Since this isn't actually about 'audio' I hope it's not too far off topic. But if anybody knows how to do this, it's likely somebody here.
I want to use a subwoofer to create a blast of air at the rate of about 2-3 hz. This is not to create an oscillating pressure wave. The objective is to actually move a mass of air down and out of a short tube, repeating the pulse as often as possible. It's effectively using the speaker as a fast bellows.
The application is to be the air source for a vortex cannon. I've found some examples of this during research, but not much that is great. Probably the best example I've seen is the Disney Research aireal project. But I'm looking to build something larger, yet mobile - so it has to be as light weight as possible.
So the concept is simple. Hook up a subwoofer so that it faces into a cone (front chamber) that will focus and accelerate the mass of air that will then shoot out of a tube. The tube will likely be a slightly expanding cone, but with an end diameter still a good bit smaller than the diameter of the speaker cone.
The length of the vortex tube is likely related to maximum frequency of the pulses. To create a vortex ring, the expelled air needs a bit of time to develop its rotation. It also needs time to separate from the tube and should not encounter any kind of back pressure/vacuum from the speaker cone relaxing. So each vortex ring will need to pass through the entire path before the next cycle begins. Since I'm clueless about the math behind this, I figure I'll just experiment with different frequencies and enclosure designs until I get something in the ball park.
The signal driving this is clearly not going to look like a sine wave. It's going to be square. The voice coil will have to retract to full x-min, hold until pressure inside is replenished, then accelerate forward to full x-max with as much power as the driver can handle, then hold at x-max until the vortex ring is generated and leaves the cannon. So the subwoofer would likely spend most of its time at x-min or x-max. I know that subwoofers aren't designed to handle this kind of signal/movement and that I'll need to be careful about overheating the coil.
I'm hoping to get much faster pulses than I've seen happen with this kind of setup. What I mean here is not the repetition rate (hz) but the actual speed with which the vortex ring moves through the air. Because this isn't a pressure wave, it can actually move very slowly. You can get fast moving rings (60+ mph) with a combustion source, but that's not what I want to do here. This is needs to be a safe application for use by older kids. There's probably a trade-off due to the fixed volume of the driver's displacement. Extra speed will likely require a shorter vortex tube and smaller ring. It probably also means that we don't want to replenish the expelled air from the tip of the vortex cannon. Instead I'm imagining a flap valve mounted on the front facing chamber. So in effect there is a one-way flow of air. Because it would be difficult to fabricate a flap valve for a conic chamber, it likely means switching to a box-like chamber with a small outlet for the beginning of the tube.
I wonder whether it would help to have a rear chamber like a typical subwoofer. The increased pressure of a sealed (or ported, or flap-valved) rear chamber might give more oomph to the forward thrust of the speaker cone - assuming the pressure doesn't prevent full retraction to x-min. Would this part of the system benefit from the same principles that govern an audio-range subwoofer enclosure?
So I said it needed to be portable. The speaker enclosure will be mounted in a backpack and connected via flexible tubing to the vortex cone. So weight is an issue. The backpack will also hold the amplifier and the power source (lipo battery). So we are looking at a car audio amplifier.
Plus the fog machine - yup, we want it to issue smoke rings. So we need to stuff a miniaturized fog machine in the backpack. Experiments with a diesel glow plug and water-based fog juice seem promising.
Therefore, all of this needs to be as light as possible. I've looked around this forum for threads on lighter materials (sono tubes, birch ply) and unusual (spherical) enclosure designs. I need any ideas that could possibly be relevant. Since audio quality is not important, this could open up a whole range of possibilities. For example, why is metal seldom used as an enclosure material? I'm thinking a steel mixing bowl might serve for one of these chambers - and would be lighter than MDF and plenty rigid. I don't think resonance would be an issue here.
We just received a 12" boss cx122 speaker and a boss AR3000D amp for testing. If we see any promise with this we'd want to find a low profile speaker that has the longest possible throw. Advice here is welcome too. Can't really go larger in diameter on the speaker, because we're constrained by the size of the backpack. In case you haven't guessed it yet - this is for a real-life TF2 medic + medi gun outfit for my son.
This amplifier likely isn't going to work out. There was nothing in the specs on Amazon about the subsonic filter that can't be set below 15hz, but that seems to be a 'feature' of this unit. Is that typical? Frequency response is another stat that isn't mentioned on this model on Amazon, but most amps that show this stat seem to mention a low end of 15 or 20 hz. Does anyone know of a 12v amp that might work down with this ultra low frequency square wave? Maybe an amplifier is the wrong way to go and this needs to be a stepped up voltage source simply switched between closed, inverted and open via a relay.
