I think the idea is that part of the voice coil is in the magnetic gap, and part of the voice coil is out of the magnetic gap compeletely-not being driven.
In other words, as long as a 6 mm wide magnetic gap has a full 6 mm worth of voice coil inside it, it should be linear.
The non-linearities occur when a excursion has carried part of the voice coil out of the gap completely. So 6 mm of magnetic gap is only reacting to 3 mm of voice coil.
But as long as a 17mm, 25 mm, or even 40 mm voice coil has 6 mm of it's length inside the 6 mm magnetic gap, it should be relatively linear.
That is why the linear excursion formula is:
(length of voice coil - magnetic gap length) / 2.
So a 17 mm voice coil with a 6 mm gap has a linear excursion of
(17mm - 6mm) / 2 = 5.5 mm.
Whether the overhung voice coil ever has quite the linearity of an underhung voice coil-one where the gap is 10 mm and the voice coil length is 6 mm-might be another question.
In other words, as long as a 6 mm wide magnetic gap has a full 6 mm worth of voice coil inside it, it should be linear.
The non-linearities occur when a excursion has carried part of the voice coil out of the gap completely. So 6 mm of magnetic gap is only reacting to 3 mm of voice coil.
But as long as a 17mm, 25 mm, or even 40 mm voice coil has 6 mm of it's length inside the 6 mm magnetic gap, it should be relatively linear.
That is why the linear excursion formula is:
(length of voice coil - magnetic gap length) / 2.
So a 17 mm voice coil with a 6 mm gap has a linear excursion of
(17mm - 6mm) / 2 = 5.5 mm.
Whether the overhung voice coil ever has quite the linearity of an underhung voice coil-one where the gap is 10 mm and the voice coil length is 6 mm-might be another question.
hi
there are 2 seperate things
as -
1) the entire motor structure being - non linear
2) only the magnet field at the magnetic gap being non linear
for no 2 - there are 3 simple appraches that will make the magnetic field symmetrical - hence jbl term - sfg , symmetrical field gap
1) an overhung sysytem in a symmetrical magnet field is never non - linear - as the driver will be designed to have a certain number of turns in the magnet field at all times
on which condition the driver specs will be measured and mentioned
- there-fore the bl
any conditions beyond which the bl will reduce - or the number of turns in the magnetic gap will reduce - is generally beyond which a driver is designed and therefore not recommended for usage
- x max
the overhang voice - coil system is linear ( other than when crossing the x- max - beyond the drivers design ) on all instances when the magnetc gap is of a - sfg design
suranjan das gupta
transducer design engineer
there are 2 seperate things
as -
1) the entire motor structure being - non linear
2) only the magnet field at the magnetic gap being non linear
for no 2 - there are 3 simple appraches that will make the magnetic field symmetrical - hence jbl term - sfg , symmetrical field gap
1) an overhung sysytem in a symmetrical magnet field is never non - linear - as the driver will be designed to have a certain number of turns in the magnet field at all times
on which condition the driver specs will be measured and mentioned
- there-fore the bl
any conditions beyond which the bl will reduce - or the number of turns in the magnetic gap will reduce - is generally beyond which a driver is designed and therefore not recommended for usage
- x max
the overhang voice - coil system is linear ( other than when crossing the x- max - beyond the drivers design ) on all instances when the magnetc gap is of a - sfg design
suranjan das gupta
transducer design engineer
Neither conventionally-gapped overhung nor underhung coils have a linear magnetic field - therefore by your reasoning neither should sound good.
Truth is, overhung coil is a much better compromise than underhung in terms of economy and distortion performance in a long excursion woofer.
Underhung coils with large gaps need HUGE magnets and many (most?) do not run the iron near the gap saturated - which means that the magnetic field of the magnet system can be modulated by the coil - creating further nonlinearity.
Overhung coils also benefit from the fringe field more and their Bl falloff is shallower than an underhung design. While Underhung may make intuitive sense, the pitfalls in practice make it impractical for large excursion designs. TC Sounds (once famous for their underhung designs) 20+mm drivers are all overhung AFAIK.
Truth is, overhung coil is a much better compromise than underhung in terms of economy and distortion performance in a long excursion woofer.
Underhung coils with large gaps need HUGE magnets and many (most?) do not run the iron near the gap saturated - which means that the magnetic field of the magnet system can be modulated by the coil - creating further nonlinearity.
Overhung coils also benefit from the fringe field more and their Bl falloff is shallower than an underhung design. While Underhung may make intuitive sense, the pitfalls in practice make it impractical for large excursion designs. TC Sounds (once famous for their underhung designs) 20+mm drivers are all overhung AFAIK.
There's also XBL^2 where multiple gaps are used, and JBL has differential drive.
I had an idea once for a commutated voice coil where it would have many coils and change between them through the excursion.... well tom danley thought of it first and had a patent on it, so i was out of luck. I've never seen it in use, but I think it has potential.
Of course you know why you never see it in use, those servo motors keep mr danley occupied!
