Lynn Olson said:
Paper is still paper in most ways. The biggest change since back then is for lighter, stiffer cones. The stiffer the cone, the higher in frequency they will play. This gives more efficiency and as a result more breakup. In the pro audio world, more output is always considered better, even if that output comes at the expense of crazy cone breakup. The other issue is that inductive rise makes the drivers electrically less efficient at the higher frequencies. Rather than have a natural rolloff though, they rely on the crazy cone breakup to extend the response. Most all pro audio 6.5" drivers had these issues. The B&C, Beyma, 18sound, etc, all rely on the cone breakup to extend their response. If you want to use these 6.5" drivers to 3Khz you are into the range where the cone breakup is providing a lot of the output. This resonance rarely ever sounds good.
In contrast with the 6.5" driver we are doing, we have a full copper sleeve on the pole with tiny inductance of .081mH. The impedance is quite flat to 10KHz. Combining this motor with the standard paper pulp cone gives output to nearly 10KHz, but the 4-10KHz range is quite a bit higher than desired. What we did was go back to one of the old school type paper pulps. It is based on the Kapok fiber, and not standard wood pulp. This Kapok is much softer and doesn't have nearly the breakup or high frequency extension, but with the low inductance motor it is a great fit. It extends very cleanly to 5KHz with virtually no breakup above there.
The main goal IMO is that if there is going to be a cone resonance/breakup, you to make sure the response is -48dB by that point. In the case of these custom Dipole 15's, they use the TD15M cone/surround assembly. They will have close to the same response characteristics of the TD15M. They should be quite flat out to 4KHz and any cone breakup is extremely well damped. This is a function of the treatment and also of the cone profile used as well. I think you'll find they are much more controlled on the top end than any other 15" driver you could find. The same would go for the 12" or 10" versions.
LineSource said:
These things like carbon fiber, kevlar, fiberglass, etc are all things you can add to get a higher strength/mass ratio. This is great for the pro audio drivers where squeezing out every last bit of output is always good, even at the expense of harsh breakups. The problem is though that none of these things add good damping characteristics. For better damping you actually want softer fibers in most cases. Roger Russell from McIntosh had told Nick that the best midrange material he had ever heard was the Kapok. This is what we experimented with in our 6.5". We found it to be much more well damped than the standard pulps of today.
We do have some 18" drivers in the works. We'll be doing a TD18 H, X, S, and M, the same as all the other TD woofers. The difference is that we needed to go up to a 2.5" VC and larger motor. As Lynn mentions, the mass goes up quickly on the larger cones, so we need more Bl to compensate. The good thing is that with the full copper sleeve on the pole, inductance is still quite low. Our 4ohm nominal 23mm Xmax AV woofers have only about .28mH inductance. For a 16mm Xmax 18" driver expect Le of about .35mH on the 2.5" coil. The larger heavier cone in this case is quite well damped as well. They should be quite well controlled up to 1KHz so use up to 250Hz or higher is quite easy.
John
gedlee said:
Using my spreadsheet you can see that you end up with at least a 18" driver for the bandwidth of 100Hz to 900Hz if you want to stay below 1% D-IM – at a relatively low 100dB max SPL limit !
For a 15"er the max SPL limit for that bandwidth and distortion drops to roughly 95dB.
Its not only the upper XO point that counts in terms of smooth driver FR - but also the bandwidth you are using that driver for – if you take distortion into account.
Michael
gedlee said:
Well - I know, Earl - but you may have noticed by all my simus / measurements and by what JohnK outlined some 1000 postings back - that driver built quality doesn't influence D-IM at all - nor does any of the specs you mentioned.
You simply can't fight D-IM and even more so BDMD – except with enlarging Sd or lowering bandwidth as outlined in a footnote in my spreadsheet.
Michael
Probably the best way to determine of IM distortion is going to be audible is to simply test it. Play a 20hz tone at high excursion and then play the highest frequency you want to use the driver at. If it's not audible at high excursions, it won't be at lower excursions either.
We need to look at 2 different things. One is the distortion/modulation that comes from the core of the coil changing with excursion and flux in the gap moving. As inductance changes with position so does the high end response. The higher the inductance, the greater the effect. Often times a driver with say 3mH inductance at rest with fully iron core will have 1mH or less inductance on extreme outward stroke as it becomes mainly air core. This greatly increases the high frequency output on the outward stroke vs the inward stroke. This is a quite audible effect. Extended pole pieces do help, but you also need to look at the outer "core" of the top plate around the coil as well. As the coil is fully within the gap the core is the same. As it begins to move out the core the inductance goes down. It's not just the pole that creates the effects, otherwise an extended pole would solve the issue. The extended pole also does nothing for the effects of flux modulation. Lowering overall inductance puts these issues out of the intended usage range of the driver where they are of no effect.
