Why is it that low frequencies consume more amplifier power then high frequencies?

[yepimonfire]

always wondered why this is. anybody know?

 

[mtrycrafts]

always wondered why this is. anybody know?

I bet this is part of it
http://www.webervst.com/fm.htm

To perceive low frequencies equal to the mid bands level needs a lot more dB spl and that needs a lot of power. Then, perhaps, the come mass and area to excite to achieve that level or loudness is also part of it.
A 10 dB spl change is a 10 fold power increase. a 30 dB spl change is a 1000 fold power increase.

 

[TLS Guy]

always wondered why this is. anybody know?

They don't. At least not unless you are used to unnatural bass heavy sound. The acoustic energy is actually in the midrange, especially that range form 100 Hz to 2 kHz.

I find that is where most speakers are sadly lacking in power handling.

There is the issue of efficiency also. Small subs are much less efficient than the other speakers, which adds to the myth.

The next issue if step loss of narrow cabinets, that requires a boost of 6 db per octave below 600 Hz for most speakers.

However the majority of the net acoustic power that has to be radiated into the room, is not below 100 Hz.

So it is a myth that a sub significantly unburdens a receiver.

Given a large efficient bass enclosure, the power that needs to be devoted to the last two octaves is small, with the exception of Hollywood's effects.

 

[GranteedEV]

If we're talking about subwoofers, it's equalization with a lot of them.

For example, a sealed sub begins to roll of maybe at 40hz. that means by 30hz it will be 6db down. It will lose sensitivity as you go down in frequency.

If you want it to be flat down to 30 hz you need to add "EQ boost". To gain 3db, you need double the power. So you need 2^2 = 4 times as much power to add that 6db of power.

The alternative is a high mass, high excursion low fs type of driver. For example my driver has a modeled F3 of 33hz in the box i've got it in. The tradeoff is that my sensitivity is only around 88db or so... which is rather low considering how much cone area I have.

Now if we're talking about speakers and powering them with an amp, where you need power depends solely on what you're listening to and its dynamic range.

 

[gmichael]

I think that for the same volume, to sustain a long note it may not take more power than a high note. But lower notes do tend to come in a single burst like a drum or an explosion. This creates a need for more headroom.

 

[TLS Guy]

I think that for the same volume, to sustain a long note it may not take more power than a high note. But lower notes do tend to come in a single burst like a drum or an explosion. This creates a need for more headroom.

I don't know a powerful blast from a bank of trombones and trumpets over the top and a Wagner tuba or two thrown in, takes an awful lot of power and peak power won't cut it.

And again everybody thinks an organ makes big bass demands and it does, but is pales in comparison the mid and HF energy, especially the 18 century instruments and their clones.

If you take a huge chorus from a Wagnerian opera with 200 piece orchestra, the midrange power required is enormous and why really good speakers cost so much.

 

[KEW]

Why is it that tweeters generally require less power?

 

[TLS Guy]

Why is it that tweeters generally require less power?

The energy spectrum of music does rapidly fall away after 2 kHz. However a tweeter crossed over below 2.5 kHz does have to handle quite bit of power especially with lower order crossovers.

Some electronic music had a very "artificial" energy spectrum and can spell death to tweeters. It is also easy to destroy tweeters with test tones, and you have to reduce power in the tweeter range.

 

[MinusTheBear]

I was always under the impression lets say for example a 3 ohm impedance dip at 100hz would be more demanding and consume more power from the amplifier than a 3 ohm dip at 5kHz? Say it ain't so?

 

[TLS Guy]

I was always under the impression lets say for example a 3 ohm impedance dip at 100hz would be more demanding and consume more power from the amplifier than a 3 ohm dip at 5kHz? Say it ain't so?

No, you are correct. However 5 kHz is not an area were there are high power demands. A dip at 1000 Hz would not be any different than one at 100 Hz though.

The impedance dips below 600 Hz as a rule so more power can be drawn from the amp for step loss compensation. However the acoustic power radiated by the speaker should not change with frequency for the same driving voltage. So a speaker not requiring step loss compensation, or one not having it when it should, will not draw this added power.

With ported enclosures there are always two peaks of impedance below 100 Hz which limit power output from the amp. A sealed enclosure will have one peak of impedance.

