el`Ol said:
ok but ATC's magnetic material and 18 Sounds's active faraday rings have nothing to do with geometry
the 3 technologies i compared are just different geometries of motors. underhung and overhung are 2 more geometries.
superlinear and ative faraday rings are ways to reduce flux modulation, they do nothing to flatten the BL curve.
differential drive also arguably lowers flux modulation because the voice coils are wound in opposite directions but this is merely a side effect. the main motivation of differential drive is what it does for the BL curve.
in fact you could use all 3 technologies in the same driver so you cant pit them one against the other.
Re: Re: How can an overhung voice coil produce good sound?
in theory yes, in practice this simply does not work.
in case of underhung driver like Aura you have the field return structure closed only on one side of the voice coil which i think makes it highly unlikely for the field to be completely uniform in the gap.
in case of overhung driver (normal driver) you will have some of the flux from the magnet shorting right to the pole piece below the top plate, and almost no flux above the top plate ... this again is neither uniform nor symmetrical so distortion is inevitable.
the first thing that was tried to improve the situation was to make what JBL called symmetrical field geometry motor, or a T-shaped pole piece reducing the flux below the top plate.
TC sounds achieved the same thing in a different way - they extended the pole piece and top plate above what could be considered the magnetic gap to create additional "stray" flux above the gap to mirror such flux below it. this created symmetry but did not provide linearity. linearity was finally restored by LMS voice coil having less turns in the middle and more on the sides. i think this is absolutely genius because instead of fighting the stray flux (with a T-shaped pole piece) they actually used it to make the driver more linear.
differential drive is equally brilliant as it also uses the nonlinearity to fight itself by positioning what is in effect two motors to cancel each others distortions out.
XBL is not as clever. i think of as differential-drive-wannabe.
jneutron said:
in theory yes, in practice this simply does not work.
in case of underhung driver like Aura you have the field return structure closed only on one side of the voice coil which i think makes it highly unlikely for the field to be completely uniform in the gap.
in case of overhung driver (normal driver) you will have some of the flux from the magnet shorting right to the pole piece below the top plate, and almost no flux above the top plate ... this again is neither uniform nor symmetrical so distortion is inevitable.
the first thing that was tried to improve the situation was to make what JBL called symmetrical field geometry motor, or a T-shaped pole piece reducing the flux below the top plate.
TC sounds achieved the same thing in a different way - they extended the pole piece and top plate above what could be considered the magnetic gap to create additional "stray" flux above the gap to mirror such flux below it. this created symmetry but did not provide linearity. linearity was finally restored by LMS voice coil having less turns in the middle and more on the sides. i think this is absolutely genius because instead of fighting the stray flux (with a T-shaped pole piece) they actually used it to make the driver more linear.
differential drive is equally brilliant as it also uses the nonlinearity to fight itself by positioning what is in effect two motors to cancel each others distortions out.
XBL is not as clever. i think of as differential-drive-wannabe.
Re: Re: Re: How can an overhung voice coil produce good sound?
Again, the physical geometry can be used to reduce those effects, and it'll cost..
Gap uniformity is easy enough given enough computing power and no cost constraints. It is also easy to work it out so the flux is gone far enough from the pole piece in either direction..
I said "easy", but must qualify that as easy "in concept". Once cost is considered, the details, as you correctly state, become difficult. But considering how far we've gone with just those two concepts, for most it is enough..
As it were, I simply pointed out to the OP, why each are used. Your correct points are considerably above the desires of the OP. But still, nice points...
Cheers, John
Yah, the distance in the return varies from one end to the other, but the uniformity could be worked out via machining if they wished, either tapering the return metal, or the gap. I could see blind holes along the gap axis to alter the effective bulk permeability, much like fluffing lamination spacing to reduce permittivity in one field direction.vasyachkin said:
vasyachkin said:
Again, the physical geometry can be used to reduce those effects, and it'll cost..
Gap uniformity is easy enough given enough computing power and no cost constraints. It is also easy to work it out so the flux is gone far enough from the pole piece in either direction..
I said "easy", but must qualify that as easy "in concept". Once cost is considered, the details, as you correctly state, become difficult. But considering how far we've gone with just those two concepts, for most it is enough..
As it were, I simply pointed out to the OP, why each are used. Your correct points are considerably above the desires of the OP. But still, nice points...
