As other posters wrote or alluded to, the ideal audio driver material should be infinitely stiff, infinitesimally heavy, and infinitely lossy, all at the same time, which is of course impossible. Failing the above, the ideal material should at least be highly intelligent, and would be able to decide correctly that Vibration A was signal, and therefore should be reproduced faithfully, and that Vibration B was just some neighboring molecules misbehaving, which should be suppressed. This seems to be impossible as well. What we are left with is compromises. Just keep in mind that lossy cannot usually distinguish between Vibration A and Vibration B. It is up to the designer to keep those two kinds as separate as possible. This makes reproducing A and suppressing B more feasible.
Not really...however what you were stating bellow is correct but doesn't take into consideration the anti-phase component appearing lower in fr on soft cones_thus giving acoustical cancellation of higher level; thus total sound response is higher before Ft on hard cones (Ft beeing more or less equal to Fx in our case); please see the Klippel link (left low corner "Bandwidth of Rigid Radiators")_thanks again to 6.283 for this interesting link!_:
http://www.klippel.de/ddl/files/Poster~OU2tYNWrdo/KLIPPEL_Radiation_Vibration_Poster_Page_1.pdf
OK, so I decided I had better make some posts with more of my data as it may be useful to some DIYers.
audio blog: Cone Treatments Demonstrated To Work (Damping)
audio blog: Cone Treatments Demonstrated To Work (Damping II)
audio blog: Cone Treatments Demonstrated To Work (Stiffening)
audio blog: Cone Treatments Demonstrated To Work (Surround Damping)
audio blog: Cone Treatments Demonstrated To Work (Geometrical Stiffening)
audio blog: Cone Treatments Demonstrated To Work (Time Elapsed)
I still have more, but that's the essential stuff. It seems to work pretty much the same no matter the size of the paper cone to start with, but I'll put up the stuff on pro 12" mid/woofers soon.
Dan
audio blog: Cone Treatments Demonstrated To Work (Damping)
audio blog: Cone Treatments Demonstrated To Work (Damping II)
audio blog: Cone Treatments Demonstrated To Work (Stiffening)
audio blog: Cone Treatments Demonstrated To Work (Surround Damping)
audio blog: Cone Treatments Demonstrated To Work (Geometrical Stiffening)
audio blog: Cone Treatments Demonstrated To Work (Time Elapsed)
I still have more, but that's the essential stuff. It seems to work pretty much the same no matter the size of the paper cone to start with, but I'll put up the stuff on pro 12" mid/woofers soon.
Dan
You're most welcome Crazy.
Where to stop is a big question as is what to use. Those were just the first things I tried that looked like they'd make sense. One thing I noticed is that stuff that stays tacky damps well. The longevity of the stuff I used and how long it would remain as effective is in question. I looked really hard to buy the stuff that they put on pro audio driver surrounds, but couldn't find a source. That's what I'd look for to damp a speaker right now. Damping certainly never hurt in my experiments. Stiffening alone certainly did if full range response is desired.
Here's what I did on large 12" pro drivers: http://dtmblabber.blogspot.com/2011/08/large-driver-cone-treatement.html
The break up is much easier to deal with after heavy doping! There also seems to be a point to where no more is getting done really.
Dan
Where to stop is a big question as is what to use. Those were just the first things I tried that looked like they'd make sense. One thing I noticed is that stuff that stays tacky damps well. The longevity of the stuff I used and how long it would remain as effective is in question. I looked really hard to buy the stuff that they put on pro audio driver surrounds, but couldn't find a source. That's what I'd look for to damp a speaker right now. Damping certainly never hurt in my experiments. Stiffening alone certainly did if full range response is desired.
Here's what I did on large 12" pro drivers: http://dtmblabber.blogspot.com/2011/08/large-driver-cone-treatement.html
The break up is much easier to deal with after heavy doping! There also seems to be a point to where no more is getting done really.
Dan
Last edited:
Thanks for the link, very interesting read, however nothing on that page contradicts what I said ?
There may well be a small dip in the frequency response on a soft cone at the point where cone breakup begins, but the amount and significance of this dip is highly design specific, (it's a potential dip that doesn't necessarily manifest) and doesn't change the basic facts about directivity at high frequencies vs rigid cones.
