Half a croissant, on a plate, with a sign in front of it saying '50c'
h a l f b a k e r y
Alas, poor spelling!

idea: add, search, annotate, link, view, overview, recent, by name, random

meta: news, help, about, links, report a problem

account: browse anonymously, or get an account and write.



Ultimate Subwoofer

Subwoofer built to move air, with frequency response from 0Hz up!
  (+2, -1)
(+2, -1)
  [vote for,

The idea behind a loud and efficient subwoofer is to move air back and forth at some low frequency (like say, 10Hz (so we can feel it even though we can't hear it) to 100Hz).

So why not use the thing that was designed to move air efficiently, i.e. a fan? Take an arrangement of two fans, sealed into the ends of two pipes. One fan extracts air from one pipe, the other blows air into the other pipe. Both fans operate at high and constant velocity. The free ends of the pipes are joined together into one fat pipe, but there is a valve at the intersection. The valve can open into one or the other pipe, or have them both open.

To produce 0Hz is easy - just open (or close) both pipes (let's assume the hissing sound of the air and fans can be filtered out with an enclosure). To produce 1Hz is easy, just swing the valve back and forth between the two pipes once a second (hence causing the air to pulsate in and out at the end of the fat pipe). The valve here is like a conventional speaker - being operated by moving coil in magnet (or vice versa).

Higher frequencies can be achieved this way, although the limiting factor will be the inertia of the valve combined with the difficulty of closing an opening the pipes when air is being forced through them. However, since this is a sub-woofer, we wouldn't need much more than about 200Hz.

toby, Nov 19 2002

Diagram of [BigBrother]'s solution http://members.aol....mjhammo/subwoof.htm
For those who can't make sense of my annotation. [BigBrother, Oct 04 2004]

Bass seeking missiles http://www.halfbake...20Seeking_20Missile
One reason why the UN should send the weapons inspectors here. [kaz, Oct 04 2004]

Acoustic piston subwoofer http://www.halfbake...a/Acoustic_20Piston
Another super-powered subwoofer idea [croissantz, Oct 04 2004]

Wikipedia: Rotary woofer https://en.wikipedi.../wiki/Rotary_woofer
The baked version of this. [notexactly, Apr 23 2019]


       rods - on the natural bass thing dont forget large animals eg. elephants that may have had some kind of effect on us in the past. if certain types/amounts of bass caused fear and so excitement that could explain an over-use of bass.
chud, Nov 19 2002

       Seems like you've just transferred the problem of moving a loudspeaker cone to that of moving a valve...
How about building a concert hall in a quarry. Then, when the musical score called for some powerful low frequencies huge explosions would be detonated (Wasn't this the basis of Hotblack Desiato's performances? - I can't remember...).
hippo, Nov 19 2002

       A couple of points:   

       The two fans would need to be perfectly synchronised. When two similar (not same) frequencies are present, a low frequency is introduced at a tone equal to the difference of the principal frequencies.   

       When a valve is shut, the fan is still operating? (I guess so because the time taken to start/stop fan ruins any chance of it operating effectivel at blowing air.) The said fan will be operating in an airtight environment and create a vacuum... or at least attempt to. This will give load the fan with extra work and (I think) make it noisier - or at least more unstable.
Jinbish, Nov 19 2002

       Although [toby]'s implementation has flaws, the basic technological concept isn't bad at all. The main trouble with the original idea is that of trying to block and unblock airflow. That requires way too much effort to be practical. However, it is much easier to have fluid flowing at constant speed and simply divert its direction.   

       Imagine instead three pipes connected in the shape of the letter 'H'. The fan is located in the crossbar and blows to the right. Flaps are located at the junctions and will direct the air flow primarily through the top half or bottom half depending on position. Now imagine the legs of this 'H' are pinched together at the top and at the bottom and merged. You now have a single opening at the top and another single opening at the bottom.   

