Jump to content

Evolution of and Replacements for the AR 11/12" Woofer


Pete B

Recommended Posts

The recent Klippel data for the Tonegen woofer offers some insight into the driver design and large signal performance.

It is interesting to look at the THD vs. frequency and level as is shown in the file AR3ATHD+N.PDF. This is a fairly reasonable plot, and shows the typical increase in THD as frequency decreases due to increased excursion and the BL and compliance non-linearity that goes along with it. This is typical for a driver that has not been highly optimized with it reaching 7% distortion at 30 Hz and 5W.

It should also be noted that as frequency increases and power decreases, within the piston range, we should expect the THD to continually decrease, but it does not, it seems to reach a lower limit of about .5% midband, this is something we should keep in mind. The increases at about 750, and 1.2 Khz are due to cone break-up modes as we leave the piston range.

We should next look at file 12N12100032.mht_w_laser.mht to see how the different components contribute to this distortion. The first two plots are compliance Cms and stiffness Kms versus x displacement. They are reciprocal measures of the same thing. There is very good symmetry for +/- x displacement but it is not very linear, the suspension is nearly twice as stiff at +/- 6mm as compared to the rest position. This is due to the spider stretching and will help to stop the motor at the ends of it's displacement, but it contributes to the distortion. Note that the derivative or slope of the curve is zero at x=0 which indicates that it is linear for small displacements.

The force factor or Bl(x) curve is not so pretty because this is a very primitive driver. The Bl peaks at almost 10 but at x= -3 mm, that's nearly half way through it's "linear" excursion. It has no small signal linear excursion due to this offset in the peak and this probably explains why the distortion does not continually decrease with decreased excursion. This can also be seen by the fact that the derivative or slope of this curve is not zero at x= 0. Indeed, the Bl varies from 10 to 9 over the small displacement of -2.5 to +2.5 mm, whereas the theoretical model (zero fringe field) suggests that the BL should be constant from -6 to +6 mm, and should fall by 50% at -12 and +12 mm. Instead we see a Bl of 9.5 at -6 mm and about 7 at +6 mm. We also see a Bl of about 7.5 at -12 mm and 3.5 at +12 mm, very poor symmetry.

It is important to look at the absolute limits of the mechanical Xmax, the point where things hit destructively and one can compute that the voice coil hits the back plate at about - 12mm where the motor should theoretically have half it's Bl strength, it actually has much more than half. I prefer this mechanical limit to be at least the point where the motor strength goes to zero which would be -18 mm for this driver, 24 mm would be even better to allow for overshoot.

Lces, Res, and Cmes are the mechanical components reflected into the primary circuit and are distorted due to the Bl curve. Le(x) is also interesting, this is the inductance of the voice coil and just as with an RF tuning slug, the inductance goes up as the VC moves in with more pole iron in the coil, and down as the VC moves out with less iron in the coil. This can also contribute to the midband distortion. All of these curves would ideally be a flat line. Shorting rings help reduce the inductance variation.

Fs(x) and Qms(x) have symmetry because they are not dependent on Bl, whereas Qes(x) and Qts(x) are distorted by the Bl curve. The driver has a Qts well under .5 at x=0, however it is nearly 2 at the positive limit of excursion. High Q systems tend to overshoot and this probably contributes to mechanical destruction of the driver.

There are many ways to improve a motor's linearity which were covered in this old paper: "Moving Coil Loudspeaker Topology as an Indicator of Linear Excursion Capability", by Mark R. Gander (JBL), JAES 1981 January/February. There are also many patents covering the subject.

The Adire Audio Extremis Bl curve:

http://www.adireaudio.com/Home/Images/eX6BL.gif

Pete B.

Link to comment
Share on other sites

  • Replies 124
  • Created
  • Last Reply

Top Posters In This Topic

Top Posters In This Topic

Posted Images

It is intersting to note that the Extremis is a 6.5" woofer with a mechanical Xmax limit of 20 mm one way, the AR is an 11/12" with 12.5 mm one way. I've used the term Xmax50 as the point where the motor has half of it's rest position Bl, this is about +/- 15 mm for the Extremis and theoretically +/- 12 for the AR. Measured for the AR is hard to define due to the asymmetry and mechanical Xmax limits.

