oh,you mean the actual test load?
I suppose it's resistive for the 4R measurement and none (?) for the open load,as the labels under the graph's state.

(I'm completely unfamiliar with this measurement,I wanted to ask anyway)

 

[Bob from Florida]

Oh,you mean the actual test load?
I suppose it's resistive for the 4R measurement and none (?) for the open load,as the labels under the graph's state.

(I'm completely unfamiliar with this measurement,I wanted to ask anyway)

Probably resistive then. If they were using a real world load they would point that out for marketing.

 

[Sokel]

Probably resistive then. If they were using a real world load they would point that out for marketing.

The graphs of the article is also put sloppy,what I keep from it (and hope it's pin compatible,at least for supply voltage) is that 6.6Khz THD+N vs power which is really low for class D.
Let's see how it will do in the third-party tests when it comes out.

 

[fpitas]

The clipping question is neatly avoided if the amp is powerful enough to deafen you near full output. Mission accomplished!

 

[Steven Holt]

Yes, if you start cranking in serious EQ at bass, the required power will soar. You may also burn out your woofers

ALL YOU HOME THEATRE GUYS THAT TREAT EQUALISATION AS A G*DD*M FETISH ARE YOU LISTENING??

 

[fpitas]

ALL YOU HOME THEATRE GUYS THAT TREAT EQUALISATION AS A G*DD*M FETISH ARE YOU LISTENING??

They'll just buy kilowatt amps.

 

[radix]

So, for a real world question: what kind of Speakers are needed to handle this sort of power? A mid-range example like Kef R3 has a recommended max power of 180W. The Kef Information Sheet does not qualify this so I don't know if that's peak (surely not) or only for a frequency and impedance range. Whatever, it's lower than 500W and nowhere near 2kW

If you look at Erin's review, the R3 meta starts compressing a lot by 102 dB SPL under 200 Hz. And the 96 dB SPL distortion below 200 Hz gets pretty bad.

KEF R3 Meta Bookshelf Speaker Review

Kef R3 Meta

www.erinsaudiocorner.com

For 87 dB sensitivity speakers, 96 dB @ 1m is only 8 watts. 102 dB is only 32 watts.

 

[HarmonicTHD]

And if there is enough wattage to cover peaks? Will adding more wattage enhance sound quality?

What I'm getting at, is that I was under the assumption that speakers are 'happy' and work best with certain wattages.
I'm now wondering if that is true, if sound quality and wattage is related. Again, assuming enough power to meet required db levels.

No.

In some simpler words.... wattage relates to "loudness", the louder you want it the more wattage is needed. If you reached the loudest level you feel comfortable with additional unused wattage does not do anything. The wattage rating of a speaker does only tell you from what wattage on up you will damage it (it does not tell you anything about "loudness" nor anything about how strong your amp should be). For this the "speaker sensitivity" is used - often around ca. 86db/1m/W (look up the exact value for your speaker). You can calculate the "loudness" i.e. SPL using for example this:

Peak SPL Calculator

Here is a not uncommon example and you see that with "just" 30 watts you can reach peaks of 96dB SPL, which is pretty loud already and is already a level where many "common" speakers start distorting. If you now have speakers which dip in impedence this wattage might e.g. double. So conservatively in this example you are usually well set with a 60W amp. However, you will also see that the amount of power required rises exponentially, in case you want it louder (and provided your speakers dont distort).

I have simplified it all a bit so you can hopefully understand better, given the knowledge you state you have.

 

[pma]

often around ca. 86db/1m/W

Definition with respect to 1W was abandoned because speaker impedance is not constant and depends on frequency. The correct way used now would be 86dB/1m/2.83V. It makes 1W at 8ohm, but 2W at 4ohm. Actual speaker impedance is neither 4ohm, nor 8ohm, it is a frequency dependent function.

 

[radix]

No.

In some simpler words.... wattage relates to "loudness", the louder you want it the more wattage is needed. If you reached the loudest level you feel comfortable with additional unused wattage does not do anything. The wattage rating of a speaker does only tell you from what wattage on up you will damage it (it does not tell you anything about "loudness" nor anything about how strong your amp should be). For this the "speaker sensitivity" is used - often around ca. 86db/1m/W (look up the exact value for your speaker). You can calculate the "loudness" i.e. SPL using for example this:

Peak SPL Calculator

Here is a not uncommon example and you see that with "just" 30 watts you can reach peaks of 96dB SPL, which is pretty loud already and is already a level where many "common" speakers start distorting. If you now have speakers which dip in impedence this wattage might e.g. double. So conservatively in this example you are usually well set with a 60W amp. However, you will also see that the amount of power required rises exponentially, in case you want it louder (and provided your speakers dont distort).

I have simplified it all a bit so you can hopefully understand better, given the knowledge you state you have.

That is a great example. The only thing I would add is that when one listens to music or measures the SPL at a location, it's a time average. There will be peaks above the "average" level. This is usually put into what's called "headroom." The calculator linked does not account for headroom.