       By adjusting the flap positions, the air can be made to flow from top to bottom or from bottom to top. Amplitude can be controlled by causing some of the air to travel an internal circuit rather than leaving the enclosure. The flaps should be mechanically linked in such a way as to force polar symmetry.
BigBrother, Nov 19 2002

       Church bells have got lots of bass. Thats part of how they carry so far. Also why the vampires hate em.
bungston, Nov 19 2002

       Thanks for the comments, guys. My implementation is definately not practical, but is there to give you an idea of how to use fans to move air, and valves to modulate the air movement. [Jinbish] i figured there would need to be "blow-off" valves on the sides of the tubes to let air in or out when the valve had closed their opening.   

       I like BigBrother's arrangement - it will even improve the efficiency since it is producing sound waves from both ends. Unlike a normal speaker, the sound waves in the back are in phase with the sound waves from the front, so no enclosure is needed!   

       I believe that this would produce a subwoofer that no "flapping-cardboard-in-the-air" speaker can reproduce (I don't care what the x-max is!). And it would do it with high efficiency, so you wouldn't need expensive high-output amplifiers, and it would do it in a relatively small package (does not need a huge air-tuned enclosure).
toby, Nov 20 2002

       In Los Angeles, I fi- I mean some people fire guns to silence subwoofers. Sometimes I- they fire at intervals which might be described as 0 Hz in terms of boom-boom-boom
thumbwax, Nov 20 2002

asp, Nov 22 2002

       Ok, this all seems well and good, but what about the little incidentals during the time that it takes the valves to close? The valves are actually moving air themselves, albeit not so much as the fans. And the valves need to be very strong indeed to defy all the air pressure that the fans are creating. The constantly changing ratio of in/out of phase air is probably not going to produce the kind of tight bass you're looking for. When the valves first open or get close to closing, the airflow into the wrong chamber is going to degrade the sound, and there may be weird little snaps as both valves open and close simultaneously. Add to that, most of the time is going to be spent in between fully closed states, which is definitely not what you were looking for.   

       Apparently, the big engineering feat here would be to minimize the distance that the valves have to travel, or reduce time spent between states. Also the valves must be able to initially defy their tendencies to stay closed, then open, travel, and close smoothly and quickly, yet still, being these robust and efficient valves, leave the moving of the air up to the fans. Maybe instead of a hinged design, you could use a linear kind of valve, like, um, an electromagnetic one? Then what do you have? Right, a speaker, in an enclosure.   

       Ok, I'm just babbling now. A big, slow croissant for the thought experiment, but this seems definitely unworkable, so it's perfectly at home right here.
absterge, Nov 22 2002

       [absterge]: I think you missed part of the point of the flaps. They are *supposed* to spend most of their time somewhere other than their extremes. And in fact, full closure of a passageway would be a bug, not a feature. Can you imagine the awful fluttery flappy slapping sounds that would result from full closure of a valve? Besides, we are interested in generating continuously varying wave forms, not square push-pull crap.   

       As for defying air pressure, a simple modification to the design goes a long way toward fixing that problem. Placing the pivot near the center of the flap drastically reduces the torque required by the driver, and at the same time it decreases the bending stress in the flap material. This happens because the air pressure has a much shorter lever arm to work through. Also, the force of the air on the leading half of the flap will tend to counter the force on the trailing half.   

       I agree that the motion of the flaps will create miniscule pressure waves opposite the direction of intent. But guess what? The waves produced by the left flap are cancelled by similar waves (but with opposite polarity) by the right flap, and vice versa. Symmetry is our friend.
BigBrother, Nov 22 2002

       I would suggest that a better approach might be to have the air blow into an object similar to a large organ pipe, but constructed so as to (1) avoid generating anything resembling a clean reflection at the open end, so as to avoid creating resonance there; and (2) mounting a transducer near the lips.   

       The reason organ flue pipes work is that a sound wave which reflects off the top of the pipe and returns to the bottom will be amplified and reflected as the small changes in pressure deflect air in and out of the pipe. I see no reason this principle couldn't be applied to amplify artificially-produced stimuli.
supercat, Nov 22 2002

       oh. so, it's supposed to be a mean, messy fluttering ordeal? ok then. In that case, I'm with thcg.
absterge, Nov 22 2002

       // The final output of your subwoofer will still be from air moving inward and outward through a series of tuned ports, correct? //   

       No, this isn't about tuned ports or resonance at all. This is about controlling the direction and magnitude of air flow from a fan. One way is to vary the speeds of opposing fans (not practical because real fans have inertia). Another way is to block and unblock air flow through two opposing fans (also not practical). A third way is practical: use flaps as control surfaces to divert air flow from a single fan either forward or backward (or partially internally recirculating) as needed.   

