I wonder why they didn't use a flare on the box panel where the fan sits?
A proper flare either side of that panel might help reduce noise.
A proper flare either side of that panel might help reduce noise.
An externally hosted image should be here but it was not working when we last tested it.
Never seen this thread before.... as some noticed, the principal is very similar to a helicopter's blade assembly....
Haveing the same problem as helicopters - huge mechanical sound levels.
Haveing the same solution as modern combat choppers, appropriatly shaped blade tips...
With down turned blade tips, you'll see on the tips of blades is a 'swept tip'. This has the effect of 'fooling' the air that the blade tip is actually going slower than it is, so delays the onset of compressability (i.e. speed of sound)
Haveing the same problem as helicopters - huge mechanical sound levels.
Haveing the same solution as modern combat choppers, appropriatly shaped blade tips...
With down turned blade tips, you'll see on the tips of blades is a 'swept tip'. This has the effect of 'fooling' the air that the blade tip is actually going slower than it is, so delays the onset of compressability (i.e. speed of sound)
outer ring?
OzMikeH, I think you are on to something with that outer rim!
This is what I needed, someone to think with fat improvements as opposed to tiny incremental measures OR radical dapartures.
I don't have a problem with the radical suggestions, but for me, they can detract from the forward progress. When someone suggests making an 8 foot ribbon driver using a slice of diamond as the membrane... OK, cool, but not the next thing I'm gonna try. Replacing the fan with a big fire to get ionized partices and then whack 'em with magnets. Ditto.
Alternately, when someone offers a tiny improvement (use the same fan subwoofer, but with better cables, etc.) that it doesn't rock my socks at all. Yeah, it might be better, but I'm looking for easier, simpler, faster and with an improvement in performance--or some combination of these. Actually, I think I would love a design that only worked half as well as Thigpen's, if I could build it for $400.
So now that the thinking caps are on...
Outer ring is brilliant.
[1] no additional "moving mass" as we'll consider the fixed rotational mass (everything that spins at constant velocity) as "stationary."
[2] huge benefit to potential stiffness of the blades and huge benefit to to the potential ability to twist/flex the blades in a controlled fashion. I've seen the outer ring on toy helicopters. It keeps the blades from getting bent on landings, and adds inertial mass to the props to keep them spinning longer.
OK, you've got me thinking. Now that we are grabbing IP rights, how about?
[1] BIG outer ring. Like a tube. Actually like a ducted propeller. Search google images for "ducted propeller" to see some
here. But you can spin the shell/shroud/duct/ring/tube thing. This may very well handle a lot of noise, and gives us the grip on the blade tips to twist against.
[2] fabric blades. Like a sail. Held taut by the outer ring and the inner hub. By fabric, I mean anything super flexible. (I think the rules state that we have to use Kapton film somewhere, right? 😀 )
[3] bicycle wheel with only 5 spokes. Center a blade on each spoke to hold the center of the blade (lengthwise). The blade could rotate with the spoke as an axis. If the blade is stiff, only one part of the blade would need to be moved.
[4] non-moving, but shape-changing blades. A blade could sort of push out one side, then the other. Possibly with a voice coil or servo. Again, it really easy to get a voice coil to move slowly. Or a Piezo element to transform the blade (or a stack of them). Lots of servo sort of possibilities here. Again, we are not worried about the "stationary" mass, just the mass that needs to be moved by the signal.
[5] use individual motors (likely stepper motors) mounted on the hub to turn the individual blades. You could even mount the motors on the outer ring. Actually, if we are still going for blade twisting (for efficiency and noise reduction) each blade could have an inner motor and an outer motor. I see having motors attached directly to the blades as a big potential improvement, as we could do away with all of the linkages. Also using a voice coil to ultimately rotate a shaft seems counter intuitive. (Like a backwards ServoDrive) Voice coils tend to be very inefficient (on the order of 1% to 4% efficiency) whereas more conventional motors can be much more efficient.
A car wheel spins about 12 times per second when driving at around 60 mph (~97 kph). This gives us a massive spinning thing or comparable magnitude. Maybe our entire rotating assembly could be put on a wheel balancer and little lead weights stuck on it, to get additional vibrations reduced? Actually, I shouldn't have mentioned this, as I don't want someone to be led to the idea of adding blades to their wheels and getting 10Hz at 120db on the highway.
OK, who's next? Record your bright ideas here for posterity. How can we start making some of these?
OzMikeH, I think you are on to something with that outer rim!
This is what I needed, someone to think with fat improvements as opposed to tiny incremental measures OR radical dapartures.