I've stated that the AR woofer is a primitive driver and it's design features are very similar to the better stamped frame drivers made in the USA such as the Swan 305 or the Eclipse W1238R:

http://www.madisound.com/images/product/SWAN305.jpg

http://www.madisound.com/images/product/small/W1238R.jpg

One area where the AR woofer has been optimized is midrange response smoothness and cone breakup. Here's the distortion performance of a modern high priced 12" woofer (SC12NRT), note that the conditions are for 96 dB SPL in a 41 liter sealed enclosure. Note distortion peaks around 300 and 700 Hz. It should also be noted that this driver is probably intended for true subwoofer use below 150 Hz:

http://www.madisound.com/images/sc12nrthar...cdistortion.gif

Pete B.

Link to comment
Share on other sites

It is interesting to take a look at a modern driver such as the Peerless XLS:

http://www.d-s-t.com/link/main/tech/xls.jpg

from: http://www.d-s-t.com/link/main/tech/xls_intro.htm

Key from the DST site:

A. Aluminum Spacer

The aluminum spacer serves as heat sink for the coil to reduce power compression.

B. Black Anodized Voice Coil

The 2-inch 4-layer voice coil is wound on thick black anodized aluminum for improved heat dissipation.

C. Stacked Magnet System

The twin stacked magnetic system is optimized by FEA (Finite Element Analysis), to create a symmetric powerful magnetic field in the air gap, and provide space for the 44 mm max excursion of the voice coil.

D. Distortion Reduced Motor

A long 4 layer voice coil normally results in high self induction and impedance varying with excursion. Its many ampere turns react on the magnetic field in the air gap. These two main factors causing motor distortion in subwoofers are practically eliminated by the combined impact of the Aluminum Short Circuiting Ring (D) and the Aluminum Spacer (A) on the pole piece.

At the same time they both contribute as heat sinks for the voice coil, reducing power compression.

The result is unbelievable clean bass reproduction.

E. Vented Cone

To eliminate compression under the dust cap the cone is vented by 8 large holes.

This way the coil is cooled and there is no need for a bore in the pole piece.

F. Nomex® Spider

The spider is made of high tech Nomex® material. This material is chosen for its high rigidity and long term stability. The suspension will stay in shape for a very long time under heavy load.

G. Rubber Surround

The surround is made from SBR rubber because of the wide operating temperature, low creep and long term reliability.

H. Fiber CompositeCone

The cone is molded from a propriety air dried wood free pulp with a blend of Nomex®, Kevlar® and glass fibers bonded together by deep impregnation with polymers.

This creates an ultra stiff and relatively light cone that will stay stable even under very large sound pressures.

I. Rigid Cast Aluminum Basket

The rigid cast basket with an aerodynamic profile provides the necessary sturdy base for the magnet structure and suspension and allows for the 44 mm max excursion of the cone.

The spider is ventilated to achieve the lowest possible compression and allow air to flow freely to create a cooling effect for the voice coil.

Link to comment
Share on other sites

>The force factor or Bl(x) curve is not so pretty because this

>is a very primitive driver. The Bl peaks at almost 10 but at

>x= -3 mm, that's nearly half way through it's "linear"

>excursion. It has no small signal linear excursion due to

>this offset in the peak and this probably explains why the

>distortion does not continually decrease with decreased

>excursion.

This is why I previously asked about the orientation of the driver, it would be interesting to tilt the enclosure so that a bias is introduced into the rest position to bring the peak of the Bl curve to x=0, then retest the distortion.

Pete B.

Link to comment
Share on other sites

Another modern driver:

http://www.northcreekmusic.com/Image25.gif

From: http://www.northcreekmusic.com/NorthCreekKits.html

Note that while the Scan Speak motor is patented, there were similar patents much earlier than Scan Speak's.

"The SCAN-SPEAK 18W/8545SC is a woofer of remarkable design. First and foremost, the driver is designed for enormous linear excursion. This is achieved through SCAN-SPEAK's development of the patent-protected SD-1 motor structure, diagramed below right. Close inspection of the conventional motor (above) and SD-1 system (below) reveals the following:

The conventional motor structure voice coil length of 12mm or less is simply too short for reasonable output at low frequencies. Conventional top plate thicknesses of 6mm yields a peak-to-peak linear excursion of only 6mm (voice coil height minus air gap height = peak-to-peak linear excursion).

The SCAN-SPEAK voice coil length is 19mm, yielding a p-p linear excursion of a full 13mm, more than twice that of conventional woofers.

The conventional motor structure suffers from high second harmonic distortion at low frequencies due to an asymmetric magnetic field about the top plate; that is, there is more flux below the top plate than above.

The SCAN-SPEAK pole is extended well beyond the top plate, creating a symmetric magnetic field which greatly reduces low frequency second harmonic distortion.

In a conventional woofer, high amplitude excursion modulates the voice coil inductance:

outward travel decreases the amount of pole surrounded by the voice coil, reducing its inductance, where as inward travel increases the amount of pole surrounded by the voice coil, increasing its inductance. This is one reason why most woofer impedance curves do not look like a simple "resistor plus inductor" at high frequencies. This displacement-dependent voice coil inductance causes intermodulation distortion between high excursion low frequencies and higher frequency information, as well as creating frequency response aberrations by perpetually misterminating the crossover network. In addition, the inductance variations create a solenoid-type force between the voice coil and pole which introduces an offset to the coil rest position, increasing second harmonic distortion.

The SCAN-SPEAK voice coil always surrounds an equal amount of pole, eliminating this source of distortion. The SCAN-SPEAK impedance curve is retraceable independent of drive level.

In a conventional motor structure, current within the voice coil creates dynamic magnetic flux which modulates the motor field by partially demagnetizing and remagnetizing the section of the pole surrounded by the voice coil. This phenomenon creates both third and intermodulation distortion throughout the midrange, and is the other reason why conventional woofer impedance curves are nonlinear.

The SCAN-SPEAK motor structure is constructed with copper rings, called "Faraday rings," bonded to the pole. The Faraday rings create a sympathetic flux which exactly matches the voice coil flux, eliminating this source of distortion.

The concave cone-shaped, vented pole top eliminates reflections between the dust cap and pole while cooling the magnetic structure.

The SCAN-SPEAK 18W/8545 cone is constructed of carbon fiber loaded paper and coated with damping compound. Damped carbon fiber is an exceptional cone material because it combines low mass with incredible strength, therefore it is extremely rigid at low frequencies yet can be easily controlled by the damping compound at mid and high frequencies. The 18W/8545SC exhibits nearly theoretically perfect performance for a 7" driver - including the rising response and gentle turnover before roll off.

One can find far more detailed information about the causes and solutions of motor-induced distortions the AES publications Loudspeakers, Vol. 1 and Vol. 2 (60 E 42nd St, NY, NY 10165). Suffice to say that the engineers at SCAN-SPEAK have used every resource to develop and manufacture a motor structure which minimizes or eliminates the common second, third, and intermodulation distortions found in virtually all conventional loudspeaker drivers, resulting in a 7" woofer with exceptional low end performance even at very high volume levels while providing midrange purity rivaling many smaller drivers. The SCAN-SPEAK motor structure is unique and patent protected."

Link to comment
Share on other sites

  • 4 weeks later...

>

>This is why I previously asked about the orientation of the

>driver, it would be interesting to tilt the enclosure so that

>a bias is introduced into the rest position to bring the peak

>of the Bl curve to x=0, then retest the distortion.

>

>Pete B.

>

An obvious solution to bring the peak in the Bl curve to x=0 is to build the driver with the voice coil slightly offset.

It is interesting that this is exactly what Bowers and Wilkins does in their midrange drivers. From page four of their white paper:

"For these drivers, a different solution was used, based on the simple recognition that it is not the flux distribution itself that matters, so much as the force factor (Bl). Simply offsetting the rest postion of the coil significantly improves force symmetry (figure 7)."

B&W 700 series white paper, in .pdf format:

http://www.bwspeakers.com/downloadFile/tec...sWhitepaper.pdf

Pete B.

Link to comment
Share on other sites

  • 2 months later...

>

>

Dear CSP members, I just read for about an hour or more some of the threads on this site about the differences of the AR 12" woofer through out its history that I never read before, I had to stop reading as it was overwhelming. I am almost completely drained to read the vast amount of knowledge and expertise from such site notables as Pete B., Tom Tyson, Vern and countless others(respects to all) that I feel like a child commenting about world affairs in the UN General Assembly when I post. I can't believe the amount of knowledge and experience that some of this site's members have. I had to stop as I was being drained and getting a headache!

Although I'm beyond impressed, I also feel belittled by such a range of helpful and interesting people on this site. I also feel if nothing else, maybe I'll just try to be the court jester here and painfully try to slightly amuse some people here if nothing else, if that's at all be possible. Or probably a better suggestion would be to just sit back quitely and learn from people who know so much more than I do. I'll listen to my system and stay away from the site as being liable of making a complete ass of myself sounds like a better idea.

I want thank all of the wonderful contributors for sharing your vast individual experiences and 'know-how'. Reading this page has made me feel that I am very grateful and lucky to have found this site in the first place less than 2 years ago. I feel secure in knowing that I have the privilege of being in such powerful company.

May I extend my deepest thanks and personal happiness I feel to read all of your opinions.

Respectfully and thankfully, frankmarsi@verizon.net

Link to comment
Share on other sites

Hi Frank;

Thank you very much for your kind words, Frank.

I look forward to everyones input here, from all over the world, even if it is above me.

Every day is a learning process for us all.

I am really looking forward to your next way out in left field humour.

I haven't recovered from your last two yet.

You really had me in stitches.

Thank you, Frank.

Link to comment
Share on other sites

Hi Frank,

Thank you for commenting, and you probably know that if we were on a photography site the tables would be turned with you being the professional expert. No matter the case I do believe that it is good to keep in perspective that this is just HiFi after all, and your humor is welcome to lighten things up. We all need a laugh from time to time and I did enjoy your humerous posts.

That's quite a rig you've got there with stacked LSTs and Phase Linears. I enjoy hearing about your listening sessions and impressions.

We've talked about our common auto, photography, electronics, and sound interestests, and there are several others here that also have many of these same interests.

Thanks again for your kind words Frank, and take care!

Pete B.

>Dear CSP members, I just read for about an hour or more some

>of the threads on this site about the differences of the AR

>12" woofer through out its history that I never read

>before, I had to stop reading as it was overwhelming. I am

>almost completely drained to read the vast amount of knowledge

>and expertise from such site notables as Pete B., Tom Tyson,

>Vern and countless others(respects to all) that I feel like a

>child commenting about world affairs in the UN General

>Assembly when I post. I can't believe the amount of knowledge

>and experience that some of this site's members have. I had to

>stop as I was being drained and getting a headache!

>Although I'm beyond impressed, I also feel belittled by such a

>range of helpful and interesting people on this site. I also

>feel if nothing else, maybe I'll just try to be the court

>jester here and painfully try to slightly amuse some people

>here if nothing else, if that's at all be possible. Or

>probably a better suggestion would be to just sit back quitely

>and learn from people who know so much more than I do. I'll

>listen to my system and stay away from the site as being

>liable of making a complete ass of myself sounds like a better

>idea.

>I want thank all of the wonderful contributors for sharing

>your vast individual experiences and 'know-how'. Reading this

>page has made me feel that I am very grateful and lucky to

>have found this site in the first place less than 2 years ago.

>I feel secure in knowing that I have the privilege of being in

>such powerful company.

>May I extend my deepest thanks and personal happiness I feel

>to read all of your opinions.

>Respectfully and thankfully, frankmarsi@verizon.net

Link to comment
Share on other sites

  • 1 month later...

I've been suspicious that the SD data used in the measurements of Bret's drivers might be in error. I mentioned this to Ken Kantor and he posted this reply, thank you by the way, which I'm posting here since it impacts the measurements that were the basis of much of this discussion. Fortunately, the larger error was in the 303 woofer, which was not the main focus of the discussion.

>Pete,

>Per your email, I have measured the cone diameter of a 303a woofer

>and an NHT 1259. This was just a ruler/eyeball measurement, made

>from halfway across the surround roll. As such, it will differ

>somewhat from effective acoustical radiating area, and is also

>probably not reliable down to the mm. But, I did measure both at

>the same time, and so my guestimation error is probably close in

>both cases.

303a woofer:

diameter = 253mm

Sd = 502.7cm^2

NHT 1259 woofer:

diameter = 251mm

Sd = 494.8cm^2

>This conforms pretty well to the published data on the 1259, which

>I did not look at before I did the measurement. On the other hand,

>my recent measurement of the 303a Sd is much higher than what is

>shown in the Klippel report I have on my website. Though I did the

> response measurements and basic SSP's in this data package myself,

>I did not do the Klippel work. I will ask the tech who did the

>Klippeling how the Sd number was derived.

>-k

>Ken Kantor

Link to comment
Share on other sites

My reply:

Thanks Ken,

I have a 1259 here and I get about the same number measured as you do with a ruler half way across the surround.

I was concerned about the 303 Sd number because it is especially important when the test box method is used for the derived T&S parameters.

Here's what I measure for the 200003s:

Cone Diameter 1/2roll -1/2roll Sd

200003 8-1/2" 9-1/8" 422.0 cm^2

Most of the 200003s were listed as 450 cm^2 in the measurements of Bret's 200003s, not too different from 422 but it does effect the mass calculation when the test box method is used. The 303 woofer was listed as 416 cm^2 and I wonder if this should have been the 200003 value.