I had an idea once for a commutated voice coil where it would have many coils and change between them through the excursion.... well tom danley thought of it first and had a patent on it, so i was out of luck. I've never seen it in use, but I think it has potential.
Of course you know why you never see it in use, those servo motors keep mr danley occupied!
overhung vs underhung
In standard overhung motors, there is more steel below the top-plate than above the top-plate making 2 problems.
a) Induction differences as the coil goes in and out of the pole.
b) Non linear magnet field from the top to the bottom of the coil.
There are 2 ways to make good overhung voice-coils in standard magnet sandwiches.
1) If the pole-piece is under-cut so the area close to the voice-coil is the same "height" as the top-plate, there is good linear motor strength and also pretty good induction linearity.
2) If the pole-piece is extended to 2x the Xmax, there is good linearity. This is great for sub-woofers because the L is high and there is a lot of extra steel to help cool the coil.
Under-hung voice-coils are naturally very linear inside the Xmax and have low induction. These are usually low Xmax drivers though and distort a LOT when overdriven. This is because the voice-coils are real short. If a voice-coil that is only 4mm long goes out of the gap by 1mm, 25% is out of the gap, if a 22mm long voice-coil goes 1mm out of the gap, it is only 4.5% outside its normal gap.
Under-hung coils are low in induction and very light because the coils are short with less turns of wire than overhung coils. This is good for high frequency extension, but not great for long term power.
Under-hung voice-coil drivers also tend to be higher Qts drivers because of a low Bl. This is not so much because of a low B, but more because of a low L.
I design drivers with both types. They all have a place. It depends on the application.
Frank
In standard overhung motors, there is more steel below the top-plate than above the top-plate making 2 problems.
a) Induction differences as the coil goes in and out of the pole.
b) Non linear magnet field from the top to the bottom of the coil.
There are 2 ways to make good overhung voice-coils in standard magnet sandwiches.
1) If the pole-piece is under-cut so the area close to the voice-coil is the same "height" as the top-plate, there is good linear motor strength and also pretty good induction linearity.
2) If the pole-piece is extended to 2x the Xmax, there is good linearity. This is great for sub-woofers because the L is high and there is a lot of extra steel to help cool the coil.
Under-hung voice-coils are naturally very linear inside the Xmax and have low induction. These are usually low Xmax drivers though and distort a LOT when overdriven. This is because the voice-coils are real short. If a voice-coil that is only 4mm long goes out of the gap by 1mm, 25% is out of the gap, if a 22mm long voice-coil goes 1mm out of the gap, it is only 4.5% outside its normal gap.
Under-hung coils are low in induction and very light because the coils are short with less turns of wire than overhung coils. This is good for high frequency extension, but not great for long term power.
Under-hung voice-coil drivers also tend to be higher Qts drivers because of a low Bl. This is not so much because of a low B, but more because of a low L.
I design drivers with both types. They all have a place. It depends on the application.
Frank
Some overhung speakers (especially for subwoofers) have high inductance due to the fact that they have a much longer coil than underhung designs. I think underhung is truely the best if cost is of no factor, but it takes massive magnets and massive amounts of steel to get massive excursion. XBL^2 offer a good comprimise
Jim Thiel claims in the interview at his site to have 1/10 of the distortion of an overhung coil with the underhung construction of his flat diaphragm midrange.
http://www.thielaudio.com/THIEL_Site05/Pages/models/Current_Models/CS3_7/cs3_7midrangeB.html
I ask myself:
Is it better to have "worse" iron in such a case to get closer to saturation?
Or does the copper ring help to overcome this problem?
http://www.thielaudio.com/THIEL_Site05/Pages/models/Current_Models/CS3_7/cs3_7midrangeB.html
I ask myself:
Is it better to have "worse" iron in such a case to get closer to saturation?
Or does the copper ring help to overcome this problem?
I agree, the underhung coil makes sense where there will be very little, but very fast motion.phase_accurate said:
The overhung coil, theoretically, should be every bit as linear as a standard coil but it will suffer from having to drag the extra mass around with it. Thus, it's sensitivity will suffer. However, they benefit by remaining linear over a greater distance; thus they should have a relatively high Xmax. Low sensitivity drivers with huge excursions driven by big amps - great for cars that go boom.
From what I have read, companies like TC Sound in very long excursion overhung coil subwoofers increase the number of turns/inch at the extreme ends of the overhung voice coil to compensate for physical spring forces in the spider plus surround. Measurements show this design technique improves linearity.
It's not possible to increase the turns density at the ends of the coil. What you can do is decrease the turns density in the middle of the coil. This is easy to do with modern coil winding machines.
The problem is that you are reducing the efficiency of the driver quite a bit to get a more linear BL curve. Rockford came up with this idea seven or eight years ago, but didn't bother to get a patent because you can get the same BL linearity improvement without the efficiency tradeoff, using other techniques. I know that the idea is now in the public domain.
The problem is that you are reducing the efficiency of the driver quite a bit to get a more linear BL curve. Rockford came up with this idea seven or eight years ago, but didn't bother to get a patent because you can get the same BL linearity improvement without the efficiency tradeoff, using other techniques. I know that the idea is now in the public domain.
