Efficiency of Two Speakers in Series

I am planning on building a nice set of tower speakers for home theater use. My thought was to use two axon 6S1 and a tweater possibly the axon TI1 or maybe a vifa or peerless instead. My dilemma would be that the 6S1's have an impedence of 4 Ohms. No problem I will wire them in series to present a load of 8 Ohms to the amplifier. However I am wondering what that effects will be on the SPL of these speakers. My amp is somewhat low powered so I would like to maximize efficiency.

I have read some of the past debates about this topic but no one has come up with a definitive answer. Some people say this combination will have the same efficiency as the single driver, others say it will increase 3 dB others say it will decrease 3 dB. What are everyones thoughts on this issue?

Thanks
Mark
 
Each one in series has 1/2 the voltage it would have had by itself. That means 1/4 the power. But there are two, so both together produce half the power that they would have had, or 3dB less. Since they are woofers, any interference pattern to affect the results will be less of a factor too. The efficiency is effectively the same, they just draw less power and put out proportunately less.
 
AudioFreak,

If you wire them in series, 1W/1M efficiency should rise by 3dB and 2.83V/1M efficiency should equal that of a single driver.

Maybe I am just confused. When you say 1W/1M efficiency should rise by 3db is that when there is 1W to each speaker or is that one what to the whole box. Because I would think that if there was one what to the whole box then each woofer would only recieve 1/2 W. There for sound would be down in each woofer 3 dB for a total of -6 dB. But since there are two woofers output would increase by 3dB for a total net gain of -3 dB. Am I correct or is there something that I do not understand?

Thanks
Mark
 
hello

anyone that says that there is a gain is wrong (sorry) when you add another speaker that is the same covering the same frequencies what happens is that the output remains the same, some people say that there is twice the surface area when you add another speaker so the sensitivity should increase, true, but you are also adding another motor structure, spiders, and so on. So the sensitivity stays the same.

you may think there are no benifits then. If you add another speaker you will (A) effectivly double the power handling of the speaker you are designing (for that frequenct range) (B) decrease the amount of distorsion, because each speaker only has to work half as much to create the same output and (C) you may be able to present the amplifier with a lower impedance and therefore increase power from the amp.

If you dont belive me, go and try it in any speaker design program, change the number of drivers, keep the input power the same and check out the results
 
Well, things are not so simple - as usual ;-)

At low frequencies, using two drivers results in better efficiency, because of increase in acoustical impedance. This is right as long as the distance between the radiating cones is small in comparison with the wavelenght. The theoretical increase of 3 dB is never reached, ususal values are 1.5 to 2 dB.

At high frequencies (1000 Hz and up) drivers must be considered as two sources, causing mutual interference, and the measured efficiency is depending on the location of the microphone. The efficiency is thus increased in the axis, but the total radiating power is the same as with one driver.


Regards, P.Lacombe
 
ok, i'll take a crack at this.

i think audiofreak gives you the answer that you want to know, although p.lacombe is also correct.

what, i think you want to know is what the ultimate "loudness" or acoustical power of your two woofer system will be.

for a two woofer system where the drivers are essentially operating as pistons, there will be a +3 db increase in spl (with the two voice coils in parallel).

Also, in general, most amps are constant voltage sources, so to speak, and you end up with an extra +3db from this (this is where you have to make clear whether you're speaking voltage sensitivity, power sensitivity, or, i suppose, driver efficiency (which doesn't change)).

Practically speaking, you will get a 6 db increase in spl with your ordinary amp into a 2 woofer acoustically coupled box with voice coils in parallel. Note your amplifier will supply twice the current in this scenario.

Series sums to extra 3 db from two acoustically loaded woofers, minus the 3 db from half the current flowing though the series voice coils. thus, the spl is the same.

Now, p. lancombe is right- once the driver is above it's piston range, the issue gets much more complex, with lobing from interference, etc. Also, as he notes, this assumes perfect acoustical coupling between the woofers.

see his interesting comments in this post.
search results
 
markkanof said:
AudioFreak,



Maybe I am just confused. When you say 1W/1M efficiency should rise by 3db is that when there is 1W to each speaker or is that one what to the whole box. Because I would think that if there was one what to the whole box then each woofer would only recieve 1/2 W. There for sound would be down in each woofer 3 dB for a total of -6 dB. But since there are two woofers output would increase by 3dB for a total net gain of -3 dB. Am I correct or is there something that I do not understand?

Thanks
Mark

No, Its when then entire array gets 1W therefore each driver will get 1/2W .... so each driver will be @ -3dB but when you add 2 wave identical waveforms on axis you gain 6dB so it should rise to 3dB above that of the single driver.....