Another issue is the doppler effect. This is essentially the same effect you hear when a car or plane goes by as the sound changes. This is based on the distance changes of the cone. The only way to lower the effect is to lower excursion of the driver so the effect is lower.
In most drivers, the issue you hear is the first relating to changing inductance and high frequency response, much more so than the second. The problem is that virtually no drivers have this linear inductance, meaning most drivers suffer from both issues.
John
We need to look at 2 different things. One is the distortion/modulation that comes from the core of the coil changing with excursion and flux in the gap moving. As inductance changes with position so does the high end response. The higher the inductance, the greater the effect. Often times a driver with say 3mH inductance at rest with fully iron core will have 1mH or less inductance on extreme outward stroke as it becomes mainly air core. This greatly increases the high frequency output on the outward stroke vs the inward stroke. This is a quite audible effect. Extended pole pieces do help, but you also need to look at the outer "core" of the top plate around the coil as well. As the coil is fully within the gap the core is the same. As it begins to move out the core the inductance goes down. It's not just the pole that creates the effects, otherwise an extended pole would solve the issue. The extended pole also does nothing for the effects of flux modulation. Lowering overall inductance puts these issues out of the intended usage range of the driver where they are of no effect.
Another issue is the doppler effect. This is essentially the same effect you hear when a car or plane goes by as the sound changes. This is based on the distance changes of the cone. The only way to lower the effect is to lower excursion of the driver so the effect is lower.
In most drivers, the issue you hear is the first relating to changing inductance and high frequency response, much more so than the second. The problem is that virtually no drivers have this linear inductance, meaning most drivers suffer from both issues.
John
mige0 said:
We've built a few. Jeff Bagby just received one and will be getting some info back to me soon. He is planning on using it instead of the PHL for the new Salk open baffle midrange speakers.
Here is the news release page on the 6.5" and there is a link to our forum with parameters and curves at the bottom.
http://www.aespeakers.com/news.php?start=0&news_id=16
John
gedlee said:
I know that you refuse to accept this kind of distortion as to be audible.
Maybe out of the same reason you find active concepts "just some more (un-useful) toys to play with".
For all those that don't want to go with your "scientific conclusions" I suggest to put their wet finger on the input of their mid or high amp to introduce just some slight mains hum into the signal.
What a joy to see some more veils dropping, vanishing stability of image, detail and beauty of presentation – no?
Michael
John is in agreement with me here, and I agree with him. I DID say that I assume a copper shorting ring so that flux modulation and inductance changes are negligable. That leaves only doppler distortion and that is very small and very low order so its not an issue. I would never use a driver over a large bandwidth that didn't have a shorting ring, to me that's just a given. In a sub I don't care either way, the bandwidth is too low.
And your test proves what?
You can believe what you want, I'm simply stating what decades of research (and design) into the subject have taught me.
And your test proves what?
You can believe what you want, I'm simply stating what decades of research (and design) into the subject have taught me.
John_E_Janowitz said:
Thanks for the infos. Good looking driver !
What about the price and availability? If not too high I'd probably like to compare to my existing 6.5"
- Telling from pix its not an under hung design.
- For my usage as a pure mid from 300Hz up a X-lin-max p-p of roughly 2mm would be more than enough – raising sensitivity by some more dB – witch is always welcome.
- is your 3.5mm Xmax meant as excursion or p-p ?
- From pix the spider isn't vented?
- I wouldn't mind to get slightly higher FS or Qts but not really a need.
Could you do that as a customer version as well ?
Michael
gedlee said:
I agree - thats a lot of interesting and useful scientifically backuped information you are providing around here - *but* - sometimes you wet finger can tell you even more - just try
And I am *only* taking about Doppler IM and BDMD (back diaphragm mirror distortion) here, so all your other points are mute (as much as they count in real world – *on top* of D-IM and BDMD)
Michael
PS
BTW -its not the first time I have proven you wrong, Earl - remember ?
gedlee said:
I agree with you 100%. 12 and 15-inch professional drivers have excellent midranges, if you select wisely, and the measurements are there to back it up - frequency response, IM distortion, cumulative decay, you name it. There's a good reason that this approach dates back to the Jim Lansing Iconic of the late Thirties - if you use similar cone materials and overall sizes, the natural crossover frequencies will fall in the same place, thanks to inherent properties of the materials. Much of what a speaker does comes down to little more than the acoustical properties of the emissive diaphragm, the phase-plug and waveguide (if any), and the skill of the designer in system integration.