We've tried a small scale proof of concept with a 3” 20 watt mini subwoofer. The basic operation was there, but the displacement was far too small to be noticeable.
Back to the voice coil heating issue. We might need to point a small fan at the voice coil? The signal will be generated by an arduino whenever the trigger switch is held closed. We imagine that the trigger would typically be depressed up to 30 seconds at a time (hopefully generating 60+ rings), but if we measure the heat as too high, we can limit the duty cycle in software. The large speaker we are playing with right now has a rating of 1400 watts and 4 ohms. But I’m guessing that doesn’t mean it can sustain 17A at about 70V with this slow square wave? If you know how to calculate the voltage range needed in this application – please speak up. I still want as much punch as this system can produce.
So any and all ideas are welcome, whether about enclosure design, amplification or control schemes.
I want to use a subwoofer to create a blast of air at the rate of about 2-3 hz. This is not to create an oscillating pressure wave. The objective is to actually move a mass of air down and out of a short tube, repeating the pulse as often as possible. It's effectively using the speaker as a fast bellows.
The application is to be the air source for a vortex cannon. I've found some examples of this during research, but not much that is great. Probably the best example I've seen is the Disney Research aireal project. But I'm looking to build something larger, yet mobile - so it has to be as light weight as possible.
So the concept is simple. Hook up a subwoofer so that it faces into a cone (front chamber) that will focus and accelerate the mass of air that will then shoot out of a tube. The tube will likely be a slightly expanding cone, but with an end diameter still a good bit smaller than the diameter of the speaker cone.
The length of the vortex tube is likely related to maximum frequency of the pulses. To create a vortex ring, the expelled air needs a bit of time to develop its rotation. It also needs time to separate from the tube and should not encounter any kind of back pressure/vacuum from the speaker cone relaxing. So each vortex ring will need to pass through the entire path before the next cycle begins. Since I'm clueless about the math behind this, I figure I'll just experiment with different frequencies and enclosure designs until I get something in the ball park.
The signal driving this is clearly not going to look like a sine wave. It's going to be square. The voice coil will have to retract to full x-min, hold until pressure inside is replenished, then accelerate forward to full x-max with as much power as the driver can handle, then hold at x-max until the vortex ring is generated and leaves the cannon. So the subwoofer would likely spend most of its time at x-min or x-max. I know that subwoofers aren't designed to handle this kind of signal/movement and that I'll need to be careful about overheating the coil.
I'm hoping to get much faster pulses than I've seen happen with this kind of setup. What I mean here is not the repetition rate (hz) but the actual speed with which the vortex ring moves through the air. Because this isn't a pressure wave, it can actually move very slowly. You can get fast moving rings (60+ mph) with a combustion source, but that's not what I want to do here. This is needs to be a safe application for use by older kids. There's probably a trade-off due to the fixed volume of the driver's displacement. Extra speed will likely require a shorter vortex tube and smaller ring. It probably also means that we don't want to replenish the expelled air from the tip of the vortex cannon. Instead I'm imagining a flap valve mounted on the front facing chamber. So in effect there is a one-way flow of air. Because it would be difficult to fabricate a flap valve for a conic chamber, it likely means switching to a box-like chamber with a small outlet for the beginning of the tube.
I wonder whether it would help to have a rear chamber like a typical subwoofer. The increased pressure of a sealed (or ported, or flap-valved) rear chamber might give more oomph to the forward thrust of the speaker cone - assuming the pressure doesn't prevent full retraction to x-min. Would this part of the system benefit from the same principles that govern an audio-range subwoofer enclosure?
So I said it needed to be portable. The speaker enclosure will be mounted in a backpack and connected via flexible tubing to the vortex cone. So weight is an issue. The backpack will also hold the amplifier and the power source (lipo battery). So we are looking at a car audio amplifier.
Plus the fog machine - yup, we want it to issue smoke rings. So we need to stuff a miniaturized fog machine in the backpack. Experiments with a diesel glow plug and water-based fog juice seem promising.