I don't have a problem with the radical suggestions, but for me, they can detract from the forward progress. When someone suggests making an 8 foot ribbon driver using a slice of diamond as the membrane... OK, cool, but not the next thing I'm gonna try. Replacing the fan with a big fire to get ionized partices and then whack 'em with magnets. Ditto.
Alternately, when someone offers a tiny improvement (use the same fan subwoofer, but with better cables, etc.) that it doesn't rock my socks at all. Yeah, it might be better, but I'm looking for easier, simpler, faster and with an improvement in performance--or some combination of these. Actually, I think I would love a design that only worked half as well as Thigpen's, if I could build it for $400.
So now that the thinking caps are on...
Outer ring is brilliant.
[1] no additional "moving mass" as we'll consider the fixed rotational mass (everything that spins at constant velocity) as "stationary."
[2] huge benefit to potential stiffness of the blades and huge benefit to to the potential ability to twist/flex the blades in a controlled fashion. I've seen the outer ring on toy helicopters. It keeps the blades from getting bent on landings, and adds inertial mass to the props to keep them spinning longer.
OK, you've got me thinking. Now that we are grabbing IP rights, how about?
[1] BIG outer ring. Like a tube. Actually like a ducted propeller. Search google images for "ducted propeller" to see some
here. But you can spin the shell/shroud/duct/ring/tube thing. This may very well handle a lot of noise, and gives us the grip on the blade tips to twist against.
[2] fabric blades. Like a sail. Held taut by the outer ring and the inner hub. By fabric, I mean anything super flexible. (I think the rules state that we have to use Kapton film somewhere, right? 😀 )
[3] bicycle wheel with only 5 spokes. Center a blade on each spoke to hold the center of the blade (lengthwise). The blade could rotate with the spoke as an axis. If the blade is stiff, only one part of the blade would need to be moved.
[4] non-moving, but shape-changing blades. A blade could sort of push out one side, then the other. Possibly with a voice coil or servo. Again, it really easy to get a voice coil to move slowly. Or a Piezo element to transform the blade (or a stack of them). Lots of servo sort of possibilities here. Again, we are not worried about the "stationary" mass, just the mass that needs to be moved by the signal.
[5] use individual motors (likely stepper motors) mounted on the hub to turn the individual blades. You could even mount the motors on the outer ring. Actually, if we are still going for blade twisting (for efficiency and noise reduction) each blade could have an inner motor and an outer motor. I see having motors attached directly to the blades as a big potential improvement, as we could do away with all of the linkages. Also using a voice coil to ultimately rotate a shaft seems counter intuitive. (Like a backwards ServoDrive) Voice coils tend to be very inefficient (on the order of 1% to 4% efficiency) whereas more conventional motors can be much more efficient.
A car wheel spins about 12 times per second when driving at around 60 mph (~97 kph). This gives us a massive spinning thing or comparable magnitude. Maybe our entire rotating assembly could be put on a wheel balancer and little lead weights stuck on it, to get additional vibrations reduced? Actually, I shouldn't have mentioned this, as I don't want someone to be led to the idea of adding blades to their wheels and getting 10Hz at 120db on the highway.
OK, who's next? Record your bright ideas here for posterity. How can we start making some of these?
Nordic and NVMDSTEvil,
Can you point me to some pictures of "swept forward" blades and "down turned blade tips" from helicopters? I would really like to understand how these are working.
Actually could I get pointed to and info somewhere on how they may fool the air, etc. I can't quite visualize how this would work.
Can you point me to some pictures of "swept forward" blades and "down turned blade tips" from helicopters? I would really like to understand how these are working.
Actually could I get pointed to and info somewhere on how they may fool the air, etc. I can't quite visualize how this would work.
bandwidth?
Actually, as long as we are brain storming here. The biggest item on my wish list for enhancements to this thing would be to increase the upper limit on frequency range.
I know that it already covers an "infinite number of octaves" from 20Hz to DC, but all but four and a half of these are below 1Hz...
Thigpen's woofer can go higher if it spins faster (I think he said 40Hz?) but the trade off is additional noise.
I'd really like to have a design that could tackle up to about 60Hz.
Actually, I'd like one that could be scaled up to large arena use. OK, if I need to use ten of them. Actually, I would even turn away from one that could only do 15 Hz to 40Hz, if I could get it really loud in a large venue.
I know this totally misses the point of the infrasub, but getting the crossover point up higher could really be, um, commercially viable.
The apparent reason for the upper limit on frequency seems to be the rotational speed. When to blades move quickly, they are only able to push on smaller masses of air. The pushes are of shorter duration at higher frequencies, so the "excursion" is much smaller.
How about designing one of these that picks up at the crossover point and is useful for 2 octaves? Use the infrasub for everything below 20Hz and and the faster-spinning version (or bigger blades, or more blades) to handle 20Hz to 80Hz.