If a speaker must be very small, a 5" or 6" midbass makes sense. The Ariel is basically an enlarged minimonitor with enhanced bass (50 Hz instead of the 80 Hz the drivers would give in conventional cabinets). Been there, done that.
If dynamics are a priority, a 5" or 6" conventional 87 dB/metre midbass is going to be severely taxed, even as a pure midrange, since it is covering the range that is close to the energy center of the musical spectrum. This will be most audible on piano music, where the music just won't have the heft, power, and sheer intensity of the real thing. Full-on choral music is the same, which is why some exhibitors at the RMAF chased me out of the room when I played it on their systems. All it takes is a dense spectrum with a lot of coherent musical information to unmask the defects of conventional audiophile speakers.
P.S. Thanks, John, especially for the fascinating comments about Kapok vs modern techno-wonder cone additives. I've been wondering for a long time why modern speakers seem to have much worse-sounding cones than the old-timers, and now I know - they've been optimizing for rigidity in the bass region (and more power, of course) at the expense of good sound in the upper midrange.
mige0 said:
Whether certain aspects are audible or not is very dependent on whether they are masked by other defficiencies. I don't think it makes sense to argue the audibility of this kind of distortion unless it's associated with a specific driver. If he does not have any analysis and measurement data available, he is just waving.
Shorting rings create additional ringing which is very audible in full range driver application.gedlee said:
Doppler effect, due to it's transient nature in normal music, is normally masked by release of stored energy. If you have drivers that decay fast, these doppler effects will gradually become more audible.
I am quite interested though how you you would describe the audible effects of flux modulation and without?
soongsc said:
I'm not quite sure what you are referring to here. This really doesn't make any sense. A shorting ring done properly lowers overall inductance, linearizes it, keeps the flux from modulating, and helps to pull heat from the coil. Unless it is loose and physically vibrating in the driver I don't know how you could have any "ringing" added.
If you actually look CSD plots from an identical driver with and without shorting ring, you'll actually see more drawn out issues without the shorting ring.
John
In one specific application, two drivers identical except that one had a shorting ring and one did not. Yes, the one with shorting ring did have lower inductance, and did show a rise above 4KHz or so, but if one looks at the impulse, it was clear that there was increased ringing, 18KHz if I remember correctly. The result in listening was cymbals had a very unnatural trailing timbre.
In another case I just took a driver with an aluminum former and used conductive paint to close the loop. Sure enough, there was that same ringing but much smaller in amplitude. Pretty much no change in the SPL. And yes, even this can be heard, which was quite unpleasant.
So my current conclusion is if one needs shorting ring to increase SPL in higher frequencies, it is much better to use equalization.
At what frequencies do you see the "drawn out issues"?
In another case I just took a driver with an aluminum former and used conductive paint to close the loop. Sure enough, there was that same ringing but much smaller in amplitude. Pretty much no change in the SPL. And yes, even this can be heard, which was quite unpleasant.
So my current conclusion is if one needs shorting ring to increase SPL in higher frequencies, it is much better to use equalization.
At what frequencies do you see the "drawn out issues"?
these papers very clearly explain benefits of shorting rings
http://www.klippel.de/pubs/papers.asp
http://www.klippel.de/pubs/papers.asp
I probably have read most if not all the paper published by Mr. Klippel. He was actually here in September. Yes, shorting rings do have benefits, but they may also create other issues. In low frequency drivers, probably below 1KHz, they may be good. However, I have not heard a full range driver that sounded good with shorting rings. And when I look at the data, it was quite evident where the problems were.
Listmania
I gathered up some specs of wide range, midrange, midbass, and bass drivers, and put into an Excel spread sheet.
