Here's something to think about...
Let's look at a typical voice coil electromagnetic force vector in 90 degree increments. 0 degrees is the starting point and outward voice coil movement coinciding with the positive going waveform soon follows due to electromagnetic force interactions with the magnet's static field. The applied positive going waveform soon reaches its peak at 90 degrees. Just past 90 and before 180 degrees the electromagnetic force is still positive in polarity, and declining, it is NOT pulling the voice coil back to its 0 voltage rest position, it is only reducing its positive force vector.
We need to be clear about the implications of this. The only reason the cone/voice coil returns home during this part of the wave cycle is because the mechanical suspension and theoretically the very small air load compliance (air load nonlinear as it may be) is trying to accomplish this.
Now, moving on to what happens from 180 to 270 degrees. At 180 degrees the voice coil electromagnetic force vector has finally reversed. Now (better late than never) we have a pulling force that is attempting to move the cone in the opposite direction. This force will be present until we reach 270 degrees and then we have the problem all over again until we reach 360 degrees.
What we have here is a condition where the cone/voice coil mass is slightly out of phase with the applied waveform between 90 and 180 degrees and between 270 and 360 degrees. The fundamental reason why is that we are using a single ended motor in application that requires a balanced symmetrical push-pull motor.
Of course it will be argued that single-ended motor topologies in dynamic drivers have existed since day one and continue to be marketed in the millions, and if this issue is such a problem, why does the design work at all? Well, yes, but that does not mean they don't all have this problem. I am certainly not an accredited driver designer, but it seems to me that relying on the less than linear restoring force and finite time constant of a dynamic driver's suspension to return it home to 0, 180, and 360 degrees does not constitute the ideal design.
Yes, the suspension has a shot of accomplishing this fairly well at frequencies below the fundamental resonance of the driver, but how can it possibly return the moving mass home in time to track and be in-phase with the applied voice coil signal at frequencies higher than the fundamental? I, for one, don't see it.
I stated in the title: “If an ICE valve-train were designed like a loudspeaker driver, we'd have a problem”. Do we as loudspeaker design aficionados realize that if an Internal Combustion Engine engineer designed his valve-train the same way a typical dynamic driver is designed what the result would be? It's called valve float, and the outcome can be, and normally is, disastrous....
If I'm correct, (and I always try my best to be) then the driver manufacturing industry needs to start paying attention to this issue so we can progress to bigger and better things.
Well as Peak Crackers would say, Much Love....
Let's look at a typical voice coil electromagnetic force vector in 90 degree increments. 0 degrees is the starting point and outward voice coil movement coinciding with the positive going waveform soon follows due to electromagnetic force interactions with the magnet's static field. The applied positive going waveform soon reaches its peak at 90 degrees. Just past 90 and before 180 degrees the electromagnetic force is still positive in polarity, and declining, it is NOT pulling the voice coil back to its 0 voltage rest position, it is only reducing its positive force vector.
We need to be clear about the implications of this. The only reason the cone/voice coil returns home during this part of the wave cycle is because the mechanical suspension and theoretically the very small air load compliance (air load nonlinear as it may be) is trying to accomplish this.
Now, moving on to what happens from 180 to 270 degrees. At 180 degrees the voice coil electromagnetic force vector has finally reversed. Now (better late than never) we have a pulling force that is attempting to move the cone in the opposite direction. This force will be present until we reach 270 degrees and then we have the problem all over again until we reach 360 degrees.
What we have here is a condition where the cone/voice coil mass is slightly out of phase with the applied waveform between 90 and 180 degrees and between 270 and 360 degrees. The fundamental reason why is that we are using a single ended motor in application that requires a balanced symmetrical push-pull motor.
Of course it will be argued that single-ended motor topologies in dynamic drivers have existed since day one and continue to be marketed in the millions, and if this issue is such a problem, why does the design work at all? Well, yes, but that does not mean they don't all have this problem. I am certainly not an accredited driver designer, but it seems to me that relying on the less than linear restoring force and finite time constant of a dynamic driver's suspension to return it home to 0, 180, and 360 degrees does not constitute the ideal design.