As one further thought, it is possible that we may need some filters at some point to notch out a few notes. Sweep test tones in a building and filter out a few resonant frequencies that set up too much sympathetc vibration. We can avoid some room creak noise this way, as vibrating a wall panel 1" ptp has got to add some out of band crapola. This may also be a necessity in some situation to avoid structural damage.
On the other hand, if we could really find resonances and really hammer at 'em, if would be fun to knock down an old Las Vegas hotel without needing any explosives...
Actually, as long as we are brain storming here. The biggest item on my wish list for enhancements to this thing would be to increase the upper limit on frequency range.
I know that it already covers an "infinite number of octaves" from 20Hz to DC, but all but four and a half of these are below 1Hz...
Thigpen's woofer can go higher if it spins faster (I think he said 40Hz?) but the trade off is additional noise.
I'd really like to have a design that could tackle up to about 60Hz.
Actually, I'd like one that could be scaled up to large arena use. OK, if I need to use ten of them. Actually, I would even turn away from one that could only do 15 Hz to 40Hz, if I could get it really loud in a large venue.
I know this totally misses the point of the infrasub, but getting the crossover point up higher could really be, um, commercially viable.
The apparent reason for the upper limit on frequency seems to be the rotational speed. When to blades move quickly, they are only able to push on smaller masses of air. The pushes are of shorter duration at higher frequencies, so the "excursion" is much smaller.
How about designing one of these that picks up at the crossover point and is useful for 2 octaves? Use the infrasub for everything below 20Hz and and the faster-spinning version (or bigger blades, or more blades) to handle 20Hz to 80Hz.
As one further thought, it is possible that we may need some filters at some point to notch out a few notes. Sweep test tones in a building and filter out a few resonant frequencies that set up too much sympathetc vibration. We can avoid some room creak noise this way, as vibrating a wall panel 1" ptp has got to add some out of band crapola. This may also be a necessity in some situation to avoid structural damage.
On the other hand, if we could really find resonances and really hammer at 'em, if would be fun to knock down an old Las Vegas hotel without needing any explosives...
neededandwanted said:
Sorry, but i've never looked close at helicopter blades, just seen a few thing here or there and always been mad at how inefficient fan designs used for PC's are the like are.. 😉
Some other items that may help are focusing the flow using reverse-direction fins on the output side of the fan (much like the Delta series "FFB" Focus Flow fans) or counter-rotating blades.
My reference to swept blades was actually having the entire blade swept forwards by the way, so all the air is forced more straight down rather than outwards.
Here's a link that shows more detail of a VPP...
http://www.ef-uk.net/data/vari-prop1.htm
Hmmmmm....I wonder if you could put one of these in a tapped horn....
http://www.ef-uk.net/data/vari-prop1.htm
Hmmmmm....I wonder if you could put one of these in a tapped horn....
stepper motors make lots of noise. no good.
Fabric blades would flutter. or be under incredible tension.
Great idea with the very heavy ring.
the blades would be very flexible pushing and pulling (along the axis of the motor shaft)
that ring is going to flop back and forth at the infra audio frequency. the extra mass will reduce that effect.
a large duct all spinnning would break the blades off with the slighest imbalance if unsupported. You would need to rest it on wheels = more noise.
the wider ring with more mass is a very good idea. the spinning ring if spun inside a duct would be partially stabilised by "ground effect" like a hard disk head as it skims over the disk. (I think)
the rim cound be prevented from flopping back and forth with magnets, same poles facing each other. a magnetic support like this has increasing force the closer it gets. it may not be needed.
As a ptrototype:
I'm picturing the ring as 12 inch OD, 10 inch ID. in two pieces, each 2 inches wide. you bolt them together and clamp the blade ends between them. A bearing on the outisde of this is a possibility, or even three sets of soft skateboard wheels (will the flex in the trucks steer it back toward centre? I'm not a skater.)
the fan blades are made with pegs on the inside, these fit in holes in the hub. probably a lump of metal with 4 holes drilled perpendicular to the shaft.
a piece of 12 inch pipe on both sides, lined with 1 inch thick sound absorbing foam.
a flare at one end, which also supports the motor. the important thing is the cross section is maintained and the turbulence is minimised.
a couple of bearings to support the shaft, and coming in the other side is a lightweight pipe that fits snugly over the shaft, the outside this shaft is connected to the voice coil of a big, ugly high excursion woofer motor.
the business end of the pipe has 2 flanges, separated by 1/2 inch. bear this in mind.
the fan blades each have a little arm sticking out at an angle with a 1/2 inch bearing bolted on, the bearing runs between the flanges on the pipe.
pushing the pipe in and out changes the pitch of the fan blades.
