Exploring Resonance And Efficiency Of Speakers And Subwoofers

[Magluvin]

First on resonance....

Recently something I have come to realize is that when we put a speaker or subwoofer(I will refer to these as sub or subs here on in) in a tuned port enclosure, the resonance gives us a useable gain in the output around that tuned freq. Ive known this for some time but only recently am I realizing that here we have a real example of resonance giving us a real gain in comparison to when we try to obtain a resonance gain in circuitry. Im in the belief now that we must be going about it, in the electronic circuit form, the wrong way if we have not been able to obtain this gain as Im seeing in applying resonance to a subwoofer system.

Second on efficiency...

The input to a sub is considered electrical power input measured in Watts. The output of the sub system(sub in enclosure) is considered acoustical power output measured in Watts also. So if we have the numbers for each, we divide the Pin by Pae and we can calculate our subs efficiency.

Here is a site that has the equations and online calculators for speakers/subs efficiencies

http://www.sengpielaudio.com/calculator-efficiency.htm

Was on this subject in another thread and it was pointed out to me that a sub I was working with was only about 1%eff.  I was in a bit of shock. Could it be? Speakers/subs are essentially coreless motors driving the cone/diaphragm. These types of motors are super efficient, even dc brushed vs a cored brushless motor.
So it took some time to find answers on this. I found this site above and it can calculate the eff of a sub just by inputting the Sensitivity spec of the sub in question. The Sub that I was told was 1%eff sensitivity spec is 96db@1w.  So I entered the spec in the calculator and it came up with 2.5%eff.  Close it would seem, but that is a 250% jump from 1%eff. 1%eff reflects a 92db@1w sub.  So i started plugging other numbers in.  Another speaker I have is 105db@1w. I ran that number and got 20%eff.  Well that is a huge jump of 2000% from a sub at 92db@1w.  I have found 2 speakers so far that are 112db@1w. I ran the numbers in the calculator and it shows 100%eff. Im a bit bewildered. 100%eff??  That seems to be unheard of, doesnt it?  Like where do we see this 100%eff number on anything?  More on that later...


So now if we begin with a 112db sub and put it in a tuned port enclosure and tune it for a gain at the tuned freq, then we should end up with a gain in acoustical power in watts output that takes us over the top of 100%eff @1w input power. Pic of plot below.

Light blue trace      Sub in 50cu sealed enclosure to simulate an infinite baffle of which the sub output is not really affected by the enclosure. If I had put in 5000cuft the plot doesnt change much and 50cu is good for example.

Green trace     Sub in recommended ported enclosure for flat response given by WinIsd speaker program(free and excellent) 4.278cuft enclosure tunes to 23.72hz. We can see a gain already in the lower freq range compared to the subs output in a virtual free air environment, infinite baffle.

Yellow trace     Sub in same 4.278cuft enclosure but with a shortened port to tune to 60hz.  At 61,97hz we have a gain up to 124.49db@1w output.

We run 124.49db in the online calculator and we get a 1774.189%eff rating @1w input.

That is a large increase above the already 100%eff subwoofer, all with 1w input..


So now the challenge is to explore this further and work out a way to convert this all to an electronic circuit that can accomplish these apparent gains using resonance like we have not been able to achieve before in the ways we have been trying so far.

Clearly the tuned resonance had a large effect on the output of the sub system with a known input and a usable output. So where are we going wrong on the electronic circuit end of it all? How can we relate the sub gain example to a circuit that resonance is applied to and get these gains that are measurable Pin vs Pout and get usable results of over 100%eff??   As we go along, we need to have a high efficiency circuit to begin with, just like the 112db@1w speaker gives us 100%eff to begin with, then apply resonance to obtain gains above 100%eff. 

That is the goal of this thread.

Mags

[fritznien]

hang a second cone in front of the first. a speaker makes a good mic.
if you have 17 times the power in then it should be easy to get 1 times on the second speaker.

[Magluvin]

hang a second cone in front of the first. a speaker makes a good mic.
if you have 17 times the power in then it should be easy to get 1 times on the second speaker.

Well here is the problem there...

When you input the same freq the box is tuned at, the speaker moves very little and there is a lot of tuned output from the port. So we would have to combine the port and the driver speaker output to affect the receiving driver to get the receiver to move a lot to get higher voltage output than what is input to the system. There are some odd sub box configurations that may help us with this shown below.  Figure 8 has me thinking.  It seems unusual to have an extra ported lone chamber to the left. Does it ring at a particular freq that gives even more gain? Or if we replace that port in that lone chamber with a receiving speaker, could that be how we apply both the driver speaker and the driver tuned port to the receiver speaker to get it to move the most and to generate the most?

Lots to think about. But I think we will learn a lot from this model. As I see it the gains are there. We just have to figure out how to reconvert the Pae to electrical power out. 

Then we have to test the supposed 100% eff speaker to see if it lives up to its specs, and then see if it can also generate as well as it motors. Like having identical dc motors shaft to shaft, the gen motor will never produce the same or higher voltage of the voltage used to drive the motor side of the setup. Is that all just due to lack of efficiency? Or might we experience the same loss using a 100%eff 'speaker' as a what you say, microphone just because it may be a great driver but not configured to be a great generator? I dont know yet. Ive never had a device that is 100%eff yet. Will have to be tested. And there may be ways to build a box specific for the microphone speaker that helps it become a more eff generator.  Again lots to do and lots to think about.