Cheers, John
Re: Re: Re: Re: How can an overhung voice coil produce good sound?
well this is probably some of the reasons why LMS driver costs $900 a piece. the entire motor is CNC machined to a complex curved shape developed with use of FEA software. Aura driver in the same in that respect, it also has complex computer-optimized shape of motor and also runs $900 a piece.
if it is at all possible to make a driver by cutting or stamping metal the manufacturer would have to be out of its mind to use CNC machining instead to produce the parts. as such it seems to be reserved for cost-no-object "statement" products.
it might very well be cheaper to produce a differential drive motor out of simple cylindrical shaped parts than a "regular" motor using parts machined to complex shapes.
so here the point about simplicity becomes moot. i mean LMS is basically a regular motor but optimized to death ... and it costs more than either XBL or Differential Drive.
but in the end unless there is a really good reason to believe that a particular optimization was implemented in the driver then i just have to assume that it was not. in other words i think that "guilty until proven innocent" is the logical way to judge drivers.
ok, im done ranting
jneutron said:
well this is probably some of the reasons why LMS driver costs $900 a piece. the entire motor is CNC machined to a complex curved shape developed with use of FEA software. Aura driver in the same in that respect, it also has complex computer-optimized shape of motor and also runs $900 a piece.
if it is at all possible to make a driver by cutting or stamping metal the manufacturer would have to be out of its mind to use CNC machining instead to produce the parts. as such it seems to be reserved for cost-no-object "statement" products.
it might very well be cheaper to produce a differential drive motor out of simple cylindrical shaped parts than a "regular" motor using parts machined to complex shapes.
so here the point about simplicity becomes moot. i mean LMS is basically a regular motor but optimized to death ... and it costs more than either XBL or Differential Drive.
but in the end unless there is a really good reason to believe that a particular optimization was implemented in the driver then i just have to assume that it was not. in other words i think that "guilty until proven innocent" is the logical way to judge drivers.
ok, im done ranting
Re: Re: Re: How can an overhung voice coil produce good sound?
It's difficult to tell the out of gap performance from that curve. Does it continue as 1/r? I didn't see a gap dimension, so can't tell from the curve, how far out one would have to go to pull in 99% or so of the integral.
And I agree, not so easily done. as a WAG from that graph, I'd think you would have to go at least another 15 or 20 mm in each direction to get low enough on the curve..too much a price in my opinion.
Cheers, John
el`Ol said:
It's difficult to tell the out of gap performance from that curve. Does it continue as 1/r? I didn't see a gap dimension, so can't tell from the curve, how far out one would have to go to pull in 99% or so of the integral.
And I agree, not so easily done. as a WAG from that graph, I'd think you would have to go at least another 15 or 20 mm in each direction to get low enough on the curve..too much a price in my opinion.
Cheers, John
454Casull said:
yes, same technology from TC Sounds.
but LMT stands for Linear Motor Technology so LMT Technology is a bit redundant ( Technology Technology ) and maybe that is why they seem to have changed the name to LMS Technology which means Linear Motor System Technology.
here is the description from TC Sounds (Audiopulse is TC Sounds Mobile) describing the tech, referring to it as LMS Technology:
http://www.audiopulse.com/products/technologies/lms/
and here is the same technology described as LMT:
http://www.soundsolutionsaudio.com/forum/index.php?showtopic=2748
i think that LMS is the designation that will be used by TC Sounds from now on.
454Casull said:
fact of the matter is LMS drivers have some of the strongest BL numbers of any drivers on the market now.
sensitivity is low because it takes a strong (and heavy) cone to handle the forces generated.
also if the BL was made any stronger the driver would not be able to produce bass because the already low QTS would become even lower.
voice coil is indeed very long ... of course imho you're way off on the "just" and "excessive" part
Variable coil winding pitch does work very well to make the BL versus X curve very linear. The big problem is that is does this by throwing away sensitivity. For a given gap dimension, you want to fill as much of the gap as possible with wire to get maximum sensitivity, because it is B cross L that determines motor strength and B is always highest in the gap. To use variable coil pitch to make the BL versus X function more linear, you must reduce the number of turns of wire in the coil (mostly in the middle). By definition, for a given gap width and magnet volume you are now reducing sensitivity.
As Jneutron seems to be saying, it isn't easy to look at a B versus X graph and tell how linear the resulting BL versus X function will be. To do this, you need B measured out very far from the center of the gap, and you must know the coil winding height, then be prepared to do some integration. Attached are some examples of B versus X and BL versus X for a couple permutations of a motor. In all cases, the winding height of the coil is 32mm.
As Jneutron seems to be saying, it isn't easy to look at a B versus X graph and tell how linear the resulting BL versus X function will be. To do this, you need B measured out very far from the center of the gap, and you must know the coil winding height, then be prepared to do some integration. Attached are some examples of B versus X and BL versus X for a couple permutations of a motor. In all cases, the winding height of the coil is 32mm.
Attachments
Jack Hidley said:
i know this. this is why TC only puts this technology on the drivers with 4" diameter voice coil and 500 ounce quad stacked magnet.
TC at first had both 4" VC and 3" VC variants of LMS driver but it seems they dropped the 3" version. i have the 4" one in any case.
yet when compared to XBL the LMS wins because it only sacrifices some of the voice coil in the gap while XBL sacrifices a full half of it. not to mention LMS makes better use of the iron without any dead space that XBL has between the two gaps.
and when compared to DD the LMS has a flatter BL curve
with LMS when you sacrifice efficiency you only sacrifice that portion of it which is nonlinear. that is to say that with LMS when you sacrifice output you only sacrifice the distortion part of it, which i think is not a bad deal
i can also think of a way to improve DD drivers with variable winding density coils ... that would be ideal probably ... as linear as LMS but with less weight and perhaps less inductance.