High stress areas: the drive pins on the fan blades. these not only handle the twisting force but the thrusting force.
All of this could be made with basic tools and a small lathe, the hardest part is the fan blades. Remember this will be really noisy, really only a mechanical proof of concept.
Fabric blades would flutter. or be under incredible tension.
Great idea with the very heavy ring.
the blades would be very flexible pushing and pulling (along the axis of the motor shaft)
that ring is going to flop back and forth at the infra audio frequency. the extra mass will reduce that effect.
a large duct all spinnning would break the blades off with the slighest imbalance if unsupported. You would need to rest it on wheels = more noise.
the wider ring with more mass is a very good idea. the spinning ring if spun inside a duct would be partially stabilised by "ground effect" like a hard disk head as it skims over the disk. (I think)
the rim cound be prevented from flopping back and forth with magnets, same poles facing each other. a magnetic support like this has increasing force the closer it gets. it may not be needed.
As a ptrototype:
I'm picturing the ring as 12 inch OD, 10 inch ID. in two pieces, each 2 inches wide. you bolt them together and clamp the blade ends between them. A bearing on the outisde of this is a possibility, or even three sets of soft skateboard wheels (will the flex in the trucks steer it back toward centre? I'm not a skater.)
the fan blades are made with pegs on the inside, these fit in holes in the hub. probably a lump of metal with 4 holes drilled perpendicular to the shaft.
a piece of 12 inch pipe on both sides, lined with 1 inch thick sound absorbing foam.
a flare at one end, which also supports the motor. the important thing is the cross section is maintained and the turbulence is minimised.
a couple of bearings to support the shaft, and coming in the other side is a lightweight pipe that fits snugly over the shaft, the outside this shaft is connected to the voice coil of a big, ugly high excursion woofer motor.
the business end of the pipe has 2 flanges, separated by 1/2 inch. bear this in mind.
the fan blades each have a little arm sticking out at an angle with a 1/2 inch bearing bolted on, the bearing runs between the flanges on the pipe.
pushing the pipe in and out changes the pitch of the fan blades.
High stress areas: the drive pins on the fan blades. these not only handle the twisting force but the thrusting force.
All of this could be made with basic tools and a small lathe, the hardest part is the fan blades. Remember this will be really noisy, really only a mechanical proof of concept.
My experience with steppers has been with small ones. If I put a hard disk drive in your attic, I doubt its noise would interrupt your experience of "Blade Runner."OzMikeH said:
We are needing forces of much higher magnitude here though. A big industrial stepper is likely out of the question for noise reasons, but I would like to use a motor with some control, or feedback. Maybe a servo motor?
As far as noise goes, the motors on the hub (in the center of the fan) will likely be inside of a nose cone housing, so likely some dampening of noise will occur there. There will likely need to be some airflow to the motors themselves for cooling purposes.
I was thinking high tension on this. Just trying to work out how to twist the blades without a lot of force. I'd like to concentrate the driving force on moving air, rather than twisting metal. We need need an easily twisted blade of high rigidity, I think.
A loose fabric would indeed flutter on each inversion. Like a ship's sail it would billow in each direction.
If we do the bit where we have blades that are skinny at the tips and wide at the base (hub, center) then we can avoid all of the twisting. The idea would be that equal amounts of air are moved by the inner and outer areas of the blade. This may be very inefficient though, as the outer blade areas have the potential to move lots of air, because of the larger (linear) distance travelled.
Actually, we give up that potential efficiency with the twisting blades as well. They won't bite much at the tips at all.
The ring may need to be tethered with some spokes to keep it from moving. We should try to keep the blades from bending out of the plane of the fan in any case, I believe. For the outer ring to hold the blade tips, likely something will need to hold the outer ring. I certainly rather putin a few spokes than wheels of magnetic levitation/stabilization.
The magnetic repulsion is a really good idea, but when I get that many magnets, I am usually doing something else with them (ribbons, planars).
Actually, we could float the entire thing in a mag field. Even rotate it like and induction direct drive turntable. Motor noise drops to zero and the only noise left comes from air turbulence.
Ground effect is likely only going to be useful at very high speeds and controlling very small masses. A hard drive spinning at 7,500 rpm can float a tiny drive head, but a big spinning assembly going at 1/10 that rotational speed may not be handled. No idea, the linear speed will be higher though, since the circumfrence will be way more than 10x higher.
propeller head?
item #1
We need someone that knows about the physics of fan and propeller design here.
My study of swept blades seems to indicate the these designs do not even out the pressure across the blade, but are designed to reduce noise. Specifically noise generated by blade tips exceeding the speed of sound.
I don't know if these designs reduce noise at lower speeds.