But I think it could be a fresh start to understanding the potential gains of resonance that we havnt experienced in resonant circuits yet.  The gains are there. We just need to figure out how to convert them back to electrical power in the most eff way possible.



Mags

[Magluvin]

Ill be showing some things about speaker boxs soon here. Im sure most will find it all interesting.

Mags

[Magluvin]

Im going to add this here on increasing efficiency by only adding more drivers while applying the same input power for each case whether it is just 1 driver or 8 drivers. I wrote this below at OUR in a thread on motors. So what Im going to say in this thread about speaker efficiency may also apply to other drivers like motors.

This is what I wrote over there.....



Ive been saying something about this for some time now. The example is with speakers/subwoofers.  Here it is...

When we apply 1w input to a speaker and we measure spl in decibels(db) from a distance 1 meter away from the speaker, then that measurement is the Sensitivity spec of the speaker.

So if a speaker is say found in the specifications as 90db@1w Sensitivity, then that should be what we read on the meter..

Now an odd phenomenon is that if we now set up 2 of the same speaker together, and we still only input 1 total watt to the system, we will increase 3db to a new total sensitivity level of 93db@1w.   Now double the speakers again. 4 speakers, 1 watt total, 1/4w each speaker, we again increase 3db. So now our system is 96db@1w....

And we can double them till you cant fit anymore, every time you double the number of drivers and divide 1w total among the larger array, we gain 3db. 

Where does that gain come from???  Its not free certainly. You have to purchase twice as many drivers to jump to the next level.. Where does it end? Does it?

Ive said this before and maybe not here, this drawing below as shown in above texts, series connection or parallel, there may be a similar gain by doubling the motors. Shaft to shaft would be most eff vs belts and pulleys.  Has anyone really done this test with 2 identical motors before? Seems like a waste of time I know and many others agree before even trying. But... considering the real gain of just doubling the numbers of speaker drivers driven at the same total input of 1w in each case, by doubling we increase output, acoustic power measured in watts. 

Here it is..  We increase the efficiency of the system by only increasing the numbers of drivers while still maintaining the same 1w input for each case.  So why is it so impossible for that to be a probability for 2 motors driving the same train vs 1 motor? +3db?? That is just a noticeable difference in level of sound we can detect. From 1 speaker to 4 speakers at 1w total in, it is a very noticeable difference of +6db. Add 4 more speakers and divide the 1 watt among all 8 speakers and you are at +9db.  10db for us sounds like a doubling of volume level. All gained by simply adding drivers and not increasing the input. Where does that gain come from???

Like I said above. Nobody has probably ever tried to specifically measure the eff of 2 ganged motors against 1 motor. Maybe it is just a noticeable gain, like 3db??? Well what about 4 ganged motors vs 1 motor. Do we get a very noticeable eff gain like 6db from speakers would be a very noticeable gain??  Who would think to check that specifically, and even then to take it next level of 4 motors after not seeing much from just 2 motors compared to 1?   For the first thoughts of coming up with it would be extinguished by preconceived notions that nahh, could never happen..  Gain??? nahh,  not wasting my time.    lol  But the speakers are doing just that.

So if anyone is game on this then that would be the test to try. Measure the eff of 1 motor vs 4 ganged(very precisely as to not cause any added binding that would hurt the test)

Id go the same route as the speakers. No matter series or parallel connections of the additional motors, the input needs to remain at a particular watt input. 10w or 100w  what ever, but the total has to be the same for 1 motor being tested and 4 ganged to be tested and drive a consistent load with each config to see if 4 at 1w tot can drive the same load better than 1 motor at 1w in.  If someone only has 2 motors to compare to 1, then the accuracy of the test needs to  be great as there may only be a 'noticeable increase in eff and if there is a slight increase that cannot be ignored then the 4 gang would need to be tested to be very sure that there is a gain from just 1 motor alone. Instead of a 3db increase with 2 motors it should emulate a 6db increase with 4 motors vs 1.  And then we will know. Put it to bed, or ride that eff gain wave.

Mags

-------

Here is an example of what Im telling you..

The green trace is 1 subwoofer at 1w total in
The yellow trace is 32 subwoofers at 1w total in.

WinIsd speaker box software free.  A lot of big guys in the business will tell you they use it all the time. Simple and gets the job done right.

Check it. Recheck it.  Its all true.  Eff gain using more drivers at the same power in.

Mags


------

Confused about the gain part?

1 speaker.  90db@1w

2w in   +3db  93db
4w in   +3db  96db
8w in   +3db  99db
Double the watts in and increase 3db


1 speaker.  90db@1w

2 speakers 1w each 2w total  +6db  96db
4 speakers 1w each 4w total  +6db  102db
8 speakers 1w each 8w total  +6db  108db
Double speakers and double total watts in for an increase of 6db

So double the watts in gives 3db increase. Likewise decrease watts in by half, reduce output by 3db


Here is the link.  If we double the speakers and the total power in, can we say that we got the 6db increase just from doubling the power in? Well we only get 3db increase by doubling the power in only.  But we get 6db more if we double the total power in AND double the drivers.  So the addition of more drivers increases efficiency all on its own.

Yup ist all true..  Been under my nose for 25years in the car audio industry.  Back in the 90s guys would do 32 8in subs and run 100w total and beat guys with 2 15s running a kw.  Because there is efficiency in numbers. 

Mags