Therefore, all of this needs to be as light as possible. I've looked around this forum for threads on lighter materials (sono tubes, birch ply) and unusual (spherical) enclosure designs. I need any ideas that could possibly be relevant. Since audio quality is not important, this could open up a whole range of possibilities. For example, why is metal seldom used as an enclosure material? I'm thinking a steel mixing bowl might serve for one of these chambers - and would be lighter than MDF and plenty rigid. I don't think resonance would be an issue here.
We just received a 12" boss cx122 speaker and a boss AR3000D amp for testing. If we see any promise with this we'd want to find a low profile speaker that has the longest possible throw. Advice here is welcome too. Can't really go larger in diameter on the speaker, because we're constrained by the size of the backpack. In case you haven't guessed it yet - this is for a real-life TF2 medic + medi gun outfit for my son.
This amplifier likely isn't going to work out. There was nothing in the specs on Amazon about the subsonic filter that can't be set below 15hz, but that seems to be a 'feature' of this unit. Is that typical? Frequency response is another stat that isn't mentioned on this model on Amazon, but most amps that show this stat seem to mention a low end of 15 or 20 hz. Does anyone know of a 12v amp that might work down with this ultra low frequency square wave? Maybe an amplifier is the wrong way to go and this needs to be a stepped up voltage source simply switched between closed, inverted and open via a relay.
We've tried a small scale proof of concept with a 3” 20 watt mini subwoofer. The basic operation was there, but the displacement was far too small to be noticeable.
Back to the voice coil heating issue. We might need to point a small fan at the voice coil? The signal will be generated by an arduino whenever the trigger switch is held closed. We imagine that the trigger would typically be depressed up to 30 seconds at a time (hopefully generating 60+ rings), but if we measure the heat as too high, we can limit the duty cycle in software. The large speaker we are playing with right now has a rating of 1400 watts and 4 ohms. But I’m guessing that doesn’t mean it can sustain 17A at about 70V with this slow square wave? If you know how to calculate the voltage range needed in this application – please speak up. I still want as much punch as this system can produce.
So any and all ideas are welcome, whether about enclosure design, amplification or control schemes.
Just putting some thoughts down:
-A driver with an RMS rating of 1400W definitely has to get less than half the power if it's a square wave, at normal frequencys! So at 2hz or 3hz, it's going to cook fast.
-You can use a small 12V ventilator, a piece of tubing and direct the cooling air in to the cooling vent hole in the backplate of the magnet.
-A class T amp is efficient and light but I'm not sure if they all have subsonic filters...
-I don't think an enclosure for the back is needed nor beneficial (so, lighter!
)
-Get a neodymium magnet driver, lightweight
-You could use multiple small woofers in a cube; 5 drivers and the sixth surface would have the vortex. Heat would be distributed in five coils.
I like the project and will follow with interest!
-A driver with an RMS rating of 1400W definitely has to get less than half the power if it's a square wave, at normal frequencys! So at 2hz or 3hz, it's going to cook fast.
-You can use a small 12V ventilator, a piece of tubing and direct the cooling air in to the cooling vent hole in the backplate of the magnet.
-A class T amp is efficient and light but I'm not sure if they all have subsonic filters...
-I don't think an enclosure for the back is needed nor beneficial (so, lighter!
)-Get a neodymium magnet driver, lightweight
-You could use multiple small woofers in a cube; 5 drivers and the sixth surface would have the vortex. Heat would be distributed in five coils.
I like the project and will follow with interest!
Sounds like an interesting project.
A voice coil driven speaker is not very efficient at moving air, especially at low frequencies. Perhaps a tank of compressed air with a valve that could be pulsed would work better. A venturi air mover would increase the air flow.
Or maybe a fan that would provide the air pressure, and be gated to provide the pulses.
If you use a woofer, you don't really don't need an amplifier, feeding the battery voltage directly to the woofer through relays (mechanical or solid state) will work fine. A 555 timer circuit should work to provide the pulses. Using narrow pulses would help keep power dissipation down.
Two 12v (13.8)batteries in series into a 4 ohm load would give you 190 watts. I have no idea how much power you would need, depends on the woofer. Using only a positive pulse would eliminate the need for a flap valve.
As for a fog machine, dry ice and water might be a simpler way to do it.
Just a few thoughts, hope some of it may be useful...
Keep us posted on your progress...
A voice coil driven speaker is not very efficient at moving air, especially at low frequencies. Perhaps a tank of compressed air with a valve that could be pulsed would work better. A venturi air mover would increase the air flow.