Regardless of the total noise involved, I would much prefer noise generated that is higher in frequency, rather than lower. We can filter our higher noise, but low bass noise will be much tougher to filter mechanically.
LF noise could possibly be actively cancelled (mic, dsp, amp, driver downstream) but would be tough to eliminate using bandpass-like enclosures or mufflers, ducting, etc. And we would need to seperate the noise from the signal if it was in the band of the subwoofers operation.
I noticed 2 or 3 different blade shapes on the Thigpen fan. At least one straight-sided and one swept-D shaped.
item #2
I have been trying to visualize the air load component to driving the blades (driving the twisting movment), and am so far thinking that there really isn't one. (Bear with me a second here)
In a flat blade slicing no air, there is theoretically no air load force. Turn the blade by some amount and there is now a force as the air gets pushed. This force may be imagined as centered at the axis or rotation (twist) in a symmetric blade supported in the middle. But here's the idea: The leading half of the blade is getting a force that is attempting to drive the blade more "open" or towards a higher angle of attack. The trailing half is getting a force that is pushing the blade towards a more "closed" position (lower angle of attack).
Am I wrong, or is the net tortional force zero here? Half the blade gets a push one way (clockwise) and the other half get a push the other way (conterclockwise).
Or visualize it this way:
A weather vane supported near the front will tend to align with the wind flow direction because the force of air will tend to straighten it. If it is supported towards the back, the force will make it turn more open (flip it around).
If the axis of rotation is at the center of pressure there will be pressure at the center (I guess that's self explanagle) but no pressure to cause the blade to twist/turn/rotate.
I had been previously thinking that a fan, especially one at high speed was going to need a strong force to turn the blades.
Now, I don't think this is the case.
Somebody help me if I am wrong here.
item #1
We need someone that knows about the physics of fan and propeller design here.
My study of swept blades seems to indicate the these designs do not even out the pressure across the blade, but are designed to reduce noise. Specifically noise generated by blade tips exceeding the speed of sound.
I don't know if these designs reduce noise at lower speeds.
Regardless of the total noise involved, I would much prefer noise generated that is higher in frequency, rather than lower. We can filter our higher noise, but low bass noise will be much tougher to filter mechanically.
LF noise could possibly be actively cancelled (mic, dsp, amp, driver downstream) but would be tough to eliminate using bandpass-like enclosures or mufflers, ducting, etc. And we would need to seperate the noise from the signal if it was in the band of the subwoofers operation.
I noticed 2 or 3 different blade shapes on the Thigpen fan. At least one straight-sided and one swept-D shaped.
item #2
I have been trying to visualize the air load component to driving the blades (driving the twisting movment), and am so far thinking that there really isn't one. (Bear with me a second here)
In a flat blade slicing no air, there is theoretically no air load force. Turn the blade by some amount and there is now a force as the air gets pushed. This force may be imagined as centered at the axis or rotation (twist) in a symmetric blade supported in the middle. But here's the idea: The leading half of the blade is getting a force that is attempting to drive the blade more "open" or towards a higher angle of attack. The trailing half is getting a force that is pushing the blade towards a more "closed" position (lower angle of attack).
Am I wrong, or is the net tortional force zero here? Half the blade gets a push one way (clockwise) and the other half get a push the other way (conterclockwise).
Or visualize it this way:
A weather vane supported near the front will tend to align with the wind flow direction because the force of air will tend to straighten it. If it is supported towards the back, the force will make it turn more open (flip it around).
If the axis of rotation is at the center of pressure there will be pressure at the center (I guess that's self explanagle) but no pressure to cause the blade to twist/turn/rotate.
I had been previously thinking that a fan, especially one at high speed was going to need a strong force to turn the blades.
Now, I don't think this is the case.
Somebody help me if I am wrong here.
If the blades/ring are magnetized the housing of aluminum around the ring could induce a stabilizing effect due to aluminum reflecting magnetism (rotation of electrons required for this to happen though).
Possibly magnetizing the duct and having the blades/ring made of aluminum may work better (preferred way to do it, first way may not react the same).
Possibly magnetizing the duct and having the blades/ring made of aluminum may work better (preferred way to do it, first way may not react the same).
Re: propeller head?
You're right 😉
The forces we are attempting to counter-act will be the blades wanting to push away from the force they are generating to move the air in the desired direction. If you watch the thigpen video(s) you'll notice the blades seem to "flap" backwards and forwards a bit due to lack of bracing.
neededandwanted said:
You're right 😉
The forces we are attempting to counter-act will be the blades wanting to push away from the force they are generating to move the air in the desired direction. If you watch the thigpen video(s) you'll notice the blades seem to "flap" backwards and forwards a bit due to lack of bracing.
