More than 1 spkr. Why?

[Mike Sulzer]

Why did (guitar amp) manufacturers begin using more than one speaker? Because, in the 1950s and for years thereafter, upping the wattage of the amp was much more expensive than adding a speaker and building a larger cabinet.

The fact that they used cabinets, often open backed, that were just large enough to accommodate the speakers attests to the fact that they weren't worrying or even thinking about the acoustic properties of the speakers and enclosure.

Adding another identical speaker does not increase the amount of acoustic power transmitted, although it could redirect it. If you match impedances properly it will be the same; if you screw up the impedance match it will be less.

Using more speakers was mostly so you could use more total power. When 12" speakers could only take 15W you needed a lot of them. 12" speakers that can take high power are usually less efficient than the old low power ones. Thus 60W into one high power 12" is not as loud as 60W into four efficient lower power speakers.

[Leo_Gnardo]

plus I was recommended a C10Q . . . £30.

C10Q a favorite here, and the price is right! Has made some tough customers happy, that's for sure.

Break in. Consider the speaker has a couple of springs. One is the "compliance", where the cone meets the frame. T'other is the "spider" at the cone bottom, that waffly thing that attaches the small end of the cone to the frame. Like any new springs, they're tight when new. Then settle in after some hours of use. Then get flabby and fail to control cone movement when they're knackered, over years of hard use. Some folks exercise new speakers by hooking them up to 6 or 12 volt transformer secondaries and let them flap around for hours. I've never done that yet but it's fun to read about it & I'm sure that method has its aficionados. Just playing, you'll break 'em in over the course of weeks, and you may notice low frequencies more easily produced as the springs loosen up a bit.

Another split window for ya. A pickup AND a VW van style, note company on the side.



Ah, Tucson! I thought I recognized those mountains.

[nickb]

Vance Dikason in his Loudspeaker Design Cookbook claims that doubling the number of speakers quadruples the acoustic power. See highlighted section:

Edit: Here is just that section but bigger:

[Sea Chief]

Vance Dikason in his Loudspeaker Design Cookbook claims that doubling the number of speakers quadruples the acoustic power. See highlighted section:

Edit: Here is just that section but bigger:

[ATTACH=CONFIG]36621[/ATTACH]

[ATTACH=CONFIG]36620[/ATTACH]

Christ those are small- I must be getting old, even my magnifying glass cant decipher the 2nd one's enlarged bit. Are you an ant?

Acoustic power.. lordy that sounds like a physics lesson me no likey.

[Malcolm Irving]

... sounds like a physics lesson ....

Yup. A physics lesson that's gone wrong somewhere.

[Leo_Gnardo]

Careful adding extra speakers Chief ol' mate, it can get to be a habit.

[Mike Sulzer]

For rf, it is called an antenna array. Direct the power into a narrower beam. For audio, it is great if you do not need to spread the power out to fill the room.

Vance Dikason in his Loudspeaker Design Cookbook claims that doubling the number of speakers quadruples the acoustic power. See highlighted section:

Edit: Here is just that section but bigger:

[ATTACH=CONFIG]36621[/ATTACH]

[ATTACH=CONFIG]36620[/ATTACH]

[R.G.]

I suspect that Mr. Dikason may have had a bit of difficulty in the wording. I'm fairly sure he did not mean that doubling the number of speakers doubles the acoustic power **for the same electrical input power**. I can see that he might have meant that it doubles the possible acoustic output power by providing another speaker to share the heat dissipation load. Possibly something else - I'm just guessing.

Here's my reasoning on this:
- Dikason says he's using 0.5% efficient speakers. That's reasonable - speakers are very, very inefficient at converting electricity into air vibrations.

- If you have an N% efficient speaker, and you feed it X watts, you get X*N/100 watts of sound pressure out. That's what an efficiency of N% means.

- If you parallel another speaker, and adjust the amplifier so that it still puts out X watts into the paralleled pair of speakers, each speaker (if they're identical) gets half the power, so each gets X/2 watts. So each one puts out (X/2)*(N/100) watts. The acoustic outputs are summed in the air in front of the speakers, and you get the same X*N/100 output you did with one speaker. This makes sense - the speakers don't know there's another speaker working with them so their efficiency (i.e. watts of sound out divided by watts of electricity in) does not change just by paralleling another speaker.

- If you parallel another speaker and *don't* adjust the amplifier so it puts out the same watts (not the same voltage, the same electrical watts of signal), then paralleling a speaker onto a solid state amp *may* cause the acoustic power to double, as solid state amps produce a constant voltage output, and paralleling another speaker feeds both speakers with the same voltage, and they each then suck down the same current as one speaker would have, so the SPL out doubles, but it does so because the electrical output from the amp has doubled because it's supplying twice the current.