In which, BL/mms shows the ability of cone acceleration, and then times Sd, thus reflecting the ability of "pumping air":
Brand`````Model```BL<TM>``Mms<kg>``Sd<m2>``BL/Mms*Sd
--------------------------------------------------------------------------------
18Sound```8M400```12.20````0.014`````0.0220````19.17
````````10NDA610``20.30````0.030`````0.0350````23.68
````````12ND610```24.00````0.049`````0.0531````26.01
````````12ND520```14.40````0.036`````0.0531````21.24
````````15MB700```17.60````0.073`````0.0850````20.49
AE``````TD15M`````17.00````0.070`````0.0855````20.76
Audax```PR170Z0```10.00````0.006`````0.0165````26.61
AN``````Super8`` ``11.02````0.012`````0.0220````21.01
````````Super 12````8.99````0.028`````0.0510````16.26
B&C`````8PE21`````16.60````0.018`````0.0220````20.29
`````````6PEV13````8.20````0.006`````0.0132````18.04
`````````12PE32````19.10```0.037`````0.0522````26.95
Beyma```8MI100`````9.80````0.017`````0.0220````12.68
`````````12MI100```21.60````0.047`````0.0530```24.36
`````````605Nd````13.15````0.010`````0.0140````18.41
`````````102Nd````25.40````0.033`````0.0380````29.25
`````````112Nd````24.20````0.045`````0.0530````28.50
`````````15KX`````14.50````0.070`````0.0880````18.23
Lowther```DX4`````10.54````0.009``````0.0211````23.58
Eminence`Delta 8```14.10````0.019``````0.0220````16.07
`````````Delta 15``14.50````0.061``````0.0856````20.35
`````````Sigma 18``22.10````0.130`````0.1140`````19.38
EV```````DL10X````20.40````0.037`````0.0370`````20.35
`````````DL12X````20.00````0.043`````0.0500`````23.47
`````````DL15X````20.00````0.076`````0.0860`````22.69
`````````DL18X````20.00````0.093`````0.1300`````28.08
`````````DL18W````22.50````0.197`````0.1300`````14.85
Focal````Audiom7k```8.94`````0.007`````0.0165`````20.21
Fostex```FE206E````11.82````0.015`````0.0220`````16.94
`````````FF225K````11.82````0.017`````0.0220`````15.03
PHL``````2520``````14.30````0.018`````0.0238`````18.91
`````````2530``````20.90````0.018`````0.0241`````27.68
RCF`````MR8N301```17.80````0.010`````0.0210`````37.38
````````MR10N301```20.70```0.032`````0.0350`````22.78
````````MB15N401```26.10```0.100`````0.0855`````22.32
PAS`````ER-1508C```21.80````0.068`````0.0790`````25.33
Selenium``10MB3P````13.60```0.033`````0.0363`````15.05
``````````15PW3````12.30````0.071````0.0681`````11.76
`````````WPU1507```20.40````0.120````0.0814`````13.81
`````````18WS600```19.20````0.123````0.1194`````18.56
`````````WPU1807```20.80````0.161````0.1194`````15.42
TAD``````TM-1201```26.00````0.060`````0.0531````23.01
`````````TL-1102````13.50````0.041`````0.0366````11.96
`````````TL-1601a```20.50````0.117`````0.0881````15.44
`````````TL-1801````21.00````0.158`````0.1220````16.22
Among them, I found some confusions. Some drivers with very good reputation show numbers which are not impressive at all. I've heard some of them and consider them very good indeed (like TAD 11" & 16" above).
The "top 1" of the list above is RCF MR8N301, with its strong motor and super light cone, shows magnificent potential. However its seal back structure makes the frequency range so narrow -- some compression drivers dive deeper than this! So, even though its number of [acceleration factor*Sd] is so high, it's still useless in delivering midbass power.
Problem in frequency response also happens in EV DL18X. Look at the numbers, it's extremely good. Its cone is so light for an 18"er. But maybe because of this, it has a 10dB hump center at about 2kHz and spreading a very broad 3.5 Oct !! It's very likely a severe breakup mentioned earlier by John. And it's odd that its decent motor, very light cone and big Sd do not contribute very much in the midbass region.
And it'd be misleading sometimes as the power requirements are not shown. As can be seen in PHL, 2520 is 8Ohm version while 2530 is 16Ohm. Under the same "power" input, I guess their performance should be very close (or simply identical) in reality.
Beside those "exceptions", I found drivers with most powerful pumping ability (on paper) are:
Beyma: 102Nd, 112Nd
18Sound: 12ND610
B&C: 12PE32
I haven't had chance to hear any of them However I believe they are all excellent.
BUT, how about those 15"ers? Why don't they shine here? Except PAS ER-1508C, none can catch up with the top notch 10" & 12". I must be missing something here.
Should I put in other factors? Any comments?
I gathered up some specs of wide range, midrange, midbass, and bass drivers, and put into an Excel spread sheet.