XBL^2 simple and very graceful decision. It something an average between overhung and underhung design. Optimum for decisions when it is necessary to have the big displacement and thermal capacity, a variant for subwoofers, woofers and midbass drivers.
When displacement and capacity small also is required light moving mass underhung it is most preferable. Ideally for midrange and highfrequency drivers.
When displacement and capacity small also is required light moving mass underhung it is most preferable. Ideally for midrange and highfrequency drivers.
yes, TC Sounds has something they call LMS which i think is a superior technology to XBL
while both XBL and LMS produce very linear drivers LMS does so while letting the entire magnetic flux go through the voice coil. XBL on the other hand achieves extra linearity at the expense of letting some flux get away.
i have been using a 18" TC Sounds LMS driver in about 6 cubic feet sealed box for about half a year now and while i was initially super-pissed at TC Sounds for making me wait months on pre-order it was worth it. the driver is simply amazing.
i also like JBL's differential drive technology ... of course JBL is not particularly interested in DIY market so they keep their best drivers to themselves. the 18" driver with two 3" voice coils and internal neodymium magnet that they have on one of their powered subwoofers (selling for like 4 grand) i am sure is nice. problem is its not for sale.
i actually would rate the technologies as follows
Differential Drive > LMS > XBL
basically all three technologies significantly improve linearity but Differential Drive uses like 120% magnetic flux (passing it twice, through two voice coils), LMS uses 100% of the flux passing it all through the voice coil ... XBL only uses about 60% of the flux by passing some of it AROUND the voice coil and not through it.
i think those are clear advantages. my LMS driver weighs 76 lbs and any XBL driver would need to weigh 100 lbs to match its output. a differential drive unit could probably match it at about 50 lbs if using ferrite motor ... of course there are no such drivers on the market, thtas why i got my LMS.
while both XBL and LMS produce very linear drivers LMS does so while letting the entire magnetic flux go through the voice coil. XBL on the other hand achieves extra linearity at the expense of letting some flux get away.
i have been using a 18" TC Sounds LMS driver in about 6 cubic feet sealed box for about half a year now and while i was initially super-pissed at TC Sounds for making me wait months on pre-order it was worth it. the driver is simply amazing.
i also like JBL's differential drive technology ... of course JBL is not particularly interested in DIY market so they keep their best drivers to themselves. the 18" driver with two 3" voice coils and internal neodymium magnet that they have on one of their powered subwoofers (selling for like 4 grand) i am sure is nice. problem is its not for sale.
i actually would rate the technologies as follows
Differential Drive > LMS > XBL
basically all three technologies significantly improve linearity but Differential Drive uses like 120% magnetic flux (passing it twice, through two voice coils), LMS uses 100% of the flux passing it all through the voice coil ... XBL only uses about 60% of the flux by passing some of it AROUND the voice coil and not through it.
i think those are clear advantages. my LMS driver weighs 76 lbs and any XBL driver would need to weigh 100 lbs to match its output. a differential drive unit could probably match it at about 50 lbs if using ferrite motor ... of course there are no such drivers on the market, thtas why i got my LMS.
Michail Bugaria said:Disadvantages of LMS technology from TC Sounds appear the large moving mass and huge inductance of voice coil. Which limits their application only in the heavy and powerful pumps with the the large Xmax
TC effectively solve the inductance issue with the monstrous shorting sleeves that they use.
large moving mass is compensated for by proportionally large BL product.
when running a sine sweep i can hear the 18" LMS all the way up to 10khz and the response seems to be RISING all the way through the bass range, so neither inductance nor voice coil mass seem to be an issue.
in fact if i could change one thing about the LMS sub it would be the cone ... i think with enough power input it can simply fail. i havent heard of LMS line cones ripping apart but this line is new and it happened to some other TC subs with non-titanium cones.
of course if you do manage to rip the cone apart the TC will probably ship you a new top assembly for free ... or you might be able to order a custom unit with a titanium cone. but regardless, i think if the LMS has a weak spot it would be the cone.
goldyrathore said:
There are two easy methods to guarantee a purely linear and constant force on the vc in response to current within it.
1. Use a small length coil within a really linear field. (underhung). The difficulty lies in making sure the entire coil remains in the "good field" region of the gap. As well as making sure the vc field doesn't alter the iron's saturation aspect too much.
2. Use a long length coil within any field whatsoever (overhung). As long as the ends of the vc are outside of any significant gap fringe field, the drive force will remain linear and constant.
In case 1, it falls upon the designer to create a really good linear field that extends the full excursion in either direction. In case 2, the designer has to make the coil longer but the field uniformity in the gap is entirely irrelevant.....far simpler to do this and incur the loss and inductance, than it is to try to maintain gap linearity at 1.5 tesla. Iron is starting to saturate at those levels, and the magnetic design does not get any easier as a consequence.
Tradeoffs...the others spoke of those quite elegantly, I learned from them...
Cheers, John
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.