Or maybe a fan that would provide the air pressure, and be gated to provide the pulses.
If you use a woofer, you don't really don't need an amplifier, feeding the battery voltage directly to the woofer through relays (mechanical or solid state) will work fine. A 555 timer circuit should work to provide the pulses. Using narrow pulses would help keep power dissipation down.
Two 12v (13.8)batteries in series into a 4 ohm load would give you 190 watts. I have no idea how much power you would need, depends on the woofer. Using only a positive pulse would eliminate the need for a flap valve.
As for a fog machine, dry ice and water might be a simpler way to do it.
Just a few thoughts, hope some of it may be useful...
Keep us posted on your progress...
2-3 Hz
Motor drive a large plywood plate...well braced.
Arduino control or 555 timer. (with motor drive cct)
(Even a Reciprocating Saw will do 50 Hz.)
Probably a lot cheaper than your speaker and amp.
Use a modified bicycle inner tube for the surround to seal the air path.
Any sort of sliding device for the guide/suspension.
Well braced styrofoam would also work and might have a better strength/weight ratio.
Motor drive a large plywood plate...well braced.
Arduino control or 555 timer. (with motor drive cct)
(Even a Reciprocating Saw will do 50 Hz.)
Probably a lot cheaper than your speaker and amp.
Use a modified bicycle inner tube for the surround to seal the air path.
Any sort of sliding device for the guide/suspension.
Well braced styrofoam would also work and might have a better strength/weight ratio.
Last edited:
Google "Airzooka" for a commercial example of a hand-powered vortex cannon that works quite well (I have one).
If you want to replace the diaphragm and elastic band with a speaker driver, you will first need to find one with the highest "xmax" you can afford in order to get the maximum air volume.
If you properly shape the driving waveform, you can minimise the required power. Also, you're using the driver effectively "open baffle", where drivers typically need only a small fraction of their rated power to reach xmax.
So work out how much voltage is required to hold the cone at xmax, and apply that long enough to move the cone back. Now apply as much voltage in the other direction as you can for a brief period, just long enough to move the cone to xmax in the forward direction. Finally, cut off the voltage or re-apply the "backward" voltage.
In theory, you don't even need an amplifier to do this. You apply a relatively low voltage source to pull the cone back to xmax, then perhaps discharge a capacitor charged to a high voltage through the driver to push it forward as fast as possible. The size of capacitor and the charged voltage needs to be calculated to push the cone to xmax in the forward direction without over-pushing it. You can do it with a timer chip (555) or Arduino, and a couple of MOSFETs. (And a capacitor and voltage source.)
Many drivers have an "xmech" figure, which is the distance the cone can move before mechanical damage occurs. This is usually significantly greater than xmax, and might be a better target to aim for - distortion rises above xmax, but is irrelevant in your application.
... and I suspect that putting the generator in a backpack and coupling it to the nozzle by a flexible hose is a bad idea. The losses in the hose will be too high.
If you want to replace the diaphragm and elastic band with a speaker driver, you will first need to find one with the highest "xmax" you can afford in order to get the maximum air volume.
If you properly shape the driving waveform, you can minimise the required power. Also, you're using the driver effectively "open baffle", where drivers typically need only a small fraction of their rated power to reach xmax.
So work out how much voltage is required to hold the cone at xmax, and apply that long enough to move the cone back. Now apply as much voltage in the other direction as you can for a brief period, just long enough to move the cone to xmax in the forward direction. Finally, cut off the voltage or re-apply the "backward" voltage.
In theory, you don't even need an amplifier to do this. You apply a relatively low voltage source to pull the cone back to xmax, then perhaps discharge a capacitor charged to a high voltage through the driver to push it forward as fast as possible. The size of capacitor and the charged voltage needs to be calculated to push the cone to xmax in the forward direction without over-pushing it. You can do it with a timer chip (555) or Arduino, and a couple of MOSFETs. (And a capacitor and voltage source.)
Many drivers have an "xmech" figure, which is the distance the cone can move before mechanical damage occurs. This is usually significantly greater than xmax, and might be a better target to aim for - distortion rises above xmax, but is irrelevant in your application.
... and I suspect that putting the generator in a backpack and coupling it to the nozzle by a flexible hose is a bad idea. The losses in the hose will be too high.
Last edited:
Speakers are generally rated with an AES band- limited pink noise signal with a 6 dB crest factor.