Yes, the suspension has a shot of accomplishing this fairly well at frequencies below the fundamental resonance of the driver, but how can it possibly return the moving mass home in time to track and be in-phase with the applied voice coil signal at frequencies higher than the fundamental? I, for one, don't see it.
I stated in the title: “If an ICE valve-train were designed like a loudspeaker driver, we'd have a problem”. Do we as loudspeaker design aficionados realize that if an Internal Combustion Engine engineer designed his valve-train the same way a typical dynamic driver is designed what the result would be? It's called valve float, and the outcome can be, and normally is, disastrous....
If I'm correct, (and I always try my best to be) then the driver manufacturing industry needs to start paying attention to this issue so we can progress to bigger and better things.
Well as Peak Crackers would say, Much Love....
A couple of comments. The relationship between phase and displacement varies with frequency. You have to look at the current phase angle (easily seen in the impedance curve) to understand it. At some frequencies the coil is pushing the cone out of the gap (displacement and force in phase). At other frequencies it is trying to push the cone back into the gap (force and displacement out of phase). This all stems from the 2nd order resonance nature of the system. The only big consequence of this that I know of is "jump phenomonon".
The valve float analogy doesn't really apply. With valves the spring force keeps the valve in contact with the cam or push rod. At some rev rate the restoring force is inadequate and the valves float (desmodromic solutions aside). The surface of the cam is falling away faster than the inertia of the valve allows the valve to follow. With the usual woofer topology the cone and coil are glued together as a unit, so "valve float" is impossible.
"the driver maunfacturing industry needs to start paying attention to this issue"
Really?
David
The valve float analogy doesn't really apply. With valves the spring force keeps the valve in contact with the cam or push rod. At some rev rate the restoring force is inadequate and the valves float (desmodromic solutions aside). The surface of the cam is falling away faster than the inertia of the valve allows the valve to follow. With the usual woofer topology the cone and coil are glued together as a unit, so "valve float" is impossible.
"the driver maunfacturing industry needs to start paying attention to this issue"
Really?
David
Yes, the valve springs (and in some valve trains, cam follower helper springs) keep the valve in contact with the rest of the valve train regardless of valve train type (Desmo valve trains being a special example) up to designed maximum RPM (frequency) only. I am trying to use an ICE valve train as one of many analogies that could be used to help understand what I'm trying to say here.
Let's keep this analogy as simple and as accurate as possible and look only at the intake valve train on a single cylinder two-valve 4 stroke cycle ICE. The valve train can be thought of as having a damped natural resonance frequency. This resonance frequency must be higher than the maximum intended frequency (RPM) this ICE is designed to operate to. If this frequency is exceeded, the valve spring cannot maintain necessary contact with the cam lobe and the valve train will float. This will in turn lead to valve train destruction if the condition is severe enough.
The key point I am trying to convey is that the valve will be under tight control and its movement will faithfully follow the cam profile up to, but not beyond the valve float threshold frequency because (and only because) it is operating in a compliance (valve spring and cam follower helper spring if the design uses one) dominated mode. When the valve float threshold frequency is exceeded, the system transitions to a mass dominated mode.
Speaker drivers whether they be woofers, midranges, or tweeters in the vast majority of designs operate in mass dominant resonant modes because driver resonance frequencies reside at the low end of the pass band they are required to operate in.
So this being the case (and hopefully explained well enough to clearly understand), I feel that one could now go to my original post to apply this operating condition to a real world driver utilizing a single-ended motor to visualize the problem I am trying to convey.
I am trying to shed light on how a typical driver's mode of operation differs from the compliance dominated tightly controlled mode of a well designed valve train. "Tightly controlled" is something I think we need to see in driver designs, and balanced symmetrical push-pull motors (not high driver resonance frequencies) is the correct way forward as far as I'm concerned.
Hope this helps...