The other issue is phase angle, which is independent of impedance, but draws current from the amp over and above what the impedance at a given frequency would lead you to expect. Whist this does not increase true power, the current for the apparent power has to be provided and helps run the amp out of gas and make it clip. The more negative the phase angle at a given frequency the more current will have to be provided to meet the apparent power demands. It is this latter problem that can make a speaker a much more difficult load then one would expect. It is this effect more than any other that causes different amps to behave differently when driving a given set of speakers and cause definite changes in sound quality.

So not all amps will sound the same into some loads. This confounds the wisdom of bench testing amps and receivers into resistive loads.

A speaker is a much more complex load than a resistor, and and causes bench testing to have much less applicability to real world conditions than most people assume.

 

[GO-NAD!]

They don't. At least not unless you are used to unnatural bass heavy sound. The acoustic energy is actually in the midrange, especially that range form 100 Hz to 2 kHz.

I find that is where most speakers are sadly lacking in power handling.

There is the issue of efficiency also. Small subs are much less efficient than the other speakers, which adds to the myth.

The next issue if step loss of narrow cabinets, that requires a boost of 6 db per octave below 600 Hz for most speakers.

However the majority of the net acoustic power that has to be radiated into the room, is not below 100 Hz.

So it is a myth that a sub significantly unburdens a receiver.

Given a large efficient bass enclosure, the power that needs to be devoted to the last two octaves is small, with the exception of Hollywood's effects.

I would phrase the question this way. Take a hypothetical mid/bass driver installed in a speaker and play a 1000 Hz tone through it at a given spl, then play a 200 Hz tone at the same spl. Would we expect a difference in the power consumed for each frequency? Or, is that question too simplistic? That is to say, would the expected change in impedance and efficiency with the change in frequency render the comparison invalid?

In other words, does it depend on the driver in question, as to which frequencies are more demanding of power? I realize that most program material will have greater power demands in the middle frequencies. But, from a purely physics/electrical standpoint, can we say that there is any difference in power consumption with a change of frequency, assuming acoustic power remains constant?

Hope my question makes sense.

 

[TLS Guy]

I would phrase the question this way. Take a hypothetical mid/bass driver installed in a speaker and play a 1000 Hz tone through it at a given spl, then play a 200 Hz tone at the same spl. Would we expect a difference in the power consumed for each frequency? Or, is that question too simplistic? That is to say, would the expected change in impedance and efficiency with the change in frequency render the comparison invalid?

In other words, does it depend on the driver in question, as to which frequencies are more demanding of power? I realize that most program material will have greater power demands in the middle frequencies. But, from a purely physics/electrical standpoint, can we say that there is any difference in power consumption with a change of frequency, assuming acoustic power remains constant?

Hope my question makes sense.

It depends on the speaker.

If the speaker had the same impedance and phase angles and no response irregularities then the power would be the same.

However lets take a narrow fronted speaker with diffraction loss starting at at 600 Hz. Then the power required for a 300 Hz tone would be doubled and double again at 150 Hz, for the same spl. Note that this not only stresses the amps, but the drivers as well. This is probably the biggest argument in favor of a big integrated speaker design. You can't accomplish step loss compensation with a sub.

 

[GO-NAD!]

Hmmm, I wonder why there isn't a market for speakers with no, or at least reduced, BSC. That way, wouldn't they be easier to position in a typical room, i.e. closer to front walls? I would think such a speaker would be far more sensitive/efficient without (or reduced) BSC.

Or, would such designs be a disaster, from a marketing standpoint?

 

[GranteedEV]

Hmmm, I wonder why there isn't a market for speakers with no, or at least reduced, BSC. That way, wouldn't they be easier to position in a typical room, i.e. closer to front walls? I would think such a speaker would be far more sensitive/efficient without (or reduced) BSC.

Or, would such designs be a disaster, from a marketing standpoint?

Most speakers DO have reduced BSC as it is.
Away from walls they lose some bass as well as lower midrange!
In-walls have no BSC. 2.5 ways can have full BSC (6db) without losing sensitivity as the second driver operates only in the BSC region.

A better solution is "changable" BSC. With passive speakers I've seen it done by guys like Kevin Voecks and Vance Dickason as well as on some professional studio monitors but ultimately the best way to do it (in MY opinion at least) is digitally active, either with in a x.5 way by simply changing the trim for that .5 driver, or with a digital filter to boost those frequencies as necessary.