All this is of coarse theory and in practice the gain is less.
 
I want to make sure I have this right, because I think that my situation is a little bit more complicated than the 2 in series v. 2 in parallel.

I am building a small line array that has 6 midwoofers. I wanted the midwoofers at the extremes to be attenuated relative to the rest of the array. Because I plan on biamping, it is not really a big deal (in fact, it is something of an advantage) if the total impedence of the woofer network is very high.

Each woofer is 8ohms nominal. The middle 4 are wired in series/parallel, so their impedence is also 8ohms. The woofer network is wired top - middle - bottom in series, so the total impedence is 24 ohms, with each group getting a third of the current and voltage as a single woofer would if hooked up to a voltage amplifier. The middle group is twice as sensitive as a single woofer w.r.t. voltage, so the whole thing is about 1/3 + 1/3 + 2/3 = 4/3 as sensitive as a single woofer? With 3 times the power handling?

Alternatively, if I am not concerned about attenuating the extremes, I can simply wire them all in series parallel, where the top and bottom 3 woofers are each wired in series (24 ohms), and the series groups are wired in parallel (12 ohms). This would yield 2x the sensitivity with 6x the power handling.

So is the math correct? What should I do?

Thanks, Won
 

Electro

Member
2002-04-12 4:41 am
I wouldn't wire any speaker driver in series except with dual voice drivers because specs will change. If you connect a resistor to a speaker driver. The Qes, Qts, Re, LVC, and SPL will change. Making the speaker driver sound worst than it did before. Also both speaker drivers will have different phase effects. Second, one speaker driver will be in phase on the positive while the second speaker driver will be slightly out of phase of the first speaker driver. Third the second speaker driver will act like a 1st order low-pass filter for the first speaker driver on each of the positive and negative part of the wave.

To hook up three or more speaker drivers in a circuit with out the side effects of series and damaging the amplifier from excess of current is to use transformers. Paralleling two speaker drivers should not be a problem for amplifiers. The crossover network also adds its own resistance, so the amplifier will be fine.

No, putting two speaker drivers in series will not give you an increase in SPL but a decrease in SPL. Putting speaker drivers in parallel will increase the SPL by 6dB (reality terms).
 

AndrewJ

Member
2001-09-05 6:37 pm
It looks like this question is destined to be asked and addressed forever and ever, but here is my contribution again :(
AudioFreak is absolutely correct from a theoretical point, assuming closely spaced drivers at low frequencies, where the wavelength is large relative to the spacing, and the drivers themselves are non directional. This applies in practice to typical drivers at frequencies below around 200Hz.
P.Lacombe is correct in a more detailed analysis where account is taken of spacing and directivity, ie in the region above around 200Hz. However I would take some issue with his statement that the full 3db is rarely achieved. I believe it is at low frequencies.
If we ignore off axis performance, then I can readily show with measurements that the on axis response of a pair of series connected drivers is identical to that of a single driver both in sensitivity and frequency response, given a constant input voltage, and +6dB for parallel connection. This would mean that I get very close to the +3dB efficiency increase at low frequencies.
 
Electro said:
No, putting two speaker drivers in series will not give you an increase in SPL but a decrease in SPL.

Practical terms, assuming 1W/1M efficiency, it stays the same @ low frequencies and close proximity.

Assuming the 2 drivers are identical or close to it, putting drivers in series shouldnt hurt the sound badly.
 
Hm. So, what would the best way to wire up 6 speakers then? Also, there is no passive crossover (as I mentioned, biamping), so concern towards that is misdirected. I would also like to avoid using a transformer.

I don't really believe that speaker 1 acts as a crossover element to speaker 2 if they are connected in series, and I'm pretty sure that they will be more or less in phase. Also, I'm fairly certain that damping would not be "killed" but only suffer a little. :) The fact that the overall Qes will change does seem valid, and I will look into that.

I suppose I should just have a little amp for each driver, then? :) Unfortunately, that isn't going to be a solution that I will pursue. Yet.

-Won
 
Another phrasing with a twist:

You get a 3dB increase, whether wired in series or parallel, because you are doubling the radiating surface.

Now if you have a SS amp that almost doubles its power into 4 ohms vrs 8 ohms then you gain another 3 dB if you parallel them, -3dB if you series them. (most home theatre amps and other cheap gear will not -- often these units put out less power into 4 ohms [or just blow-up]).

If you have a tube amp, you change the taps you are using so you get no gain (or loss) in power.

dave