Acoustic Research Crossover Inductors (Classic)

[RoyC]

Whoa, great info as always, Tom!!!!

This would be a good time to mention again that the debate regarding changing out caps DOES NOT apply to the inductors. The 17ga inductor wire provides a specific resistance to the circuit, which is required to maintain the original sound. Changing them out for fancy low impedance coils of any type is asking for trouble. In fact, the thicker the coil wire (lower gauge number) the worse the speakers are likely to sound.

I have found a couple of defective AR coils along the way, apparently caused by deformed plastic bobbins. It is preferable to rewind them than to buy new ones with the wrong size wire (good luck, finding 17 ga inductors).

Roy

[johnieo]

Tom:

Thank you for this very useful document! It is dated May 8, 1974, therefore speakers out of, or not yet into production by that date would not be listed.

The x-o drawings for the AR-1, -1W, and -3 list a woofer coil with the notation <143 turns>. This is the first document that I have seen, which designates an AR coil number (#3) for that critter.

The early AR-4x used the #4 inductor (0.88 mH); so it would appear that it was changed to the #5 (1.187 mH) coil before mid-1974.

I believe the AR-2ax also used two inductors, but that x-o change was due to a change in tweeter by mid-'74.

We see two sets of inductors for the AR-11, one in shaded typeface. The AR-11B used the other three coils. Since the AR-11 was just being introduced at that time, does this mean the shaded set was used in the AR-11A (brass nameplate)?

Someone opening an AR-2xa or -4x cabinet might be confused, unless they reconciled the date of manufacture of their unit with the date on the AR-published data sheet.

We discussed earlier the fact that many early AR-3a cabinets contain an AR #7 coil. If anyone finds this coil in their 3a, they should replace it with the correct #9. Who knows how the #7 got there, but it doesn't agree with information on any early AR-3a document. It is obvious from modeling and listening tests (mostly by Roy C) that the #7 is incorrect.

Thanks again for the most useful chart!!!

[Carlspeak]

Tom. Is there some way to shrink down the size of your inductor table? It's enormous on my PC. I also opened the .jpg attachment and got the same image, only 1/2 of which I can print.

Carl

Carl's Custom Loudspeakers

[dynaco_dan]

>Tom. Is there some way to shrink down the size of your

>inductor table? It's enormous on my PC. I also opened the .jpg

>attachment and got the same image, only 1/2 of which I can

>print.

>

>

>Carl

>Carl's Custom Loudspeakers

Hi Carl;

I was having the same problem.

I saved the file to my hard drive.

I use a free utility called Irfanview.

When I went to my saved file, I moused it, Irfanview came up and it was still too large to see it all.

I then went up to that programs toolbar, and reduced it down to a manageable size and turned it right 90 degrees and printed out on a 8 1/2 x 11 page, no problemo.

Thank you Tom for some more valuable archival data.

Good luck, Carl.

[tysontom]

>Tom. Is there some way to shrink down the size of your

>inductor table? It's enormous on my PC. I also opened the .jpg

>attachment and got the same image, only 1/2 of which I can

>print.

>

>

>Carl

>Carl's Custom Loudspeakers

Carl,

I have reduced the file size significantly, but the resolution has dimished as well -- the catch-22 with files downloaded onto this server. I've tried to compromise somewhat, but see if this one is more manageable. If you put your cursor on the picture, the "Save-Print-Mail-Open-My-Pictures" symbol box should display in the upper left corner of the image. Otherwise, you can right-click on the picture and do a "Save-Picture-As" or "Print Picture." Be sure to go into your pull-down menu under File... Page Setup and go ahead and set your printer to Landscape mode. I would also save the file to your My Pictures folder on your computer's drive, and then you can (usually, depending on your image program) modify the image once you have it in your system. In any event, I have reduced the size of the image below:

http://www.classicspeakerpages.net/dc/user_files/939.jpg

--Tom Tyson

[Pete B]

I'm going to have to disagree here based on simulations of early impedance and frequency response measurements from AR. The only way to get agreement with the early data was to use a 1.9 mH (Coil #7, rated 1.88 mH) inductor. That there were two different inductor values is also suggested by the fact that one set of measurements shows the 3a woofer as "flat" under half space measurement, and later measurements show it being "flat" into full space. Roy Allison's own published measurements with the same AR-3a show that this is not possible for the same speaker as I discussed here:

http://www.classicspeakerpages.net/dc/dcbo...&mode=full#6202

Allison discusses in that paper design trends regarding 2pi and 4pi system analysis and I believe the thinking in this paper was the basis for the change in inductor value.

I'd say leave the #7 inductor if people want their system to be historically correct.

Pete B.

>

>We discussed earlier the fact that many early AR-3a cabinets

>contain an AR #7 coil. If anyone finds this coil in their 3a,

>they should replace it with the correct #9. Who knows how the

>#7 got there, but it doesn't agree with information on any

>early AR-3a document. It is obvious from modeling and

>listening tests (mostly by Roy C) that the #7 is incorrect.

[Steve F]

I still have serious doubts as to whether AR ever incorporated BSC into the 3a's design. If they did, they never mentioned it--and one would have thought that AR would have said something, since it would have been worth talking about from a sales/marketing standpoint.

In any event, my response in that thread #6845 http://www.classicspeakerpages.net/dc/dcbo...&mode=full#6202

still makes the most sense to me:

"It’s questionable whether the 3a ever incorporated BSC into its design. If it did, AR never mentioned it in any of their post-1973 literature.

Roy Allison’s independent research in 1972-93 identified and quantified many of these issues for the first time in the industry. The 3a never was presented to the public by AR as addressing these issues.