For a given diameter a well designed soft cone is capable of going much higher in frequency on axis and also provide better off axis performance at high frequencies than a rigid cone that is operating below cone breakup, and that page demonstrates this quite nicely.
One problem with large rigid cones at high frequencies is that the depth of the cone causes on axis cancellation due to the cone geometry - the cone is not a coherent point source at higher frequencies where the cone depth is significant, even if it's still below mechanical cone breakup. (Also shown in the bottom left of that linked page)
On a soft cone with well controlled cone breakup the bending waves travel at a finite velocity from the voice coil to the edge of the cone which is faster than the speed of sound in air, but not infinite. If the taper of the cone is properly chosen so the forward vector of the bending wave propagation speed through the material in question is equal to the speed of sound in air, the cone can launch a fully coherent forward facing wave-front even at frequencies where the cone is several wavelengths in diameter, something that is not possible with a similar sized hard cone.
While hard cones have their advantages, they certainly cant compete with a well designed soft cone for either high frequency extension or high frequency off axis response for the same diameter of cone.
Thanks. if you think about it, this is what sandwiching or doping strive to achieve. A very rigid core to maintain signal transmission longitudinally along the cone profile, and lossy layers on top of it, to suppress propagation in a direction normal to the profile. The degree of success varies but this "forced anisotropicity" is in essence an attemp to endow a cone with intelligence, so to speak.That's an almost close to poetry astonishing expressed statement but I agree
There may well be a small dip in the frequency response on a soft cone at the point where cone breakup begins, but the amount and significance of this dip is highly design specific, (it's a potential dip that doesn't necessarily manifest) and doesn't change the basic facts about directivity at high frequencies vs rigid cones. As said previously the subject of this thread isn't in the ability of soft cones to increase high fr bandwidth but about audibility of the (yes very small) higher power response of rigid ones before Ft.
Indeed 4lexander believes it's audible. I'm middle of the pond so to say but in fact much closer to the bank of an unaudibility of such a little difference.
On a soft cone with well controlled cone breakup the bending waves travel at a finite velocity from the voice coil to the edge of the cone which is faster than the speed of sound in air, but not infinite. If the taper of the cone is properly chosen so the forward vector of the bending wave propagation speed through the material in question is equal to the speed of sound in air...One must be a very brilliant technician to define the proper geometry of a cone to be able to control the speed of shear waves, no?
It's not clear that there is a higher power response of the hard cone driver before Ft in this particular case. If you look very closely at the seas spec sheets 0, 30, 60 degree responses in the 2Khz - 3Khz region especially they don't agree with the results measured by 4lexander ? In fact the results are opposite as far showing which driver has better off axis performance. There may be too much measurement uncertainty in 4lexanders graphs to draw solid conclusions.crazyhub said:
You're looking at the problem back to front. The propagation speed of the bending waves above cone breakup is dictated by the materials property - thickness, stiffness, material composition etc.
You choose your cone material and properties first based on other criteria, and end up with a certain propagation speed. What the exact speed is doesn't matter, and you're not trying to adjust it - consider it a constant for a given cone material.
Only then do you determine the amount of taper (depth) of the cone required for coherent wave-front launch. (Directly forward component of the bending wave velocity matches the speed of sound in air, so radiation from all parts of the cone arrive at the listener at the same time) Too shallow or too deep will cause periodic ripples in the on axis treble response, so a few iterations of the cone shape will determine the optimal angle.
Although you could no doubt do it more quickly by computer analysis these days, driver designers were doing it iteratively decades before computer FEA was available, it just took them a little longer and a bit of trial and error.
Last edited:
Hmm, strange. Maybe I'm misreading them but they don't seem to correspond well. I'll check again.
Electronics for a while yes, but speakers and audio have only ever been a hobby for me, like most people here.
Thinking about how soft cones work at high frequencies comes from a lot of tweaking and playing around with full range drivers - drivers which push the limits of how high in frequency a largeish driver can go. (Sometimes 1, 2, or even 3 octaves higher than normal for a given size)
As an extreme example of a cone driver a full range driver relies on every trick in the book to work, including matching the propagation speed of bending waves to the cone angle. (At least on well designed ones!)
I wouldn't say it's rare knowledge, I've seen and read articles on it before.