Not a lot of force to turn, but torque will vary greatly to maintain the constant fan speed with maximum pitch.
because the blade pitch always swings though zero you can use a smaller motor, if the "flywheel" is big enough to maintain the inertial at maximum pitch (excursion)
think of pushing a disc through the duct with your hand at the speed the air will be moving (peak airflow in one direction only)
this would roughly be the force required to be exerted by the motor, how that force on you hand equates to torque, I have no idea. I'm sure someone can calculate the excursion and maximum cone velocity required for a theoretical 10 inch woofer to generate 10Hz. Of most interest is the peak cone velocity, this would be the airspeed we need to aim for at full positive (or negative) pitch.
Reversing airflow may cause cavitation issues, we want pressure variations (sound) but consider the variations in pressure within the cylinder of air that is moving back and forth through the fan.
The varying force on the fan blades would be like a plane flying through severe tubulence. Assuming an indestructable mechanism the cavitation effect would be the limiting factor of maximum volume at a given frequency. Would it limit max volume at lowest freq say 10Hz) or limit max frequency at a useable volume (rolloff at high frequencies).
As the air pressure is lower the fan would move less of it.
You would have to size the fan to the room volume, and I expect most plaster ceilings would fall to pieces. Is a sealed room or a vented room better? Those tube amp filaments aren't going to last long.
the idea of trying to make all of the noises a high frequency is a very good one. If the ends of the duct are flares you could even put damping material completely covering the opening, think of a loose industrial air conditioner filter, Even a quilt or a blanket would work. Transparent at low frequencies, absorbent at high frequencies.
Faster rotation with less pitch would be better I think.
this will raise the noise frequency and make it easier to keep quiet. Along shaft on the motor, put it out in the garage. The muffler is a good idea too.
because the blade pitch always swings though zero you can use a smaller motor, if the "flywheel" is big enough to maintain the inertial at maximum pitch (excursion)
think of pushing a disc through the duct with your hand at the speed the air will be moving (peak airflow in one direction only)
this would roughly be the force required to be exerted by the motor, how that force on you hand equates to torque, I have no idea. I'm sure someone can calculate the excursion and maximum cone velocity required for a theoretical 10 inch woofer to generate 10Hz. Of most interest is the peak cone velocity, this would be the airspeed we need to aim for at full positive (or negative) pitch.
Reversing airflow may cause cavitation issues, we want pressure variations (sound) but consider the variations in pressure within the cylinder of air that is moving back and forth through the fan.
The varying force on the fan blades would be like a plane flying through severe tubulence. Assuming an indestructable mechanism the cavitation effect would be the limiting factor of maximum volume at a given frequency. Would it limit max volume at lowest freq say 10Hz) or limit max frequency at a useable volume (rolloff at high frequencies).
As the air pressure is lower the fan would move less of it.
You would have to size the fan to the room volume, and I expect most plaster ceilings would fall to pieces. Is a sealed room or a vented room better? Those tube amp filaments aren't going to last long.
the idea of trying to make all of the noises a high frequency is a very good one. If the ends of the duct are flares you could even put damping material completely covering the opening, think of a loose industrial air conditioner filter, Even a quilt or a blanket would work. Transparent at low frequencies, absorbent at high frequencies.
Faster rotation with less pitch would be better I think.
this will raise the noise frequency and make it easier to keep quiet. Along shaft on the motor, put it out in the garage. The muffler is a good idea too.
That's a new one for me. Where can I read about "aluminum reflecting magnetism?"NVMDSTEvil said:
I'm not sure, but I think most AC motors will draw additional current in order to maintain constant speed. This applies to constant velocity motors only, but I believe that the speed is governed by the frequency of the AC (50Hz or 60Hz).OzMikeH said:
If we have a good bit of mass being spun by the motor (not the signal) then the speed of the fan will tend to average out and smooth out the wow/flutter/ripple of the motor speed as well.
If we need to control the speed of the motor, we could just feed it a strong sine wave of the desired freqency, rather than wall current. The wall current does however give us a big head start--we don't need any amplifier to crank out a 2kW 60Hz signal.
I need to wrap my head around this a bit.
In the fan, rotational speed x blade pitch will equate to frequecy x excursion in a conventional driver (roughly).
This will set our upper frequency limit. Higher frequencies will need a higher cone velocity.
A higher frequency tone is going to have much less time to push air on each cycle. This equates to lower displacement. The displacement approaches infinity as the frequency approaches DC. (and halves for every octave we go up).
The fan device has a sort of built in crossover this way. (Actually, it would be interesting to match a fan sub with a dipole sub, since those have a built in slope in the opposite direction: 6db down per octave.)
You lost me here.
As a fan blade goes from sucking to blowing in one direction, it will be trying to push rarefied air. It has less air mass to push against. Is that what you are saying?
