Split... Stevens speaker design

I'm not seeing where the boost would come from... see if you can work that out, the additional amplitude will come from creating additional difference between the compressed and decompressed air nodes in the travelling sound wave.

A pipe will have a resonance at a single frequency.
And [not to bastardize the description of a horn for the sake of a brief statement], a horn opens up exponentially in such a way that its surface resonates at multiple frequencies, down to whatever frequency resonance corresponds to the opening of the horn (the larger the horn opening, the lower it can play).

As far as how instruments play, yes.. it's a single note at a time, at least for wind instruments.
Even for string instruments, each string, one note. Multiple simultanious tones come from multiple strings, like playing a chord on a guitar, or a piano.

With loudspeakers, they simply emulate what the electrical signal does.

If you picture a sine wave, that's the easiest.
Pretend the X-axis represents time, the distance above the line represents increasing excursion of the cone in one direction, and distance below the line represents increasing excursion of the cone in the other direction.
If you tracked your position on that sine wave graph as time began moving, if the loudspeaker is doing it's job ideally, it's cone motion will parallel the point you are tracking on the sine wave at any point in time. At the very peak of the sine wave, the cone would be maximally excurted in one direction. At the point where the sine wave passes the 0 line in the center, the cone would be passing through it's natural "at rest" position, headed to excurt in the other direction.

Now, that's obviously a pure representation of one single note.
When you combine multiple frequencies, you create complex waveforms, which not only define the note, but the tibre and all other qualities of that tone that make it recognizable as a particular instrument and note. It's all frequency based, by fundamentals and harmonics, that are all wrapped up in what works out to be a very complex wave, looking nothing like a pure sine wave.
But plotted in the same way, it still would represent a curving moving line, this one snaking around in a much more randomly appearing manner.

But what the loudspeaker does with that signal is exactly the same.
If you tracked your position on that complex wave 'graph' as time began moving, if the loudspeaker is doing it's job ideally, it's cone motion will parallel the point you are tracking on the sine wave at any point in time.
You could point to any point on that wave, and know what the cone position would be when the speaker got to that portion of the signal.

So, while a speaker can technically do more than one note at once, it's really just as simple, it's the waveform that's more complex.

With a string instrument, you have resonances related to the length and mass of the string, so when you pluck it, it resonates at it's own internal resonance. One note.
With a wind instrument, you blow across a reed, or blow in such a way that your lips are the reed, they resonate at a single frequency, essentially, as the air moving across the reed is disturbed in a predictable and consistent manner - one note results, unless something changes, and then the note changes.

So, with this system, is the goal a single frequency resonance? (obviously not)
If not, then what has to be done, to get a static length and area pipe (which will inherently have a single resonant frequency of it's own, like an organ pipe) to be able to not just produce their resonant frequency, but to be able to actually reproduce complex waveforms?
The speaker takes care of the complex waveform part, but what takes care of the coloration the single resonance would add?

And again, then how does the fan add anything - boost? to the equation?

Phew.. :detective:

:bag:

 

I'm just reading it, and seeing the reaction of those involved. I see merit in most posts, but I see some apparent misunderstanding in some others. There have definately been 1 or 2 points I will comment on, and a couple I am using to possibly change my thought on our prior conversations about this design.

Lemme absorb this a bit more. Also, I like the diagram. I definately had a differant idea in mind when you brought this up to me origionally. Gives me a visual to get it working in my head. I see mention of why you decided a fan in this case, rather than a compressor, but I see less realism in it being used. I still think the basic principle of a servo woofer as a bit better, but I see your point in the lower frequencies working, but the uppers being non-existant. I still don't buy the full range ability of the speaker as per this design. I just fail to see the rapid compression of the air exiting the port that would be required to reproduce say 3000Hz or higher. It almost seems as if the surface area of the air being compressed is too great for high frequency wave development (without cancellation and/or absorbtion to the opposite side of the port as the exciter I mean).

 

http://www.phys.unsw.edu.au/~jw/clarinetacoustics.html

thot that'd be interesting


the noise (im guessing) is coming from the 'wind' noise
usually it's broadband noise, so it's pretty hard to control
only way is to lessen the turbulence

 

What I'm concerned with is that if the fan were on, or off, it wouldn't seem to make a difference...
Because the only pressure fluctuations that are being made are being made by the driver installed in the tube, and the difference between the high and low pressure nodes created (the amplitude of the sound) are no larger than the driver installed in the tube can make them.

Also, the frequency response will be controlled by the driver as well... any aspirations of full-range-sound will still be limited by this driver's abilities.

Put another way, if the fan were on, and no driver, there would be no sound, just air moving.
If the fan were on, and driver working, the sound would be as loud as the driver could make pressure fluctuations, no louder.
So if you turned the fan off, the pressure fluctuations wouldn't be any smaller, the sound would still be as loud, carry the same amplitude, as the driver could make it.

This is all because the driver is adjacent to the air movement, as opposed to being integral with the air movement.
I am picturing the driver would need to be integral to the air flow itself, controlling the release of either air pressure or vacuum, the driver being in control of the frequency being played (ability to reproduce a complex waveform), and in this way the fan (or fans) would be in control of the air pressure fluctuations (the amplitude).

I drew a diagram of what I am picturing, two chambers, one positive air pressure, one negative air pressure, a loudspeaker type motor (or could be a servo motor) controlling two valves on a common shaft.
When the valve is pushed into the chamber, it would be considered sealed.
When the valve is pushed out of the respective chamber, it would be open to airflow... rubber, so that the more it were open, the more air could escape/enter.
I drew it with the driver slightly pushing up, so the bottom chamber would be slightly open, the top chamber sealed.