I usually add 6-10 dB of headroom to accommodate classical or some instrumental music that's not been compressed to hell. For other types of music, 3-6 dB might be sufficient.

So, in the above example, that 96 dB would be at peak power, and you might expect your average loudness to be closer to 90-93 dB without clipping the amp.

 

[HarmonicTHD]

Yes. All correct. My intention was to make it a bit simpler for the OP because I got the impression from post 5 onward he might have gotten difficulties to follow.

 

[Henryk]

Yes. All correct. My intention was to make it a bit simpler for the OP because I got the impression from post 5 onward he might have gotten difficulties to follow.

You ain't wrong, my friend!

All that said, I read over everything carefully and I do get the gist!
Even though some of it is hard to follow, it's nice to have more information than less.

 

[Henryk]

Yes, if you start cranking in serious EQ at bass, the required power will soar. You may also burn out your woofers

Very good to know!
So then, if one is using room equalization, you need to pay close attention to your amplification.

 

[fpitas]

Very good to know!
So then, if one is using room equalization, you need to pay close attention to your amplification.

If you boost, definitely. A lot of people don't have much luck boosting to fill in holes in FR, they simply knock down peaks.

 

[Deleted member 21219]

If you boost, definitely. A lot of people don't have much luck boosting to fill in holes in FR, they simply knock down peaks.

Soooooo true! Many years ago, I was told, "Cut first, boost later." For a long time, I didn't understand why. Now perhaps I do ..... maybe ..... I hope.

What follows is a generality. It doesn't hold true in all cases, but it does for most.

Peaks are generally resonances, either in the room or associated with the driver. (Rarely are they constructive interference. The one instance where they are is between drivers at angles divergent from the best response axis. The cure in that case is simple; find the best response axis.) Resonances are frequency bands of too great a level of energy, so you can "cure" them by reducing the energy.

Nulls in the frequency response, OTOH, are usually areas of destructive interference. These areas occur all throughout the frequency band, but as frequency rises, the nulls become very narrow, and thus difficult for our hearing to distinguish. The bass portion of the spectrum, however, is different. The waves are longer, which means that we can hear (and detect) the null more easily.

But because those waves are longer, they dominate the room's dimensions more easily. Moving the location of the speakers or moving the location of the listener has progressively less and less noticeable effect as frequency becomes lower and lower. (This is why speaker bass response is clearer outside than inside; fewer surfaces for interference.)

Destructive interference is very interesting. Because it is defined as the subtractive interaction between waves, the level doesn't really make all that much difference. So you can have a null that you notice at one watt output, and it will still be there at ten watts output ..... and twenty ..... and a hundred ..... and a thousand.

Notice that I'm not saying that all nulls are caused by destructive interference. But if they are, you're (almost) screwed.

The reason I say "almost" is that by adding energy to the room in different locations (with resultant slight differences in phase), you can address the issue productively. The only system of which I'm aware that does this is the one by @Duke, called "The Swarm". I have heard that there are others, but his was the only one that I knew of.

Jim

 

[Henryk]

Soooooo true! Many years ago, I was told, "Cut first, boost later." For a long time, I didn't understand why. Now perhaps I do ..... maybe ..... I hope.

What follows is a generality. It doesn't hold true in all cases, but it does for most.

Peaks are generally resonances, either in the room or associated with the driver. (Rarely are they constructive interference. The one instance where they are is between drivers at angles divergent from the best response axis. The cure in that case is simple; find the best response axis.) Resonances are frequency bands of too great a level of energy, so you can "cure" them by reducing the energy.

Nulls in the frequency response, OTOH, are usually areas of destructive interference. These areas occur all throughout the frequency band, but as frequency rises, the nulls become very narrow, and thus difficult for our hearing to distinguish. The bass portion of the spectrum, however, is different. The waves are longer, which means that we can hear (and detect) the null more easily.

But because those waves are longer, they dominate the room's dimensions more easily. Moving the location of the speakers or moving the location of the listener has progressively less and less noticeable effect as frequency becomes lower and lower. (This is why speaker bass response is clearer outside than inside; fewer surfaces for interference.)

Destructive interference is very interesting. Because it is defined as the subtractive interaction between waves, the level doesn't really make all that much difference. So you can have a null that you notice at one watt output, and it will still be there at ten watts output ..... and twenty ..... and a hundred ..... and a thousand.

Notice that I'm not saying that all nulls are caused by destructive interference. But if they are, you're (almost) screwed.

The reason I say "almost" is that by adding energy to the room in different locations (with resultant slight differences in phase), you can address the issue productively. The only system of which I'm aware that does this is the one by @Duke, called "The Swarm". I have heard that there are others, but his was the only one that I knew of.

Jim

That was all incredibly helpful and insightful.
Thank you for taking the time

Also, your signature is extremely true.
Unfortunately, so few people are aware of it.
Good of you to publicize it!

 

[MaxwellsEq]

But since clipping is going to happen at higher volumes anyway, shouldn’t we be just as concerned with how the amp recovers from the clipping episode?

Yes. There are some standards for burst testing.