A sine wave has 3 dB crest factor, a square wave 0 dB crest factor, a 1400 watt AES speaker would handle 700 watt sine, 350 square wave.
Square waves (DC current) also heat voice coils more rapidly than sine waves.
A speaker has very poor efficiency for what you are trying to do, the speaker has only 2-3% efficiency to start with, then when it is "stuck" at the forward or rear position it is wasting the power in those positions, presumably 50% of the duty cycle.
A reciprocating drive motor (as Dug suggested) using a Geneva movement driving a piston (or a flexible balloon-like diaphragm) would provide far more air percussive air movement with far less power using cheaper materials.
Battery powered reciprocating saws are available quite cheaply, Geneva movements can be harvested from a variety of machines, or built using nylon plastic.
Art
Last edited:
I've always wondered about this... presumably, a square wave is the fundamental sine, plus odd order harmonics going up the frequency range.
In the world of subwoofers, we normal use lopass filters to limit the bandwidth of the box. in this case, unless we were producing a very low freq square wave (below 30hz), would we not generally be eliminating (or at least drastically reducing the level of) all but possibly the 3rd order harmonic via the lopass filter, thus rendering the square wave.... back to a sine(ish) wave?
The OP wrote: "I want to use a subwoofer to create a blast of air at the rate of about 2-3 hz. The signal driving this is clearly not going to look like a sine wave. It's going to be square. "
When a 2-3 Hz square wave is run through a speaker, the speaker simply goes forward, holds, goes backwards, and holds, repeat.
If the low pass was close enough to that frequency it would "soften" the square, but also reduce the "puff" vortex.
At any rate, using a speaker for an air puffer just seems grossly inefficient, and the OP wants lightweight and low power, there are better ways to achieve that than a speaker.
Art
simple crank mechanism
A small DC motor driving a simple (single sided) crank mechanism, moving a 1/8" thick Aluminum plate with a glued or clamped on rubber surround would probably work very well.
That would have the added benefit of variable frequency.
On that note, one down side is it would start at 0 Hz and spin up to frequency, and then spin down when you shut it off.
Or, getting fancier, a stepper motor would have almost instant full speed and not require a speed reducer.
.
As far as the crank, Picture an Aluminum disc with a set screw to clamp a motor shaft in it's center hole.
Then a Heim joint riding on a shoulder screw tapped into the face of the aluminum disc at the radius of the Xmax you want.
Connect the other end of the Heim joint to a round Aluminum disc with a rubber surround, or to the cone of a heavy coned loudspeaker.
Just an idea,
Dave
A small DC motor driving a simple (single sided) crank mechanism, moving a 1/8" thick Aluminum plate with a glued or clamped on rubber surround would probably work very well.
That would have the added benefit of variable frequency.
On that note, one down side is it would start at 0 Hz and spin up to frequency, and then spin down when you shut it off.
Or, getting fancier, a stepper motor would have almost instant full speed and not require a speed reducer.
.
As far as the crank, Picture an Aluminum disc with a set screw to clamp a motor shaft in it's center hole.
Then a Heim joint riding on a shoulder screw tapped into the face of the aluminum disc at the radius of the Xmax you want.
Connect the other end of the Heim joint to a round Aluminum disc with a rubber surround, or to the cone of a heavy coned loudspeaker.
Just an idea,
Dave
I have 3 grandkids that would love to have something like this. It looks like the woofer is probably the simplest way to go.
This guy looks to have the right idea:
Vortex Ring Machine Gun - YouTube
This guy looks to have the right idea:
Vortex Ring Machine Gun - YouTube
That one is cute, but this one can blow out candles and shoot rings 180 feet:
http://www.geekologie.com/2008/03/worlds-largest-air-vortex-cann.ph
There is no replacement for displacement!
Attachments
A low rpm motor turning a commutator to switch DC driving the highest excursion driver required will do this. Think if it was made with a real phase plug to mate with the Tube. Remember to make the vortex tube length at an even harmonic is important for optimal pressure and velocity where it terminates, driver dampening.
Also should say, I built one sub design that if you thumped the signal would hit you from across the room with an unintentional vortex.
Startled the heck out of me when first experienced.
You didn't hear any port noise or the like, was like having a silent fan buffeting you from time to time, actually nice.
It would launch NERF balls too
Startled the heck out of me when first experienced.
You didn't hear any port noise or the like, was like having a silent fan buffeting you from time to time, actually nice.
It would launch NERF balls too
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.