An ICE is an air pump, plain and simple. Also the analogy of resonance and rpm is wrong. The rpm limit is based of combustion rate of a given fuel and failure rate due mechanical G forces imposed by velocity x mass of the moving parts and not harmonically related. Valves would only float if the spring tension is exceeded (that it if it has springs), effectively flying off the cam. Honda Racing R&D has done much research into what the limits are with ICE designs and backup everything mentioned. Also wonder why you mention 4 stroke, if we want to keep it simple why not 2 stroke? Perhaps a Sterling or a rotary analogy 
In your second letter you refer to the mass controlled region above resonance. Of course in that region compliance and compliance nonlinearity are not issues. It is, after all, mass controlled. "Returnng a driver to home" is never a problem at those frequencies.
For low frequencies there are some interesting effects as I mentioned before. DC offset or Jump phenomenon are an interesting study. They aren't always related to asymmetry, more frequently caused by dL/dx issues. The best cure is actually a especially non-linear spyder (read T.H.Wiik on the subject).
As to push pull motors, there have been a number designed, expecially by doug Button at JBL.
It seems to me that you are presupposing some big problems when you have no evidence that such problems exist (not that I would take away any audiophiles right to do so). What we might expect to get from a push-pull motor is symmetry. Still, there are a number of "single ended" motors that are highly symmetrical. Especially the well undercut core pole type motors (and some with top side extended poles) we can see very good performance with regard to symmetry. If even order distortion is low, as it frequently is, what would you hope to pick up? Perhaps you are advocating for a cure without a disease.
Regards,
David S.
Also wonder why you mention 4 stroke, if we want to keep it simple why not 2 stroke? Perhaps a Sterling or a rotary analogy
At least 2 strokes use genuine resonance (in the exhaust system) to make them run better.
Yes, 2 stroke expansion chambers. The engine worlds counterpart to the 1/4 wave line. I wonder if they have discovered long haired wool yet?
The problem is the OP was really making comparison with the valve train of internal combustion engines. That leaves out 2 strokes that run without valves.
He also gets it exactly backwards in that it would be better if the valve train were designed like the loudspeaker rather than the other way around. With regular valves the spring force keeps the valve and cam follower in contact with the cam profile. Spring force is quite high and valve float is set well above the usual rev range by design.
Still, if you could pull the valve up (strong fingers!) and release it, between the spring force and the mass of the parts you would see that it accelerates back to the valve seat at a finite speed. As long as the cam profile doesn't ever exceed that, there is no problem.
If the engine designers were clever as the speaker designers are, they would glue "voice coil and cone" together rather than try to maintain contact with spring force.
David
The problem is the OP was really making comparison with the valve train of internal combustion engines. That leaves out 2 strokes that run without valves.
He also gets it exactly backwards in that it would be better if the valve train were designed like the loudspeaker rather than the other way around. With regular valves the spring force keeps the valve and cam follower in contact with the cam profile. Spring force is quite high and valve float is set well above the usual rev range by design.
Still, if you could pull the valve up (strong fingers!) and release it, between the spring force and the mass of the parts you would see that it accelerates back to the valve seat at a finite speed. As long as the cam profile doesn't ever exceed that, there is no problem.
If the engine designers were clever as the speaker designers are, they would glue "voice coil and cone" together rather than try to maintain contact with spring force.
David
I, too, thought ICE Power amps.
Interesting analogy, but I don't understand how it applies. An engine valve train is push-push, isn't it? The spring pushes the valve closed, the cam pushes the valve open. Or Dr. Desmo's system with a cam is pushing both ways.
A normal loudspeaker is seems push-pull to me. The spider pulls the cone into place, current thru the voice coil pushes it toward me, or away from me, depending on the current direction. The spider pulls it back. Of course during most music, the voice coil current may be "pullng" too. Pulling toward the other side.
Add to that the fact the valves move only away from their seat and back again - not to the other side of the seat, and they do so with always the same magnitude, unlike a speaker - and I don't see them being much the same.
Interesting analogy, but I don't understand how it applies. An engine valve train is push-push, isn't it? The spring pushes the valve closed, the cam pushes the valve open. Or Dr. Desmo's system with a cam is pushing both ways.
A normal loudspeaker is seems push-pull to me. The spider pulls the cone into place, current thru the voice coil pushes it toward me, or away from me, depending on the current direction. The spider pulls it back. Of course during most music, the voice coil current may be "pullng" too. Pulling toward the other side.