- If you did that last item with a tube amp, the output electrical power may go up or down, depending on how well the new paralleled speaker load as reflected onto the output tube plates matches the peak in output power for those particular power tubes at that power supply voltage and conditions. For output tubes, both under-and over-loading the tubes' plate to plate load causes electrical signal power to fall a bit. If electrical output power falls, it doesn't matter how many speakers are connected to the output, sound pressure level will fall because the speakers don't get more efficient because other speakers are added; X watts out times an efficiency of N% gives the same acoustic power level out.

Mother Nature doesn't care what you build your speakers out of, nor how many you use. The laws of thermodynamics say that if you have a certain electrical power and convert it to acoustic power at some percent efficiency, the acoustic output power remains constant if the total power in doesn't change and the efficiency doesn't change. She doesn't care how many speakers you use. The answer's the same.

You can fiddle the answer by using beaming, to gather the spread-out radiation of one speaker into a smaller volume; that's the antenna array mentioned. Sounds louder in the beam, quieter outside the beam, but the total power radiated remains constant. This happens in the same way that a lens can intercept mild sunlight and concentrate it into a burning spot; the power's the same over the whole area, it's just gathered into a small area, leaving other places darker.

[Malcolm Irving]

In his book, Dikason was claiming that 2 speakers put out 4 times the acoustic power of a single speaker (for the same electrical power input).

Two speakers can put out more acoustic power (per input electrical watt) if the two speakers somehow work together to give higher efficiency.
This could be explained in various ways: Doubling of the cone area gives a more efficient mechanical ‘transformation’ to the air. (Presumably for that explanation the two speakers do not need to be close together). Or, the air pressure loading on each speaker from the other creates a higher acoustic impedance for the cone to work into. (For that explanation the two speakers have to be close together. The distance between the speakers would have to be short, relative to the wavelength of interest.)

We know that exponential horn loading massively increases the efficiency of a driver. That’s supposed to be because the higher acoustic impedance at the narrow end of the horn is a good acoustic impedance match to the driver, while the open end of the horn is a good acoustic impedance match to ‘free air’. (The exponential horn acts as an acoustic ‘transformer’.) However, if it’s just a transformer the electrical input would have to increase to supply the extra acoustic power. To get the extra acoustic power for free it also has to be a lower loss transformer.

It’s obviously pretty complicated and if you take any two explanations as independent factors you could end up with a factor of times 4. I think there may be some ‘double counting’ going on here.

If it is real physics it must be supported by experimental or theoretical evidence. I am not aware of that evidence. (Doesn’t mean there isn’t any, of course.)

[Malcolm Irving]

If it is correct that ‘doubling the cone area’ creates higher efficiency, let’s follow up the consequences:

Assuming that the two speakers are far enough apart for the sound fields not to interact.

Say we have a 12” speaker which converts 1 electrical watt into an acoustic power of 0.01 watts. (The efficiency of the speaker is 1%.)
If the ‘doubling cone area’ theory is correct, two of these speakers with 0.5 watts into each would create a higher total acoustic output of say 0.02 watts. Therefore, considering each speaker individually (which we can do, as they are far enough apart) the efficiency of each speaker is now 2%.

So if this theory is valid, the efficiency of the individual speaker doubles when we reduce the input power from 1 watt down to 0.5 watt. That means that we get the same acoustic output from an individual speaker at 0.5 watt and at 1 watt. The theory is nonsense.

[nickb]

If it is correct that ‘doubling the cone area’ creates higher efficiency, let’s follow up the consequences:

Assuming that the two speakers are far enough apart for the sound fields not to interact.

Say we have a 12” speaker which converts 1 electrical watt into an acoustic power of 0.01 watts. (The efficiency of the speaker is 1%.)
If the ‘doubling cone area’ theory is correct, two of these speakers with 0.5 watts into each would create a higher total acoustic output of say 0.02 watts. Therefore, considering each speaker individually (which we can do, as they are far enough apart) the efficiency of each speaker is now 2%.

So if this theory is valid, the efficiency of the individual speaker doubles when we reduce the input power from 1 watt down to 0.5 watt. That means that we get the same acoustic output from an individual speaker at 0.5 watt and at 1 watt. The theory is nonsense.

It seemed such a basic conservation of energy issue that it couldn't possibly be true. I admit I stopped reading his book when i got to that bit- I lost confidence in anything he had to say. The reality is that the (area-averaged) power out will be reduced very slightly due to losses in the extra speaker.

In this thread, one poster claimed that it doubled the power and another who said it didn't so I thought I would throw in some controversy to stimulate discussion.