Motor interest
Thought I'd revive this thread since it's the closest thing to a full-on discussion of motors I came up with on searching.
Thermal behavior has been touched on, but only lightly. Long-term thermal compression is one issue, and is addressed by airflow management, metal phase plugs acting as heatsinks, shorting rings improving transfer of heat into the pole, and other issues.
What do people think of short term? I'd think that a single-layer coil, on a large diameter aluminum former(preferably with a split to keep Qms high), would seem to be about as effective as it could be, with as many long-term cooling options in place as possible. This maximizes surface area exposed to air, which is unfortunately the best option we have without ferrofluid (which has its own problems). Large shorting rings are good but the coil has to get hot, to heat the air, to heat the ring. I think having the gap exposed somewhat also would tend to help. I've heard of the inside of cabs being significantly warmer than ambient, and that part of the reason for venting pro cabs is to give a path to outside air (sometimes).
Regarding shorting rings, it's been shown that bigger isn't always better from a Le/X perspective, but they certainly are for reducing flux modulation and heatsinking ability. The Eminence Magnum 15HO has a ludicrous amount of heatsink ability in its motor. The undercut polepiece has a large copper ring below, and the aluminum phaseplug acts as a ginormous one that's also exposed to air. Much of this is duplicated in the Apollo versions of the Lambda motors, with the addition of a sleeve through the gap. Seems that a lot of shorting is very desirable.
Thought I'd revive this thread since it's the closest thing to a full-on discussion of motors I came up with on searching.
Thermal behavior has been touched on, but only lightly. Long-term thermal compression is one issue, and is addressed by airflow management, metal phase plugs acting as heatsinks, shorting rings improving transfer of heat into the pole, and other issues.
What do people think of short term? I'd think that a single-layer coil, on a large diameter aluminum former(preferably with a split to keep Qms high), would seem to be about as effective as it could be, with as many long-term cooling options in place as possible. This maximizes surface area exposed to air, which is unfortunately the best option we have without ferrofluid (which has its own problems). Large shorting rings are good but the coil has to get hot, to heat the air, to heat the ring. I think having the gap exposed somewhat also would tend to help. I've heard of the inside of cabs being significantly warmer than ambient, and that part of the reason for venting pro cabs is to give a path to outside air (sometimes).
Regarding shorting rings, it's been shown that bigger isn't always better from a Le/X perspective, but they certainly are for reducing flux modulation and heatsinking ability. The Eminence Magnum 15HO has a ludicrous amount of heatsink ability in its motor. The undercut polepiece has a large copper ring below, and the aluminum phaseplug acts as a ginormous one that's also exposed to air. Much of this is duplicated in the Apollo versions of the Lambda motors, with the addition of a sleeve through the gap. Seems that a lot of shorting is very desirable.
Oh, wow, no idea why I never responded to this.vasyachkin said:
fact of the matter is LMS drivers have some of the strongest BL numbers of any drivers on the market now.
sensitivity is low because it takes a strong (and heavy) cone to handle the forces generated.
also if the BL was made any stronger the driver would not be able to produce bass because the already low QTS would become even lower.
voice coil is indeed very long ... of course imho you're way off on the "just" and "excessive" part
A high BL doesn't mean that there isn't any L (or B, for that matter) being wasted. Make an ultra-overhung coil infinitely long with any standard motor and the BL isn't going to change worth a squat.
The sensitivity is probably low because of the mass and the DCR.
Hi,
the only useful answer to this is: if it works, it works.
Many spekaer builders, regardless if DIY or professional, concentrate on one particular thing. This is wrong. If you see a chassis, look at the measurement results (especially amplitude response and harmonic distortion) and take your pick. Don't bother with productional details, just concentrate on the results.
Bye
Baseballbat
the only useful answer to this is: if it works, it works.
Many spekaer builders, regardless if DIY or professional, concentrate on one particular thing. This is wrong. If you see a chassis, look at the measurement results (especially amplitude response and harmonic distortion) and take your pick. Don't bother with productional details, just concentrate on the results.
Bye
Baseballbat
Wow! A lot of discussion under the assumption that a linear Bl is the goal. What if its NOT the goal! I don't want a linear BL! Linearity is highly overrated - even undesirable in my mind.
No BL curve can be linear for any excursion - period. It MUST go to zero at some point. HOW it goes to zero is the real issue! Here's a hint! Going to zero sharply is NOT the answer.
No BL curve can be linear for any excursion - period. It MUST go to zero at some point. HOW it goes to zero is the real issue! Here's a hint! Going to zero sharply is NOT the answer.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.