If so, I don't see how this is any different than any conventional driver. I don't necessarily see this as an issue and I can't imagine that we would get anywhere near 2atm on one side and vacuum on the other. That would be the theoretical loudest sound possible I guess.
Yep. I think we are going to be going pretty solid-state for everything. I would imagine that, short of destruction of home and gear, feedback will become an issue. Long before the tube element breaks, it will likely vibrate sympathetically and add coloration.
Even line-level cables could start getting microphonic. The piano down the hall will start playing itself, and I can't even imagine what a turntable would do!
Actually, most fabrics start blocking really low frequencies. Think of a microphone pop filter. Pantyhose stretched over a frame blocks ultra low plosive sounds. Ordinarily, these are seen as a brief DC bias of the desired signal, but it is really just a short term tone (half wavelength) that could be above 10Hz even.
A blanket is going to block wind. We are dealing with frequencies close to wind here. Even most superlight speaker grille cloth will block the ultra low frequencies. Pull off a cover and try to blow on your hand through it. Grill covers get used as microphone pop filters in a pinch.
Likely the sound absorbtion will need to rely more on materials that absorb on reflection, rather than transmission. Make a curved path to the listening room and line it with some material that is non-reflective from 100Hz up or something.
Otherwise some experiments may need to be done to get the transmission spectra of various materials. This is similar to working with video cameras that are sensitive to Infra Red light. It's hard to tell what will reflect IR, until you look through the camera.
Likely the motor can be encased. A really long shaft is going to need extra bearings for support.
Faster rotation (I'm told) increases the noise as the square of the velocity. Double the speed and quadruple the noise. So that's a factor anyway.
Otherwise, higher speed / lower pitch is desireable for lots of reasons. The blades will be stiffer to the air impact at lower angles of attack. The modulation will be more linear (the excursion is sort of a sine function). We will have more potential headroom to open the blades up more for higher pitched notes or just louder ones. Lower moving mass of the blades themselves (we want to spend our amp power moving air, not blades).
How did I miss this page?
http://www.eminent-tech.com/main.html
Here's a Thigpen rotary woofer made for a car!
7 inch fan! Looks useable up to 80Hz!
OK. So I guess we know it can be made smaller, faster etc.
http://www.eminent-tech.com/main.html
Here's a Thigpen rotary woofer made for a car!
7 inch fan! Looks useable up to 80Hz!
OK. So I guess we know it can be made smaller, faster etc.
neededandwanted - Unfortunately I dont have the link to the specific source where I got the info (aluminum reflects magnetism).. which is a shame because it had quite a bit of info regarding other things like gyroscopes as well. The guy who figured it out was big into gyroscopes and got tossed from his teaching position or something by suggesting they were exerting force(s) other than what we would normally expect... or something along those lines. Suffice to say, a magnet was levitated above a spinning aluminum plate (from a a hard drive I believe).
Anyways, here's some info you might find handy on the subject.
http://www.google.ca/search?hl=en&safe=off&q=aluminum+repells+magnetism&btnG=Search&meta=
"The link to M.R. Morrow - Lab Facilities" looks interesting.. http://72.14.253.104/search?q=cache...um+repells+magnetism&hl=en&ct=clnk&cd=6&gl=ca[/url]
Another search.. http://www.google.ca/search?hl=en&safe=off&q=magnet+levitated+above+aluminum&btnG=Search&meta=
Pira Bibliography looks good: http://72.14.253.104/search?q=cache:zpI_7I6e1tcJ:physicslearning.colorado.edu/pira/PiraSubTOC.asp%3FSTopic%3D5K20+magnet+levitated+above+aluminum&hl=en&ct=clnk&cd=10&gl=ca
edit - had to toss a peice of vb code in there so the link would quit showing a smiley... you'll have to copy/paste it all to get it to work.
Anyways, here's some info you might find handy on the subject.
http://www.google.ca/search?hl=en&safe=off&q=aluminum+repells+magnetism&btnG=Search&meta=
"The link to M.R. Morrow - Lab Facilities" looks interesting.. http://72.14.253.104/search?q=cache...um+repells+magnetism&hl=en&ct=clnk&cd=6&gl=ca[/url]
Another search.. http://www.google.ca/search?hl=en&safe=off&q=magnet+levitated+above+aluminum&btnG=Search&meta=
Pira Bibliography looks good: http://72.14.253.104/search?q=cache:zpI_7I6e1tcJ:physicslearning.colorado.edu/pira/PiraSubTOC.asp%3FSTopic%3D5K20+magnet+levitated+above+aluminum&hl=en&ct=clnk&cd=10&gl=ca
edit - had to toss a peice of vb code in there so the link would quit showing a smiley... you'll have to copy/paste it all to get it to work.