In which, BL/mms shows the ability of cone acceleration, and then times Sd, thus reflecting the ability of "pumping air":
Brand`````Model```BL<TM>``Mms<kg>``Sd<m2>``BL/Mms*Sd
--------------------------------------------------------------------------------
18Sound```8M400```12.20````0.014`````0.0220````19.17
````````10NDA610``20.30````0.030`````0.0350````23.68
````````12ND610```24.00````0.049`````0.0531````26.01
````````12ND520```14.40````0.036`````0.0531````21.24
````````15MB700```17.60````0.073`````0.0850````20.49
AE``````TD15M`````17.00````0.070`````0.0855````20.76
Audax```PR170Z0```10.00````0.006`````0.0165````26.61
AN``````Super8`` ``11.02````0.012`````0.0220````21.01
````````Super 12````8.99````0.028`````0.0510````16.26
B&C`````8PE21`````16.60````0.018`````0.0220````20.29
`````````6PEV13````8.20````0.006`````0.0132````18.04
`````````12PE32````19.10```0.037`````0.0522````26.95
Beyma```8MI100`````9.80````0.017`````0.0220````12.68
`````````12MI100```21.60````0.047`````0.0530```24.36
`````````605Nd````13.15````0.010`````0.0140````18.41
`````````102Nd````25.40````0.033`````0.0380````29.25
`````````112Nd````24.20````0.045`````0.0530````28.50
`````````15KX`````14.50````0.070`````0.0880````18.23
Lowther```DX4`````10.54````0.009``````0.0211````23.58
Eminence`Delta 8```14.10````0.019``````0.0220````16.07
`````````Delta 15``14.50````0.061``````0.0856````20.35
`````````Sigma 18``22.10````0.130`````0.1140`````19.38
EV```````DL10X````20.40````0.037`````0.0370`````20.35
`````````DL12X````20.00````0.043`````0.0500`````23.47
`````````DL15X````20.00````0.076`````0.0860`````22.69
`````````DL18X````20.00````0.093`````0.1300`````28.08
`````````DL18W````22.50````0.197`````0.1300`````14.85
Focal````Audiom7k```8.94`````0.007`````0.0165`````20.21
Fostex```FE206E````11.82````0.015`````0.0220`````16.94
`````````FF225K````11.82````0.017`````0.0220`````15.03
PHL``````2520``````14.30````0.018`````0.0238`````18.91
`````````2530``````20.90````0.018`````0.0241`````27.68
RCF`````MR8N301```17.80````0.010`````0.0210`````37.38
````````MR10N301```20.70```0.032`````0.0350`````22.78
````````MB15N401```26.10```0.100`````0.0855`````22.32
PAS`````ER-1508C```21.80````0.068`````0.0790`````25.33
Selenium``10MB3P````13.60```0.033`````0.0363`````15.05
``````````15PW3````12.30````0.071````0.0681`````11.76
`````````WPU1507```20.40````0.120````0.0814`````13.81
`````````18WS600```19.20````0.123````0.1194`````18.56
`````````WPU1807```20.80````0.161````0.1194`````15.42
TAD``````TM-1201```26.00````0.060`````0.0531````23.01
`````````TL-1102````13.50````0.041`````0.0366````11.96
`````````TL-1601a```20.50````0.117`````0.0881````15.44
`````````TL-1801````21.00````0.158`````0.1220````16.22
Among them, I found some confusions. Some drivers with very good reputation show numbers which are not impressive at all. I've heard some of them and consider them very good indeed (like TAD 11" & 16" above).
The "top 1" of the list above is RCF MR8N301, with its strong motor and super light cone, shows magnificent potential. However its seal back structure makes the frequency range so narrow -- some compression drivers dive deeper than this! So, even though its number of [acceleration factor*Sd] is so high, it's still useless in delivering midbass power.
Problem in frequency response also happens in EV DL18X. Look at the numbers, it's extremely good. Its cone is so light for an 18"er. But maybe because of this, it has a 10dB hump center at about 2kHz and spreading a very broad 3.5 Oct !! It's very likely a severe breakup mentioned earlier by John. And it's odd that its decent motor, very light cone and big Sd do not contribute very much in the midbass region.
And it'd be misleading sometimes as the power requirements are not shown. As can be seen in PHL, 2520 is 8Ohm version while 2530 is 16Ohm. Under the same "power" input, I guess their performance should be very close (or simply identical) in reality.
Beside those "exceptions", I found drivers with most powerful pumping ability (on paper) are:
Beyma: 102Nd, 112Nd
18Sound: 12ND610
B&C: 12PE32
I haven't had chance to hear any of them
However I believe they are all excellent.BUT, how about those 15"ers? Why don't they shine here? Except PAS ER-1508C, none can catch up with the top notch 10" & 12". I must be missing something here.
Should I put in other factors? Any comments?