It's very worrying that supposed authorities can make such a fundamental boo-boo and have it published. Now, I have to find a decent speaker book

PS: Sea Chief, sorry about the size - it was the best I could do.

[Malcolm Irving]

I think in the design of an individual speaker, if you keep everything else the same (magnet, coil, etc.) but double the cone area, you will get a more efficient speaker (for low frequency anyway). Not sure by how much.

The fallacy then arises because people think ‘Ah, two speakers – that must be like one speaker with a bigger cone area’. It doesn’t necessarily follow.

However, two speakers close together (e.g. in a 2x12 cab) do interact acoustically to produce higher efficiency. It’s because the air pressure from each speaker provides a ‘higher acoustic impedance’ for the other to work into. (Something similar to the higher pressure at the driver end of an exponential horn.)

[Malcolm Irving]

A couple of other factors that I think do come into it for two speakers:

Power compression (as mentioned by pdf64): with prolonged high power use, the voice coil heats up, its resistance increases, making the efficiency worse. Two speakers will heat up their coils less.

Nonlinearity of cone suspension: The voice-coil/cone suspension may become stiffer as the cone excursion gets nearer to its design limit - again reducing efficiency. Two speakers will not get so close to their excursion limit.

[R.G.]

I think in the design of an individual speaker, if you keep everything else the same (magnet, coil, etc.) but double the cone area, you will get a more efficient speaker (for low frequency anyway). Not sure by how much.

You do get a more efficient speaker at frequencies below the one where the centers of the speaker cones are um... one quarter of the wavelength of the sound in air IIRC. But that's because the efficiency of the speaker's coupling to the air is already falling off because the impedance of the cone compared to the impedance of the air it "grips" is mismatched ever more as frequencies falls. This is an extension of bass response, not an increase in general efficiency overall. And it's why you use bigger speakers for bass - the bigger cone has a better "grip" on the air.

A bigger speaker cone makes the speaker's treble efficiency worse because the efficiency of the cone and magnet "motor" driving the cone falls off with higher frequencies. The cone's greater mass simply can't be accelerated as well at treble. The cone stops moving as a piston and starts flexing. The treble comes increasingly from the center of the cone at high frequencies. In the limit, you wind up with only the very center of the speaker radiating. This is why there are "dome tweeters", with only a hard dome over the end of a voice coil. Some MI speakers have aluminum domes over the voice coil, and these have a vastly greater treble response, for better or worse.

However, two speakers close together (e.g. in a 2x12 cab) do interact acoustically to produce higher efficiency. It’s because the air pressure from each speaker provides a ‘higher acoustic impedance’ for the other to work into. (Something similar to the higher pressure at the driver end of an exponential horn.)

It's highly frequency sensitive. That's why the bass response is extended; at least it's another way to say what I said up top. But the "higher acoustic impedance" only exists at frequencies where the spacing of the speakers produces a mutual reinforcement. At frequencies where you get a mutual cancellation, the acoustic impedance is worse locally, and one speaker makes the other suck. The trick here is to get the speakers so close that the reinforcement/cancellations as high in frequency as possible, and that's why speakers in an array are as close as they can be put mechanically. This makes them ever closer to a single speaker with a wide, flat diaphram for frequencies where the effective gap is less than 1/4 wave between the effective radiating area of the cone, which gets smaller as frequency rises. And that makes for peaky response from speaker arrays as frequency rises, and that is why big speaker arrays are always made with crossovers and different sized speakers (apart from the considerations in getting better treble response by using physically smaller speakers, too).

And then there are horns. Horns really are "acoustic impedance transformers". They transform a low acoustic impedance in free air to a high one that the much higher-mass diaphram of the speaker motor at the throat can work better with. That dramatically increases the grip the speaker coil/magnet motor can get on the air and raises that 0.0x% speaker efficiency a lot. Horns are the real winners in speaker efficiency - and they're BIG! I remember reading a DIY home horn speaker article from the 60s where a hifi nutzo made the whole wall of his living room be a horn mouth and wound the tapered tube around the house and into his garage for the speaker to work into.

Speaker efficiency is so very low that even tiny changes make for big changes in SPL. The really tough part is to make that efficiency change be linear across the 1000 to 1 frequency range of human hearing. Practically every kooky theory you can imagine has been tried, and even marketed as the next big thing in speakers. But by and large, we keep coming back to the same solutions.

There must be a reason for coming back, eh?

[Mike Sulzer]

However, two speakers close together (e.g. in a 2x12 cab) do interact acoustically to produce higher efficiency. It’s because the air pressure from each speaker provides a ‘higher acoustic impedance’ for the other to work into. (Something similar to the higher pressure at the driver end of an exponential horn.)

I am having trouble figuring out how this works. How does it happen? Anybody have a reference?