Probably the best single article/white-paper I've seen on the design of a soft cone driver that goes into tons of detail is the section on the FST midrange driver in the following B&W Nautilus 801 design white-paper:
http://www.hifiportal.co.uk/Articles/Article0026-Development of the Nautilus 801.pdf
B&W's design goes beyond the traditional radially symmetric soft cone approach and they discuss why. Great reading for anyone interested in how soft cones really work.
There is also a brief albeit slightly marketing-speak summary with a nice little animation here:
Bowers & Wilkins - Kevlar
Thanks Simon for the explanations and link to 801 design white paper!
After having read all this stuff and after having listened to the 801... God! all this tech inside a speaker to get such an average listening result!
Note that's only my (short) point of view and not the sudject of this thread...
After having read all this stuff and after having listened to the 801... God! all this tech inside a speaker to get such an average listening result!
Note that's only my (short) point of view and not the sudject of this thread...
B&W 801 or B&W Nautilus 801 ? Two very different speakers
I've only heard the original version of the Nautilus 802 - the one with the 2x 8" woofers instead of 1x15", and the original metal dome tweeter. (Not the later diamond tweeter of the updated version)
I do like the midrange on them, very smooth, but as speakers they're not without their flaws overall. I wasn't too keen on the characteristic of the treble (not a fan of metal domes) and found the overall response a bit tipped up towards the treble and quite "clinical".
Although I think they're good quality, they're not quite my cup of tea in overall (sound) presentation or design philosophy...
I can't fault their analysis of the design of the midrange driver though, and the behaviour of the standing waves on the cone, which is what I find most interesting about the article.
I wouldn't mind betting that their decision to run the FST up to 4khz and with a simple crossover might help contribute to the tilted response and possible some of the treble quality.
I've experimented to a degree with the FST driver myself and it lives up to what you might expect considering the engineering presented in the white papers - it measures superbly (As Zaph also showed) and it sounds fantastic. It may have a very well controlled breakup region, but running it out to 4khz? Not for me.
It's a few years since I've read that white paper right through, (I only recently re-discovered it after losing the printed copy I had) but I don't recall whether they disclosed the crossover design in it ? I did hear rumours elsewhere that the midrange to tweeter crossover is 1st order, but have never seen a circuit diagram to substantiate that. Is it true its 1st order ?
Is the FST driver available to buy ? Or did you just manage to get your hands on one somehow
Given the extensive work they put into optimizing the high frequency and off axis response, I think 4Khz is entirely reasonable, (otherwise why go to all that trouble) but 1st order, if that's what it is, is not, and that could well be what I didn't like about the sound of the treble.
I cross over my 8" dual cone full rangers at 4Khz to a ribbon tweeter very successfully, but I've done a lot of cone damping tweaks to them affecting the 3-5Khz region, as well as using a 3rd order (electrical) crossover. I wouldn't even consider using 1st order as too much cone breakup and directionality in the high treble would show through. With the 3rd order (and possibly even better with 4th order, will try it one day) the result is very clean.
Perhaps the B&W's could have benefited from a higher order filter, if they really are 1st order...
PS do you have a link to Zaph's measurements ?
Last edited:
You can find the crossover schematics for B&Ws older (ie not current) ranges at the B&W group.
Service Manuals
They use first order electrical on the tweeter and 2nd order electrical on the FST.
Is the FST driver available to buy ? Or did you just manage to get your hands on one somehow
Well B&W sell replacement parts for their loudspeakers so technically anyone can buy the parts if they want. Again the B&W group for North America has a price list, but being in the UK myself I enquired directly to B&W about prices. Ordering a pair was very straight forward. The new Neo prices however more then doubled the price on some of the 800 series FST drivers in the last month or so.
Its frequency response might be well extended, but it's off axis performance isn't that amazing.
Well as I said before they do use 2nd order electrical on the FST. However if you look at Stereophile's review of the latest 800D2 you can quite clearly see the controlled breakup of the FST plastered over the upper octaves.
Here is a thread that I started on the HTguide forums after seeing Zaph's review. I have added a lot of my own measurements on the FST to the thread too. Zaph's full set of measurements is available on his blog.
Attachments
Brilliant! Thanks for that link.