Here's my diagram, showing this airflow-integral pneumatic type:

I don't know if anthing like this has ever been built, I just drew that up as one way to get the driver to become integral in controlling the airflow.

 

geo
forget the conventional speaker in that pic - just use
a servo motor...

EDIT: didnt see u already said servo....

geo: for your pic, i already see a problem - you're gonna
have a lot of hissing noise when the flaps open...
:unsure:

 

stephen, i see what you're getting at, but i dont that the wing will work
the airflows will just merge again after passing through the wing
and probably have turbulence after leaving the end of the wing...
i'm just imagining those smoke patterns u see when you have a wing in
a wind tunnel

as for your full flap covering the tube, that might work... but i think
you'll only be able to create the 'positive' pressure and not the
'negative' pressure part of the wave... not sure tho

and for the clarinet, the reed does move faster or slower depending
on how long the pipe is (which holes are covered) so they're working
in conjunction with the pipe - since the pipe is considered as a closed
pipe with one end open, there's only certain frequencies you can play
out of that length pipe...

but the thing is that you have to change the length of the pipe - not just
the vibration frequency of the reed

 

I just got what you were saying earlier with the wing

moving the air up or down like that with the wing will not
make any difference - seems like your physically trying
to recreate a wave in the air by moving a wing up and down
but when you see a wave drawn out on paper, it the magnitude
of the pressure (or velocity) you're seeing, not a physical
shift in the air up and down (or left and right). The air
is actually vibrating in a direction perpendicular to the source -
think of a speaker moving back and forth.

we're sorta saying the same thing... but changing the vibration
frequencies of the reed without changing the length of the tube
will decrease the amount of 'sound' coming from the tube
The two are inherently linked - the reed will tend to vibrate at
the natural frequency conditions defined by the length of the tube

you have a closed end tube, and the change in impedance wants
to make it so that there's always a pressure antinode at the open
end... you're gonna have standing wave conditions such that at 1/4
wavelengths, it'll be much louder than at non 1/4 wavelength
frequencies


if i'm wrong, plz correct me... and sorry if it sounds like i'm bashing
your ideas, just trying to help work it out...

 

i read about half of it, so IM READY TO REPLY

first things first...the diagram that was first shown seems inaccurate to me. the waves that are made of moving air that is excited will propagate differently than in the picture. the driver will compress and rarify the air, making the sound output be much like that of a port. the other thing is, when the air is compressed by the driver, theoretically (and likely functionally) would flow backward out the fan, or at least slow it down, causing a change in the frequency of the fan whine, making it hard to block out.

thats it for now...probably incorrect, but its my .02 nonetheless.

eric

 

I see our server is down again today (red 'X' for a picture)
It'll be back up in a bit, hopefully.
Anyway-

There hopefully wouldn't be hissing... if you can't avoid it the way I am trying to, I think there will be a real challenge avoiding a hissing in ANY pneumatic driver.

As described, they would be made out of rubber, they would be pushed/pulled at their centers, and secured at their outside edges, so "at rest" that burgundy line would be straight. I drew the picture with the driver starting to push up, starting to open the bottom chamber, leaving the top chamber sealed.
At whatever would be deemed to be "max excursion" (whose design determination would depend more on the question of how much air could be forced through a given size opening without hissing), the burgundy line - the rubber - would look more like a stretched '^' (or 'v' when opening in the other direction) than it does in the picture even.

In this way, when the valve starts to open, just a little air escapes. More air by volume can escape as the valve opens farther and farther, until it begins closing again and the airflow is progressively reduced until the valve seals the chamber again. That would describe 1/2 of a sine wave - this driver would be able to reproduce complex waveforms as well, as the valve motions would simply parallel the waveform.
So, in this way, the resulting sound escaping the driver would actually replicate the very un-binary (more progressive) pressure fluctuations of a real speaker. B)

I'm not sure a servo would be practical in this, honestly.
Efficiency is a great thing when you are talking about needing the driver to do ALL the work, excursion being the thing controlling amplitude.
But in this case, you just need enough power sent through the driver to be able to open the valves as far as is needed... it's the fans, and the maximum pressure generation you can run them at that will dictate the amplitude. :gunsmilie:

My only real concern is that a servo driver would not be able to replicate the higher frequencies as well, where a voice coil motor might have a broader frequency range.

But then again heck, I'm not sure a driver like this inherently could replicate the higher frequencies all that well, as it's literally going to have construction tolerances related to the actual on-the-fly reassembly of two halves of a sine wave, in the common "release" chamber there. Might get tricky. h34r:

 

That's definitely more like what I was thinking...
In that case, you have the wing in the path, directly simulating the air.
I can almost picture the disturbed air behind the wing.

That one makes more sense, but I'd be curious to see it operate in actuality.
After all, we're concerned with creating air pressure fluctuations.
As the wing moves, air is diverted around to the other side of the wing... but by the time you got to the other side of the wing, the air pressure hasn't changed.. at least much, right?
I can see disturbances in the air... but would it make sound?

And if it did, would it be the fan blowing that would be controlling the amplitude? Or the movement of the wing - because it's movement will be dictated by amplitude also.
What do you think?

 

I had posted a picture, but part of our server is down right now.
Have no fear, it'll be up again, and as soon as it is, you'll be able to see the picture.

It's actually two chambers, one positive pressure, one negative pressure... and essentially two valves, two flaps, one valve opening for the positive half of the wave (and progressively opening, proportional to how far above the '0' line on the sine wave graph you are), the other valve opening for the negative half of the wave (also progressively opening, mimicking the waveform).

Hopefully our server will be back up soon, anyway. :blush: