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

[Pete B]

Thank you very much for this data, very nice work.

I'm not familiar with that part number, would you or Tom please provide some details about this driver. Is it a ceramic magnet driver? Early or late production? Or is this a Tonegen?

Pete B.

[tysontom]

>Hello everyone,

>

>This is Ruchi Goel, Assistant Research Engineer to Ken Kantor

>at Tymphany Corporation, USA. As Ken had mentioned in the

>discussion forums, I was given the task to measure AR 12"

>woofer (12N12100032) sent to him by Mr. Tom Tyson. I am

>delighted to say that I have finished the task and have the

>measurements ready to be posted. The link to the measurements

>is : http://aural.org/ar_hist/AR_OCT05/

>

>Kindly have a look at the data and I will be glad to do any

>other measurements requested by anyone.

>

>

>Thank you,

>

>Ruchi Goel

>Assistant Research Engineer

>Tymphany Corporation

>Ph: 408-200-3126

>

Hello everyone,

This is Ruchi Goel, Assistant Research Engineer to Ken Kantor at Tymphany Corporation, USA. As Ken had mentioned in the discussion forums, I was given the task to measure AR 12" woofer (12N12100032) sent to him by Mr. Tom Tyson. I am delighted to say that I have finished the task and have the measurements ready to be posted. The link to the measurements is : http://aural.org/ar_hist/AR_OCT05/

Kindly have a look at the data and I will be glad to do any other measurements requested by anyone.

Thank you,

Ruchi Goel

Assistant Research Engineer

Tymphany Corporation

Ph: 408-200-3126

Ruchi,

Thanks so much for performing these tests for us. Once again, we are greatly indebted to Ken and Ruchi for this work! Great job!

I sent a “test” enclosure, which was a standard 1.7 cu. ft. AR-3 box, and I set the crossover so that the woofer could be tested with or without the crossover coil in the circuit. Other parts of the crossover were disabled. After all, this is a test for the low-frequency characteristics of the various AR 12-inch woofers. The midrange and tweeter drivers were removed and the holes filled. I was very careful to make sure that the enclosure was acoustically air-tight, and the enclosure fiberglass was the “standard” weight and density for an AR-3 or early AR-3a, but this could be altered if needed.

If I remember correctly, I sent along with the enclosure (1) Tonegen PN12100032 woofer and (2) and early Alnico PN 3700 woofer, although I am not now sure I sent the latter. The one Ruchi tested would clearly have been the Tonegen, and you can see from the response curve that there is close correlation with 2Pi tests that AR did outdoors when testing the Alnico (PN 3700) and ferrite (200003/1210003) woofers. I am going to get an early AR-3a 200003-type woofer (with the proper 5/8-inch surround) out to Ruchi to test, and from that we can see how the different, mounted woofers perform.

Thanks again, Ruchi!

--Tom Tyson

[Guest rgoel]

Thank you Mr. Tyson,

Ken mentioned only one woofer (12N12100032), the one I tested. I will be happy to test the other one.

The tests I did were without the crossover coil in the circuit. The cabinet was really airtight with no air leaks, it was really a great job.

Looking forward to receiving the other woofer to test.

Have a great day!

Ruchi

[kkantor]

Just a quick note...

1- I only got one woofer with this shipment.

2- My site is hosted by Verio and Florida, and has been impacted by the hurricane. As such, people might find access to the data intermittent. I would suggest that Mark move the data over to CSP.

-k

[administrator]

>Just a quick note...

>

>1- I only got one woofer with this shipment.

>

>2- My site is hosted by Verio and Florida, and has been

>impacted by the hurricane. As such, people might find access

>to the data intermittent. I would suggest that Mark move the

>data over to CSP.

Hi Ken!

I will get the data up in the Library this weekend, or sooner if work permits. The only problem I seem to have over here with connectivity is when my housekeeper blows fuses or disconnects network cables when I'm in other countries. Like last week.

Mark

[tysontom]

>Just a quick note...

>

>1- I only got one woofer with this shipment.

>

>2- My site is hosted by Verio and Florida, and has been

>impacted by the hurricane. As such, people might find access

>to the data intermittent. I would suggest that Mark move the

>data over to CSP.

>

>-k

>

Ken,

You are correct. I only *thought* I sent two woofers. I actually shipped the Tonegen woofer with the cabinet, and I did not get around to shipping the early Alnico 3700 woofer. I will do that, in addition to a 200003 AR-3a ferrite woofer.

Sometimes I think I might be having short-term memory loss. Sometimes I think I might be having short-term memory loss.

--Tom Tyson

[Pete B]

Hi Ruchi,

I have a few questions if it's not too much trouble. The caption in the Freq Resp .pdf file reads: AR3a 12" Woofer, Re 2.83 Ohms ...

I take it that should be Volts?

So this was 1W into 8 ohms or the standard voltage sensitivity type of measurement? Further, I take it the mic was 5 cm from the cone or near field so that the environment is not an issue?

Are the 1W, 2W, and 5W figures in the distortion plots also referenced to 8 ohms, or Re of the driver, or?

I'd like to see 2nd and 3rd harmonic distortion level plotted since it gives an indication as to the source of the non-linearity, if it's not too much trouble in the future.

I notice that some of the parameters in the Klippel measurements are prefixed by (imported), Re and Mms for example, does this mean that they were measured outside of the current test. It looks like Mms was measured in the small signal section using the laser, then imported to the other tests. Am I reading this correctly?

One last request, I'm having trouble reading the large signal plots, it would help if they were 2 or 4 time larger again if it's not too much trouble in the future.

Thanks again for this effort,

Pete B.

[Pete B]

There's a lot of data to digest, but something jumped out at me in this data.

I suggested that the problem with the Tonegens might be a high Bl, yet this driver is slightly under 10, comparable to the lowest in the last group tested. Bl is not the issue. Fs is slighly high with this Tonegen but let's ignore this for the moment.

Question for the group here to stimulate conversation: There is one parameter that stands out to me as compared to the 200003 woofers tested, does anyone see it?

Pete B.

[Pete B]

Hi Ruchi,

I just wanted to ask, again if it's not too much trouble, for an in box input impedance plot to go along with the frequency response plot.

It would also help to know Fc and Qtc in box, with and without the crossover network.

Thanks again,

Pete B.

[Guest rgoel]

>Hi Ruchi,

>

>I have a few questions if it's not too much trouble. The

>caption in the Freq Resp .pdf file reads: AR3a 12" Woofer, Re

>2.83 Ohms ...

>I take it that should be Volts?

Hello,

I have the habit of writing DCR of the woofer in the comments whenever I am doing nay measurements, so Re - 2.83 Ohms is the DCR of the woofer.

>So this was 1W into 8 ohms or the standard voltage sensitivity

>type of measurement? Further, I take it the mic was 5 cm from

>the cone or near field so that the environment is not an

>issue?

The reference used was 1 W in 8 Ohms. And, the microphone was indeed 5 cm from the center of the cone.

>

>Are the 1W, 2W, and 5W figures in the distortion plots also

>referenced to 8 ohms, or Re of the driver, or?

Yes, the distortion plots are also referenced to 8 Ohms.

>I'd like to see 2nd and 3rd harmonic distortion level plotted

>since it gives an indication as to the source of the

>non-linearity, if it's not too much trouble in the future.

I will definitely try to get those plots as soon as I can. No problem at all.

>

>I notice that some of the parameters in the Klippel

>measurements are prefixed by (imported), Re and Mms for

>example, does this mean that they were measured outside of the

>current test. It looks like Mms was measured in the small

>signal section using the laser, then imported to the other

>tests. Am I reading this correctly?

>

On Klippel, I do small signal test to list the Thiele-Small parameters. For large signal tests, Bl, Re and Mms are imported. You are definitley reading it right.

>One last request, I'm having trouble reading the large signal

>plots, it would help if they were 2 or 4 time larger again if

>it's not too much trouble in the future.

>

I will try if the plots can be saved at a different resolution.

Thank you

Ruchi

[Guest rgoel]

>Hi Ruchi,

>

>I just wanted to ask, again if it's not too much trouble, for

>an in box input impedance plot to go along with the frequency

>response plot.

>It would also help to know Fc and Qtc in box, with and without

>the crossover network.

>

>Thanks again,

>Pete B.

Hello,

I will try to get the impedance plot as soon as I can.

Thank you

Ruchi

[Pete B]

Thanks very much Ruchi, look forward to seeing that data.

Pete B.

[Pete B]

Hi Ruchi,

I was just wondering if you tested the woofer face up, down, or in a vertical position?

Pete B.

[Guest rgoel]

>Hi Ruchi,

>

>I was just wondering if you tested the woofer face up, down,

>or in a vertical position?

>

>Pete B.

Hello

The inbox measurements were done in the vertical position and Klippel measurements were done with the woofer mounted in the Klippel stand and laser pointing to center of the woofer.

Kindly let me know if that answers your question.

Thank you

Ruchi

[tysontom]

>>Hi Ruchi,

>>

>>I was just wondering if you tested the woofer face up, down,

>>or in a vertical position?

>>

>>Pete B.

>

>Hello

>

>The inbox measurements were done in the vertical position and

>Klippel measurements were done with the woofer mounted in the

>Klippel stand and laser pointing to center of the woofer.

>

>Kindly let me know if that answers your question.

>Thank you

>Ruchi

Pete,

With the woofer mounted in the enclosure, it probably makes no difference whatsoever how the enclosure is pointed. Even in free air it probably doesn't matter that much, but I would think that vertical mounting, such as the Ken's Klippel setup below, would be the norm.

http://www.classicspeakerpages.net/dc/user_files/677.jpg

Klippel test jig with AR 12-inch woofer (Ken Kantor)

--Tom Tyson

[Pete B]

I believe I've seen pictures of the Klippel stand with the driver in a vertical orientation and if this is correct then yes it answers my question and thank you once again.

I think Tom Tyson mentioned that this driver had the correct 5/8" foam and I read this to mean that the driver was refoamed. Tom, anyone, would you please verify this?

It would be very interesting to see Klippel data for a new Tonegen and also the NHT1259/Tonegen as a baseline comparison. I'm just thinking out loud and don't expect you to do this.

Thanks,

Pete B.

Edit: I did not see Tom's post above as we were writing at the same time, so yes I see it is vertical.

[Pete B]

Just wondering if this picture is of the current woofer that we have test data for: Tonegen 12N12100032. It doesn't look like the Tonegen currently being sold, rather it looks like a 200003 type.

Pete B.

[Pete B]

Tom,

I ask this specifically because the Klippel data shows significant voice coil offset or magnetic field asymmetry. This is to be expected if they assemble the driver assuming that the magnetic field is symmetric when in reality it is not. It is hard to tell if this is a manufacturing tolerance issue which I would not expect from Tonegen, or if it is due to assumptions made about the symmetry of the magnetic field and the center position chosen for the voice coil.

An even better solution is a simple mod to the voice coil that can be used to compensate for an asymmetric magnetic field.

Pete B.

>With the woofer mounted in the enclosure, it probably makes no

>difference whatsoever how the enclosure is pointed. Even in

>free air it probably doesn't matter that much, but I would

>think that vertical mounting, such as the Ken's Klippel setup

>below, would be the norm.

>

>

>--Tom Tyson

>

[tysontom]

>Just wondering if this picture is of the current woofer that

>we have test data for: Tonegen 12N12100032. It doesn't look

>like the Tonegen currently being sold, rather it looks like a

>200003 type.

>

>Pete B.

Pete,

I'm not sure who supplied that AR woofer in Ken's Klippel test jig, but I can tell that it was an earlier 200003 or 1210003 factory woofer that had been repaired with a new surround. It might have been one of Bret's woofers; I'm not sure. It is not the Tonegen version.

Incidentally, I plan to send (to Ken) an earlier AR-3/AR-3a PN 3700 Alnico woofer and an early AR-3a ARNP 200003-0 woofer recently refurbished with a new 5/8-inch surround. Regarding surrounds, I use Speaker Works Northwest surround kits with satisfactory results, although I am sure there are numerous other good suppliers of the proper surround. These Speaker Works surrounds are similar to samples I received from Minh Luong a few years ago that had the proper 5/8-inch half-round size and the angled-inner flange.

http://www.speakerworks.net/

Description: "11-inch Acoustic Research Angle Attach Surround Kit." Price is $24.00 per set, with a small discount for larger numbers. Unfortunately, the kit is incorrectly labelled as an "11-inch" surround, but I guess it's too much trouble to change the description at this point. Despite the incorrect description, the surround kit works very well in my experience. Shims are not supplied nor are they recommended, and the dust cap is left alone. Slow-drying glue is used, similar to white glue, and this method works the best in my opinion.

--Tom Tyson

[Guest Bret]

>It might have been one of Bret's woofers; I'm not sure.<

Tom -

You can be sure. This was my "woofer in bondage," a 1978 200003.

This is a Simply Speakers refoamed 200003 woofer from an AR-9 manufactured circa '78.

It is one of the two drivers which were refoamed with the spider-sag, meaning that the speaker had almost no rearward "throw." The fact that this was not any more obvious than it was when the data was recorded almost makes me wonder about our testing methods. These speakers had "no chance" of sounding "right" or testing correctly.

They are listed as "Edge-It" in Ken's spreadsheet of data on the six drivers I sent him.

This driver in this configuration no longer exists. This one, and its mate, were sent to Millersound, they were un-repaired and then repaired correctly - cured of spider-sag. They are happily at home in the 9s from which they came.

Bret

[Pete B]

This picture, from an ebay sale, has Fs 18 Hz written on the back. I wonder if it was from the time of manufacture, my guess is not since CPS was in common use not Hz, and the Villchur patent uses the term Fr for free air resonance not Fs which is a more recent term:

http://www.unclepapa.com/AR1Woofer_11.JPG

These drivers have accordion edges, which I don't believe is correct:

http://www.unclepapa.com/AR1Woofer_2.JPG

I believe that these woofers have been worked on and the 18 Hz figure was measured after the repair.

[Pete B]

I posed a question about the new data for the Tonegen driver:

http://www.classicspeakerpages.net/dc/dcbo..._id=&page=#7359

The parameter that stands out is moving mass which for this Tonegen is about 68 grams, whereas the older 200003's measured the last time were around 110 grams. I have a 200003 replacement woofer with the cone, voice coil, and spider removed as a unit. I weighed it some time back as 67.5 grams on a scale noting the difference from the other units tested here. I believe AR moved to a lower moving mass in later years and this, along with Bl differences are the reasons for the lack of "growl" from the newer woofers.

Some here state that we will never know the sound of the older AR systems which is not true since they've been documented in the journals and AR's own literature. "The Sound Field in Home Listening Rooms" by Allison and Berkovitz, AES July/August 1972 gives many measurements of an AR-3a. The authors write, "is 6 dB down at 30 Hz exactly what an AR-3a woofer is supposed to do.", note that this is half space.

An NHT document here lists 33 Hz as the -6 dB point for an AR-3a and 24.5 Hz for the AR303, anechoic. An AR303 woofer was measured in the last group of drivers and had a moving mass of 139 grams.

We see large differences in moving mass which correspond to large differences in F3/F6 and efficiency as can be seen below:

I've added the Tonegen to the list:

------ moving mass (Mms)

DUT2 103 g (AR-9 1978, new surround by Simply Speakers)

DUT3 104.36 g (AR-9 1978, new surround by Simply Speakers)

DUT4 123.95 g (AR-9 1978)

DUT5 107 g (early 10pi soft cone, surround, spider by Tri-State)

DUT6 115.16 g (early 10pi soft cone, surround, spider, VC by Tri-State)

DUT7 164.88 g (mass added by reconer) (AR-9 1978)

DUT1 139g (actually a 303 woofer)

Tonegen 68.1 g

200003 67.5 g (Replacement driver, Cone stamped 3013F on back)

------ Bl ------- Ref Eff

DUT2 11.76 Tm .572 %

DUT3 11.78 Tm .531 %

DUT4 11.59 Tm .358 % (estimated used RDC = 2.5 ohms)

DUT5 10.10 Tm .391 %

DUT6 9.89 Tm .319 %

DUT7 13 Tm .289 % (was the voice coil replaced?)

DUT1 12.94 Tm .320 % (actually a 303 woofer)

Tonegen 9.52 Tm .731 %

------ Qms

DUT2 2.89

DUT3 3.11

DUT4 2.82

DUT5 7.92

DUT6 7.89

DUT7 9.3 (reconned)

DUT1 5.02 (actually a 303 woofer)

Tonegen 3.53

I believe that there is hope for the Tonegen woofers and adding mass is probably the way to duplicate the deeper extension of the older drivers. Adding mass will lower Fs which indicates that the spider is probably not the issue, this is also supported by the 161 l Vas value. One has to be cautious about voice coil heating and midrange break up modes when adding mass, and therefore it is not so simple.

Pete B.

[tysontom]

>

>These drivers have accordion edges, which I don't believe is

>correct:

>

The term "Fs 18Hz" is definitely written by someone after the woofers were manufactured because of "Fs" and "Hz." Neither of these terms was used in the 1950s. The 18 Hz value is correct for this woofer.

The first AR woofers with sand-cast frames also did have "pleated" surrounds, which was not shown in Villchur's patent. His patent shows the half-round surround, but Henry Kloss felt the pleated surround would work better since it was the conventional method of the day. Therefore, Villchur decided to compare the harmonic distortion of woofers with both the pleated surround and the half-round surround, and the once that had lower distortion would set the rule. After the initial production run of AR-1s and AR-1Ws, the pleated surround was changed to half-round.

http://www.classicspeakerpages.net/dc/user_files/679.jpg

Very large file on AR-1 Ser. No. 0006. Shows the pleated surround. Skiver masonite has separated, but underneath the hand-made voice coil can be seen.

>I believe that these woofers have been worked on and the 18 Hz

>figure was measured after the repair.

The surrounds have been treated with some material, but the pleated surrounds are original.

--Tom Tyson

[Pete B]

>The term "Fs 18Hz" is definitely written by someone after the

>woofers were manufactured because of "Fs" and "Hz." Neither

>of these terms was used in the 1950s. The 18 Hz value is

>correct for this woofer.

I believe that Fs went down with the half roll edge as I was told, by the audio engineering professor that I studied under, that it was in the low teens for the very early woofers. I'm not sure if this low figure was what it should have been according to the patent or actual, a figure of 11 Hz was mentioned, and I've also heard the 14 Hz figure. This professor was critical as a scientist of several other areas/issues concerning AR.

>The first AR woofers with sand-cast frames also did have

>"pleated" surrounds, which was not shown in Villchur's patent.

> His patent shows the half-round surround, but Henry Kloss

>felt the pleated surround would work better since it was the

>conventional method of the day. Therefore, Villchur decided

>to compare the harmonic distortion of woofers with both the

>pleated surround and the half-round surround, and the once

>that had lower distortion would set the rule. After the

>initial production run of AR-1s and AR-1Ws, the pleated

>surround was changed to half-round.

I've read this somewhere, probably your writing Tom but had the impression that they made the comparison in the prototype phase before shipping product, but yes it makes sense that this was a production unit. I wondered, after posting above, where a repair shop would get the pleated surround for the non-standard cone and also thought that it looked original.

>The surrounds have been treated with some material, but the

>pleated surrounds are original.

Yes, this makes sense.

Pete B.

[Pete B]

>I'm not sure who supplied that AR woofer in Ken's Klippel test

>jig, but I can tell that it was an earlier 200003 or 1210003

>factory woofer that had been repaired with a new surround. It

>might have been one of Bret's woofers; I'm not sure. It is

>not the Tonegen version.

Ah, I had the impression that Bret's woofers were not tested on a Klippel system probably because I only noticed the small signal data, I have not looked at the data for some time. This makes sense just never made the connection with the old drivers.

>Incidentally, I plan to send (to Ken) an earlier AR-3/AR-3a PN

>3700 Alnico woofer and an early AR-3a ARNP 200003-0 woofer

>recently refurbished with a new 5/8-inch surround.

Should be interesting to see the results.

>Regarding

>surrounds, I use Speaker Works Northwest surround kits with

>satisfactory results, although I am sure there are numerous

>other good suppliers of the proper surround. These Speaker

>Works surrounds are similar to samples I received from Minh

>Luong a few years ago that had the proper 5/8-inch half-round

>size and the angled-inner flange.

Good to have the feedback on this supplier.

>http://www.speakerworks.net/

>

>Description: "11-inch Acoustic Research Angle Attach Surround

>Kit." Price is $24.00 per set, with a small discount for

>larger numbers. Unfortunately, the kit is incorrectly

>labelled as an "11-inch" surround, but I guess it's too much

>trouble to change the description at this point. Despite the

>incorrect description, the surround kit works very well in my

>experience. Shims are not supplied nor are they recommended,

>and the dust cap is left alone. Slow-drying glue is used,

>similar to white glue, and this method works the best in my

>opinion.

>

>--Tom Tyson

Yes this works well in my experience also.

Pete B.