       Efficiency is not the point. In fact, it's not even a consideration. The point is to extend the frequency range to zero hertz.   

       A conventional speaker cone has finite travel. Since sound amplitude is related to cone velocity, lower frequencies require greater travel to give the same amplitude (maintain velocity for longer time) as higher frequencies. Near-zero frequencies require near-infinite travel.   

       A fan effectively simulates infinite cone travel. That allows frequency response reaching all the way to zero hertz. But since a fan's push is unidirectional, the flaps are needed to re-route that push in the opposite direction when needed.
BigBrother, Nov 22 2002

       Have you looked at the picture in my link? That's the design I'm referencing in my babble.   

       Oh, and yes it is an awful lot of work for something one doesn't hear. I don't think the justifying application is an audio device. More like a Dolby Cinema Super Surround Massage™ type of system. (You know -- for those scenes where an earthquake levels most of the hero's home town.)
BigBrother, Nov 22 2002

       Nice pictures, BigBrother. As I noted, it's probably simpler to use a small transducer to shift the airflow laterally, rather than a large flap valve.   

       Actually, though, I think what's really needed is to have the seats in a movie theater connected mechanically to electric motion devices which are cued off the soundtrack. After all, the purpose of the sub-subwoofers is to produce vibrations which can't be heard, but can be felt. Rather than coupling them through the air, it may be better to couple them mechanically.
supercat, Nov 22 2002

       A related method can be borrowed from pnuematic logic systems.
Compressed air travels through a pipe, and comes to a Y shaped divide. The air tends to favour one of the Y legs. A pilot line is drilled into the junction of the Y shape. When the pilot is pressurised, it gives just enough push to divert the main air stream into the other leg of the Y.
Therefore, with a small pilot, a large volume of air can be diverted from one direction to another.
This would be easier than operating a big mechanical valve.
Ling, Apr 07 2004

       Ling: Hmm... sounds like the bottom of an organ pipe [indeed, the principles are the same].
supercat, Apr 07 2004

       Yes, it does appear to be similar. Perhaps the organ pipe "pilot" is the pressure wave, coming back down from the top of the pipe.
Ling, Apr 07 2004

       //Perhaps the organ pipe "pilot" is the pressure wave, coming back down from the top of the pipe.//   

       Precisely. Now for the $10,000 question--one which I didn't figure out the answer to until well after I graduated college: by what mechanism do the holes on a flute or recorder change the pitch?
supercat, Apr 09 2004

       What would be the effect if one were to use a fan which was attached to a very powerful servo motor whose rotational speed would be controlled by the specified sound waveform? So for a 15Hz sound the fan would switch back and forth between clockwise and counterclockwise motion 15 times/second. To be sure, the electronics would be rather tricky since one would want to capture as much of the kinetic energy as possible every time the fan switched direction (since that would have a huge effect on efficiency).   

       BTW, is sound really a function of air motion? I thought it was a function of air pressure? In which case even a fan wouldn't produce zero Hz since air would equilibrate to move constantly at the old pressure level.
supercat, Apr 09 2004

       0Hz: a strange beast. It does not define what the pressure is: only that it is non-cyclic.
[supercat], inertia will be your enemy.
Do I get $10,000 for answering "resonance"?
Please donate it to [jutta] c/o hb.
Ling, Apr 09 2004

       Okay, since nobody seems to have quite got it (and I can't blame any of you, since it took me years to understand this) the key is phase vectors and their summation.   

       When the pressure wave reaches a hole in the side of the tube, some of its energy will be reflected and some will not. These different reflected waves produce vectors that must be summed, but not in the seemingly-obvious way (producing separate pitches).   

       Imagine that you have a speaker located at the junction a pipe assembly that will return exactly two equal-strength in-phase waves with delays of 0.9ms and 1.1ms; further imagine that the speaker is producing a 1kHz sinewave.   

       During the time it takes the first reflection to reach the speaker, it will have produced 0.9 waves, so the phase of the reflection will lead the speaker by 0.1 cycle. During the time it takes the second reflection to reach the speaker, it will have produced 1.1 waves,so the phase of that reflection will trail the speaker by 0.1 cycle. If the two reflections were of equal strength, their sum will precisely match the phase of the speaker, and so the speaker will resonate at 1.0khz.   

       If the faster wave produced a stronger reflection, the sum of the two reflections' phase vectors would lead the phase of the speaker, and so it wouldn't resonate at 1.0Khz. At some slightly higher frequency, the sum of the vectors would match in phase, and so the system would resonate at that slightly higher frequency.
supercat, Apr 09 2004

       If there are n different reflected/delayed copies of the signal, each of which has amplidute A[i] and delay D[i], then compute two functions over all frequencies f:   

       p(f) = sum[i=1..n]( A[i] * sin(2pi*f/D[i]) ) and
q(f) = sum[i=1..n]( A[i] * cos(2pi*f/D[i]) )

       Sounds will be produced at all frequencies f where p(f) is zero and q(f) is positive. If there is only one reflected/delayed copy of the signal and its amplitude is positive, there will be tones produced at all frequencies n/D[1] (where n is any positive integer). If there is one reflected/delayed copy and its amplitude is negative, there will be tones produced at all multiples of (1+2n)/2D[1]. If you consider the requirement that q(f) be positive, you'll see the reason for the different sets of harmonics.
supercat, Apr 09 2004

       OK so far, but is there an explanation for the Octave shift when blowing harder?
Ling, Apr 12 2004

       //OK so far, but is there an explanation for the Octave shift when blowing harder?//   

       Yes. The formula I gave will indicate several possible frequencies the pipe might produce, but will not indicate which one or which ones will be produced with what amplitudes. Because the amplifying element in the system (the "mouth" of the pipe) has very non-linear behavior, different harmonics are not generated independently. Generally one will dominate; the behavior of the pipe mouth will influence which one that is. Since the pipe mouth can behave differently with different amounts of winding, that can in turn cause a pipe to oscillate in any of several different modes.
supercat, Apr 12 2004

       Well, folks, as much as I've enjoyed the lovely tangent about the mechanics of resonance in wind instruments, I think it's too far off topic to take up so much space on this idea page. So I'm taking the initiative to clean house a bit and remove the bulk of my irrelevant annotations. I would encourage [Ling] and [supercat] to do likewise.
BigBrother, Apr 12 2004

       I might clean up some stuff, but the behavior of resonant air columns could be quite relevant to anyone seeking to improve upon the subwoofer concept.   

       For example, while it would be quite impractical to have a slider in a long air column behind a speaker which would position itself for optimal resonance at whatever frequency was being output, it might not be unreasonable to have an air column with some electrically-actuated valves like those on a flute or saxophone (but bigger). Given that electrons move much faster than sound, a DSP might be able to adjust the accoustics of the air column in real time.
supercat, Apr 14 2004

       Surely the ultimate subwoofer would just be to hire someone to wheel your chair backwards and forwards at a low frequency, say 0.1 Hz?
hippo, Apr 15 2004

       //Given that electrons move much faster than sound//
You sure about that? Their effect may move very quickly, but the electrons themselves at a snail's pace (relatively speaking). Much, much lower than the speed of sound in air.
coprocephalous, Jan 26 2009

       How many Hz does the shockwave of an explosive blast have? And can something really be described as oscillating at 0Hz? Isn't that just staying still? Can anything wobble at -ve Hz?
zen_tom, Jan 26 2009

       //How many Hz does the shockwave of an explosive blast have//
I suppose it depends on the force of the blast, i.e., how long before it takes for the air to rush back to the epicentre.
Where's [8th of 7] when you need them?
coprocephalous, Jan 26 2009

       //Can anything wobble at -ve Hz?// No. Think about it.
coprocephalous, Jan 26 2009

       Just musing on the imaginary oscillation of (5 + 3)iHz - it could oscillate on a different plane but be useful in the long run for some Real World solutions.
Probably already happens, come to think of it.
gnomethang, Jan 26 2009


back: main index

business  computer  culture  fashion  food  halfbakery  home  other  product  public  science  sport  vehicle