Odd, because I just downloaded the circuit for the Nautilus 802 and the tweeter crossover appears to be 3rd order. (Unless they changed the crossover in later revisions ?)
I haven't studied the component values though, (which are on separate pages) but when I get a chance I'll throw the circuits into Esketch and see how they simulate.
Haha, I never thought of that
I guess I just assumed they would only sell spare parts to authorized dealers, or with some sort of proof of ownership of the model you're asking for a driver for...Yeah those measurements don't actually look as good as I expected to be honest, and I presume that the on axis upwards tilt in the response is corrected in their crossover.
One thing that springs to mind about the overall response though is I assume those measurements and the ones taken by Zaph were done on an infinite baffle, while in the Nautilus they're in a zero diffraction spherical enclosure, which will give somewhat different results. Perhaps the response of the driver has been tailored with the spherical enclosure in mind, and may require more equalization on a conventional baffle.
Yeah I've seen that stereophile review before - I'm not sure that it's the same model FST driver though, and I'm really surprised they opted not to correct the rather large bump at 3.5Khz, that would definitely be audible.
The tweeters response is surprisingly non flat too, if they measure similarly to the ones I've heard that could explain why I didn't like the treble.
I'm getting a considerably flatter response than that from 2Khz to 10Khz with my full range + ribbon tweeter, albeit with two notch filters on the FR, one about 1/3 octave -3dB at 2Khz, and the other 1/5 octave -2dB at 4Khz to smooth a couple of small resonances. After that its easily +/-2dB from 2Khz to 10Khz
Is it worth me going to the trouble to correct those two resonances ? Absolutely! The audible improvement is quite dramatic. I wonder what the 800D2 would have sounded like if they'd taken the trouble to correct the 3.5Khz bump ?
Thanks, interesting thread. I might have to sign up over there too.
Brilliant! Thanks for that link.
Odd, because I just downloaded the circuit for the Nautilus 802 and the tweeter crossover appears to be 3rd order. (Unless they changed the crossover in later revisions ?)
Indeed the tweeter does use a third order network with the nautilus versions of the 800 series. The 800D series use a first order though, the diamond must do something quite different then the alu dome did.
Also some of the older Nautilus speakers did use a more complex crossover then the current range does.
Haha, I never thought of that
You can buy direct from B&W for some of the (non current) parts providing you are in the right country etc. Being in the UK you can't really get any better
I'd been waiting for a mid-range like this to come along and after seeing Zaph's measurements I knew I had to get me a pair. I was half expecting B&W to request extra details but they never asked and were extremely helpful. I don't know what their stance would be if they knew a few people were buying their drivers for DIY purposes, but why not? A sale is a sale and it's great advertising too.I get the impression the the FST was designed to get the most performance out of a 'soft' cone driver without considerations as to what enclosure it would go in, except of course that the low end bandwidth would be very limited.
It wont be the same model of FST, in the original Nautilus versions I think they used ferrite magnets whereas the more recent versions swapped to the Neo magnet with a thicker top plate (underhung motor). I do believe however that the cone is probably very similar and would behave in a similar way.
I do have a review of the original 800D in a Hifi news mag and the frequency response of that model is far smoother then the 'D2' version.
Stereophile also reviewed the 802D and it's frequency response is noticeably smoother then the 800D2s. What's interesting to notice here is that the individual driver measurements are done near-field, so would be reasonably immune to diffraction effects and the more extreme parts of the cone breakup. Yet go far field for a proper response and things fall to bits a little with the resonances and diffraction effects altering things significantly and for the worse.
I've included a few of my own musings in that thread, but one thing that stands out is that the cone appears to have a bell mode type resonance at 3.5khz, which is responsible for the peak seen in the 800D2 (and yes I agree this will be audible). If the driver didn't have such a low distortion motor this peak would really set the upper limit on how high you'd want to take the driver - heck it still does but it's not really a 'problem' due to how good the motor is. I can understand B&Ws desire to keep the filters simple and not wanting to add more series connected components then is necessary, but the addition of parallel notch filters to flatten the response I'd have thought would be somewhat essential if neutrality is the goal. Of course I have speculated somewhat that running the FST up that high, with the somewhat 'benign issues' it has could be what's responsible for the B&W 'sound' and is something they want to keep.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.