Either way, total power being radiated by the speaker is the same at 200w and at 1khz. The only differences is their direction, since some goes backwards below the baffle step you only need more to go forward towards YOU.

 

[yepimonfire]

sorry, but what exactly is bsc?

 

[TLS Guy]

Hmmm, I wonder why there isn't a market for speakers with no, or at least reduced, BSC. That way, wouldn't they be easier to position in a typical room, i.e. closer to front walls? I would think such a speaker would be far more sensitive/efficient without (or reduced) BSC.

Or, would such designs be a disaster, from a marketing standpoint?

You guys are asking all the tough questions!

Most speakers in fact have no, or reduced BSC. The whole problem with BSC is that it is so speaker position dependent. Of course an in wall speaker needs no BSC.

The issue is controversial. Most designers feel speakers should have as narrower front as possible and be placed 10 to 14 inches from room boundaries. Some like Ted Jordan vehemently disagree. I would have to admit he is in the minority. He is an advocate of large front baffles. I won't say that his approach is without merit, but I know many who would be all over me for that remark. However we continue to get posts and one recent one about preferring old 70s and eighties speakers with 10, 12 and 15 inch drivers and not being satisfied with modern offerings.

This narrow front approach does result in reduction of output in the tenor range hence BSC.

The reason for the narrow front and being away from boundaries is reflection reduction, and consequently reduction in peaks and nulls from interference by reflected waves.

The whole confounding issue is rooms and room gain. One level of BSC will be right for only one room and one location. For other locations it will be too much or too little.

So I have moved to making BSC active and not passive. Then I have continuous and infinite control of it.

Prior to that my passive designs had switchable variation in BSC. However you can not make passives continuously variable, but have to be content with a number of switchable curves.

This approach adds a lot of expense to the speaker if done properly. You have to have switchable inductors. Actually what I do is add them in series as BSC is reduced. Now they are in the signal path and of high value. So to do it properly you need large value air core inductors with large gauge wire.

This is the system I use in my surround speakers, which used to be my location monitors when I was doing live recordings. They had to be able to work in different environments continuously.

One thing I will say against these narrow cabinets requiring BSC is that in addition to stressing amps and receivers, it stresses the drivers even more. It greatly adds to distortion and thermal compression issues. That is why I gave the duty of BSC compensation to the upper 10 inch driver rather than the two 7 inch drivers, in my mains via an active controllable design. This contributes greatly to the glorious effortless sound of these speakers, noted by Jake in his review.

Now I can do what I want. I don't have to market and sell a speaker. That's another reason I'm glad I roll my own and don't have to go speaker shopping.

I would say this issue is just one more reason to advocate to move away from passive speakers and go active. Dealing with this issue and having appropriate control over it becomes duck soup.

One thing I don't like is a narrow fronted speaker with inadequate BSC compensation. From my forays into dealers showrooms, that seems to encompass a lot of speakers. It results in atrocious reproduction of the violin, cellos, the tenor, contralto and mezzo voice for starters. Unfortunately there are many confounding circumstances to getting it right, and yet it involves a portion of the frequency spectrum where the ear is highly sensitive to error.

 

[TLS Guy]

sorry, but what exactly is bsc?

This will explain it, so I don't have to again.

 

[Swerd]

sorry, but what exactly is bsc?

Read these two links

BSC made simple by Jim Salk

BSC by Martin King

 

[Swerd]

… One thing I don't like is a narrow fronted speaker with inadequate BSC compensation. From my forays into dealers showrooms, that seems to encompass a lot of speakers. It results in atrocious reproduction of the violin, cellos, the tenor, contralto and mezzo voice for starters. Unfortunately there are many confounding circumstances to getting it right, and yet it involves a portion of the frequency spectrum where the ear is highly sensitive to error.

Agreed!!

Martin King has an article describing a variable BSC circuit. I've never seen it implemented. What do you think of it?

 

[TLS Guy]

Agreed!!

Martin King has an article describing a variable BSC circuit. I've never seen it implemented. What do you think of it?

That's fine for a highly efficient driver like the Lowther, that will deafen you with less than a watt of power.

However for higher powered speakers, the variable resistor will soon go up in smoke at the pint of contact. Been there and done that.

So my approach is to vary the turnover frequency of the correction.