Makes a lot of sense with the fabric blocking low frequencies. I should have thought of the pop filter example.
And one more half-baked thought:
Ducted fans are more efficient for blowing air.
What we want is not to blow air in one direction but to pull it back and forth, so the air inside the duct closest to the fan might end up spinning around along with the fan.
And one more half-baked thought:
Ducted fans are more efficient for blowing air.
What we want is not to blow air in one direction but to pull it back and forth, so the air inside the duct closest to the fan might end up spinning around along with the fan.
Don't know if that would be a factor or not. It's hard to visualize the "what if's?" but when someone says "here's what happens" I can grasp it pretty well.OzMikeH said:
Does a ducted fan spin the air like a whirlpool? Does it eject a vortex of spinning air that generates other turbulence?
Is this more or less of a factor when the fan is larger in diameter?
I'm guessing that it won't be a factor because of the short duration on the cycles. At 20Hz, the air would not be pushed around many times. If we have a blade spinning at 12rps, it wouldn't even be a full spin.
We may be using faster motor rotation to get higher frequencies to be louder, but then we have proportionally shorter durations.
A fast blade playing a really low note would be the worst case, then.
With low angle of attack blades, the air is getting pushed much more out, than around.
Would it be worse if we spun the duct as well? (likely)
Is it possible that we will be putting our signal into compressed or rarefied air as a result? If so, we may have an impedance mismatch (blades to near air or near air to far air) or possibly we'd get a built in helpful transformer to match impedances.
The velocity and excursion calculations may play a big part in understanding this. Maybe we need to ensure that the highest frequency notes blow the air clear of the duct before sucking back in? A calculation of the equivalent pistonic excursion may show that the air itself never leaves the duct.
I imagine that this is not the case, especially for low frequencies, because the equivalent conventional driver excursion for the volume levels (for the same diameter) can be many feet.
Who knows? Maybe a duct stops air from spinning and an unducted fan generated severe vortexes?
Maybe a design that is sort of isobaric? Put two fans face to face so that they are counter-rotating.
I'm sure that not gonna help with the noise much though.
oh, oh oh.... that "half-baked thought" just blew a fuse in my mind! 😀
Take a look at this if you guys havent seen it before.. its based on the principle of air bouncing in the neck of an enclosure 😉
Very much worth the read: http://www.diyaudio.com/forums/showthread.php?s=&postid=1146138#post1146138
Take a look at this if you guys havent seen it before.. its based on the principle of air bouncing in the neck of an enclosure 😉
Very much worth the read: http://www.diyaudio.com/forums/showthread.php?s=&postid=1146138#post1146138
Hi. I've taken the time to real through this thread from first post to the present, and I have a suggestion: you guys should petition the moderators to get the title of this thread changed to - "The Wonder Years Thread"!
Back in my wonder years, my favourite TV show was Battlestar Galactica. I would sit in class at school and daydream about being Starbuck and flying a Viper, and shooting down cylons. I would sketch and design how I could build a Viper, and actually started to build one in my back yard from plywood scrapes and bits of sheet metal from old ductwork. I would sit in the seat(borrowed from a '74 Comet) and pretend I was tearing through space, pursuing an evil cylon. Ah...such sweet memories.
I think that there is a general thought that this device is easy to build. WELL, as an experience Viper constructor, I can say that's NOT the case. What happened to the 3 or 4 different members who said they were going to "take a crack" at building "a ghetto version" or a "rough and ready" fan subwoofer? Did they get distracted by homework or perhaps SpongeBob Squarepants?
You guys need to call them back, their contribution is VITAL to the success of this endevor.
And of course, it goes without saying that you can count on my full support.
Gentlemen, goodluck and Godspeed!
Back in my wonder years, my favourite TV show was Battlestar Galactica. I would sit in class at school and daydream about being Starbuck and flying a Viper, and shooting down cylons. I would sketch and design how I could build a Viper, and actually started to build one in my back yard from plywood scrapes and bits of sheet metal from old ductwork. I would sit in the seat(borrowed from a '74 Comet) and pretend I was tearing through space, pursuing an evil cylon. Ah...such sweet memories.
I think that there is a general thought that this device is easy to build. WELL, as an experience Viper constructor, I can say that's NOT the case. What happened to the 3 or 4 different members who said they were going to "take a crack" at building "a ghetto version" or a "rough and ready" fan subwoofer? Did they get distracted by homework or perhaps SpongeBob Squarepants?
You guys need to call them back, their contribution is VITAL to the success of this endevor.
And of course, it goes without saying that you can count on my full support.
Gentlemen, goodluck and Godspeed!
- Status
- Not open for further replies.