Thanks again for confirming and looking into this.

Pete B.

Link to comment
Share on other sites

People continue to write here and on the web, making statments as if there are no woofers made today that will provide bass like the AR 11/12" in the same size (small) enclosure - this is simply a false statement. They also make claims to the effect that no advancements have been made in modern woofers. Obviously, as I've shown above in this thread and the JBL papers I've refernced show that much work has been done to provide more linear motors.

The AR woofer operates well into the lower midrange whereas many of todays woofers are intended more for subwoofer duty, with much more Xmax.

One exception that includes several innovative design features is the Aura NS12-513-4A:

http://www.madisound.com/cgi-bin/index.cgi...8.7078&pid=2096

pdf file: http://www.madisound.com/pdf/aurasound/NS12-513-4A.pdf

http://www.madisound.com/images/product/NS12-513-4A.jpg

I'm not suggesting to swap this in any vintage AR, just offering it for discussion about design improvements and innovation.

Note that this woofer is a true 12" with a cone area of 503 cm^2, it also has an Xmax of 11 mm which is nearly twice that of the AR. The NS12's volume displacement (VD) is 1 cm * 503 = 503 cm^3 whereas the value for the AR is: .6 * 420 = 252 cm^3.

Nearly twice the VD of the AR, and this is not a large Xmax value by today's standards. It also has a mechanical Xmax limit of 22mm also nearly twice the AR's value of 12mm.

From Aura's marketing literature:

"feature Aura's patented Neo-Radial Technology (NRT) magnetic structure yielding unsurpassed linearity and exceptionally low distortion. The cones are made of aluminum alloy, which has a much higher stiffness-to-weight ratio than traditional paper or polypropylene cones. This allows the cones to resist deformation in small enclosures and to withstand the abuse of competition use. It also allows the cones to operate as true pistons at all power levels ..."

This driver has an underhung motor where rather than a long coil in a short gap, a long gap is used with a relatively short coil. This allows Xmax to be increased simply by using more magnet, the coil length does not have to be increased with the associated increase in voice coil inductance. The driver also uses shorting rings to lower voice coil inductance and the associated non-linearity.

This driver should have similar, likely better, low end small signal performance as compared to the AR. It's moving mass is 137 g combined with its higher surface area should provide a slightly lower Fc, similar to the AR-303, for deeper extension. The AR woofer is unusual in that it has lower than typical voice coil resistance and the Aura is similar with a value of 2.86 ohms.

The Aura is different having a stiff cone, which provides good piston band performance up to the point where the behavior is very different from a soft cone driver having a large peak just above 2kHz. This would be notched in a modern system, but it may not be necessary if a very low crossover point is chosen.

This seems to be an excellent driver based on specs, it would be important to see (Dumax or Klippel) and hear the large signal performance.

Pete B.

Link to comment
Share on other sites

These 12" Auras would be high on my list for evaluation if I were to do a modernized acoustic suspension 3-way system. However, they are rather expensive. While these would work fine in a 4-way, I'd probably choose a driver with even more Xmax for such systems.

Pete B.

Link to comment
Share on other sites

  • 2 years later...

Read thru and downloaded this thread today and was impressed with the volume of info. provided. However, many important details are no longer accessible. In particular, KK's testing of a driver provided by TT. KK's research assistant provided a few links in a few posts that now are dead ends (e.g. file no longer found).

KK: can this data be accessed again? Your fine report on the 3/4 inch tweeter testing project also is no longer accessible.

Just a suggestion here, but I've found numerous instances of dead end links to important information in threads going back a few years.

I think all important data should be embedded in a post to best preserve prosperity should the site be once again upgraded to another format by Mark.. Another thought would be for key authors to CC the library when making important posts. Perhaps that location is less vulnerable to 'evaporation' issues.

Link to comment
Share on other sites

  • 9 years later...

I agree Carl, it is unfortunate that there are so many dead links in this thread and

the Klippel data seems to be gone.  I see that your post is nearly 10 years old.

Link to comment
Share on other sites

  • 2 years later...

Arthur,

Be sure to download the AR-3a restoration guide http://www.classicspeakerpages.net/library/acoustic_research/original_models_1954-1974/original_models_schematicss/restoring_the_ar-3a/

It shows several iterations of the 12" driver on pages 27-28. Certainly the one in figure A.1 is not suitable. A.2, A.3, A.4, A.5 probably are fine. As Aadams said, post pics and if you can, note the model numbers. Also see Tom Tyson's comments on the first page of this thread.

Kent

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...