Add to that the fact the valves move only away from their seat and back again - not to the other side of the seat, and they do so with always the same magnitude, unlike a speaker - and I don't see them being much the same.
ICE valve trains are designed to operate in the compliance controlled region. This would be well below the resonance of a loudspeaker. Loudspeaker drivers are designed to operate predominately in the mass controlled region where the equation of motion for the drive reduces to effectively F = ma. In this region the compliance plays the smallest roll. Damping plays a more significant roll though it too is not the dominant force.
The comparison is meaningless.
The comparison is meaningless.
The differences are too numerous to count, but here's a few:
The thermodynamic efficiency of an ICE is around 10%, from fuel to power at the wheels, while the electrodynamic efficiency of most loudspeakers on the market is from 0.5% to 1%, from electrical watts in to acoustical watts out.
The usable torque/RPM range of an ICE ranges from 1:3 to 1:6, with the ICE either stalling-out or tachometer redline below and above the usable RPM range. By comparison, loudspeakers have to span a nearly three-decade range, from 20 Hz to 20,000 Hz, with most loudspeakers on the market covering 50 Hz to 15 kHz, which is still 1:300, or two and a half decades.
The low efficiency of loudspeakers is a direct result of the extreme demand for bandwidth; buzzers and car horns are a lot more efficient, but they don't play music.
An ICE is an air-pump that exploits temperature differences from combustion to derive crankshaft power, while a loudspeaker is a linear electrical motor with an mechanical-to-acoustic radiator attached (the cone). The most direct automotive analogy is an electric motor, which has far higher energy efficiency than any possible ICE (in the 90% range). The problem with electric-motor cars are related to the low energy storage density and high cost of batteries, not the motors.
Although the electrical-to-mechanical conversion in a (dynamic direct-radiator) loudspeaker is quite efficient, the mechanical-to-acoustical conversion is extremely lossy. Horns can improve this by an order of magnitude, which is why they were used at the birth of loudspeakers and continue to be used in applications where high efficiency and high output are important.
The absolute energy levels are quite different; one acoustic watt will drive most listeners out of the living room, while an electrical watt is a tiny fraction of what it takes to power a bicycle. A realistic minimum for a small car would be 50 kilowatts, with 70 to 200 kilowatts more suitable for an average-sized car at freeway speeds.
The thermodynamic efficiency of an ICE is around 10%, from fuel to power at the wheels, while the electrodynamic efficiency of most loudspeakers on the market is from 0.5% to 1%, from electrical watts in to acoustical watts out.
The usable torque/RPM range of an ICE ranges from 1:3 to 1:6, with the ICE either stalling-out or tachometer redline below and above the usable RPM range. By comparison, loudspeakers have to span a nearly three-decade range, from 20 Hz to 20,000 Hz, with most loudspeakers on the market covering 50 Hz to 15 kHz, which is still 1:300, or two and a half decades.
The low efficiency of loudspeakers is a direct result of the extreme demand for bandwidth; buzzers and car horns are a lot more efficient, but they don't play music.
An ICE is an air-pump that exploits temperature differences from combustion to derive crankshaft power, while a loudspeaker is a linear electrical motor with an mechanical-to-acoustic radiator attached (the cone). The most direct automotive analogy is an electric motor, which has far higher energy efficiency than any possible ICE (in the 90% range). The problem with electric-motor cars are related to the low energy storage density and high cost of batteries, not the motors.
Although the electrical-to-mechanical conversion in a (dynamic direct-radiator) loudspeaker is quite efficient, the mechanical-to-acoustical conversion is extremely lossy. Horns can improve this by an order of magnitude, which is why they were used at the birth of loudspeakers and continue to be used in applications where high efficiency and high output are important.
The absolute energy levels are quite different; one acoustic watt will drive most listeners out of the living room, while an electrical watt is a tiny fraction of what it takes to power a bicycle. A realistic minimum for a small car would be 50 kilowatts, with 70 to 200 kilowatts more suitable for an average-sized car at freeway speeds.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.