Of course, it’s possible that the 2.85 mH inductor change to the 3a was meant as some sort of EQ for that effect. AR’s marketing during that time period was so incredibly inept that it is not unthinkable that the engineers could have incorporated such a change into the design but the Sales/Marketing departments didn’t know how to publicize it properly.

Personally, I doubt it. I still think that the 3a’s basic design approach was unchanged, and that any component/measurement changes over the years were simply the result of the normal migration of parts used and assembly methods employed—as always happens with any item that is sourced and manufactured for a long time at multiple locations.

But it’s possible—I certainly have no proof either way, only a hunch.

Perhaps Tom might have some internal strategy documents that would shed some light on this."

Steve F.

[Pete B]

BSC is a term that was not commonly used in those days or in any of the Allison literature IIRC. The change might have been marketed as the "Truth in Listening" slogan - going from memory here as far as that phrase goes.

I would not be surprised if they down played the change to allow old customers to assume that they were the same, and new ones to hear them in the show room with more bass.

It's possible that the change came after Allison left where the AR engineers continued to follow his publications.

You write:

>Personally, I doubt it. I still think that the 3a’s basic

>design approach was unchanged, and that any

>component/measurement changes over the years were simply the

>result of the normal migration of parts used and assembly

>methods employed—as always happens with any item that is

>sourced and manufactured for a long time at multiple

>locations.

>

>But it’s possible—I certainly have no proof either way, only a

>hunch.

My point is that the design approach did change, early on they designed for "flat" into half space, later for "flat" into full space, this is a significant change and AR's own measurements support this theory. There is no guessing here, it is right there in the data. My simulations further support the theory.

Pete B.

[RoyC]

Hi Guys,

Since I am part of the great AR-3a coil conspiracy investigation, I figured I should check in.

First of all we have documented around 20 cabinets that have #7 inductors and alnico woofers. These serial numbers range from 07xxx to 25,xxx. What are the odds of NEVER finding a 3a containing an alnico woofer served by a #9 inductor, if #9 was THE inductor and #7 was the exception to the rule?! I also have personal experience with many more than 20 3a's containing the ceramic magnet woofer with the expected #9...but not ONE with a #7.

The answer to me is very clear. #7 inductors are in the minority because 3a's utilizing alnico magnet woofers are in the minority. #7 inductors and alnico magnet woofers were mated...period. When the ceramics appeared in the 3a so did the #9 inductor, the combination that comprises the majority of all AR-3a's in existence.

The second question would then be, why did AR switch to the #9 coil, and for that matter, reduce the amount of fiberglass to 20oz when the ceramic woofer was introduced?

I will ultimately defer that question to those who know their way around the engineering end of things better than I do.

I have a number of the older AR woofers (sorry Pete, nothing from the AR-11 era) that I have been testing using a program called Woofer Tester II. I will share data with anyone who would like to have it. It shows that the alnico woofer is not exactly the same having a lower Q than the ceramic magnet 71 and 72 woofers.

In listening tests I found that the two versions of 3a's sound somewhat different from each other. The alnico/#7 combo has a more forward midrange and the ceramic/#9 combo more "bassy" with a more reticent midrange.

It is tough to believe that the existence of the #7 coil had anything to do with parts availability issues. We have found evidence of other "evolutionary changes" in other models as well. They do not seem haphazard at all..yet are probably not as well documented as we would have hoped for.

Speculation is no substitution for braving the fiberglass!:-)

Respectfully,

Roy

[Pete B]

Roy your findings are not inconsistent with my theory. It is possible that the effort was a redesign, and that the #9 coil, new woofer, and measurement into full space were all part of the new design.

It seems to me that #7 coils and ALNICO magnet woofers are in the minority because they are older and fewer survived or were made.

The coil changed mainly because of measurement differences, the fiberglass changed to obtain the target Qtc with the new woofer.

A speaker designed for half space works better against a wall, and one for full space a few feet out into the room. I believe that AR changed their measurement to better match the typical in room response. This was covered in Allison's AES paper where an average of 20 different speakers of the day were measured based on typical use. The curve showed a drooping response, and a dip around 200 Hz, the new #9 coil compensated for the droop.

Pete B.

>First of all we have documented around 20 cabinets that have

>#7 inductors and alnico woofers. These serial numbers range

>from 07xxx to 25,xxx. What are the odds of NEVER finding a 3a

>containing an alnico woofer served by a #9 inductor, if #9 was

>THE inductor and #7 was the exception to the rule?! I also

>have personal experience with many more than 20 3a's

>containing the ceramic magnet woofer with the expected

>#9...but not ONE with a #7.

>

>The answer to me is very clear. #7 inductors are in the

>minority because 3a's utilizing alnico magnet woofers are in

>the minority. #7 inductors and alnico magnet woofers were

>mated...period. When the ceramics appeared in the 3a so did

>the #9 inductor, the combination that comprises the majority

>of all AR-3a's in existence.

>

>The second question would then be, why did AR switch to the #9

>coil, and for that matter, reduce the amount of fiberglass to

>20oz when the ceramic woofer was introduced?

[Steve F]

"The change might have been marketed as the "Truth in Listening" slogan - going from memory here as far as that phrase goes.

I would not be surprised if they down played the change to allow old customers to assume that they were the same, and new ones to hear them in the show room with more bass."

The Truth in Listening phrase applied to the ADD Series, not the 3a.

And remember, by 1973, AR had virtually no "showroom" sales. They were not a retail-oriented company, and none of the major stereo chains at the time (Tech Hi Fi, Tweeter, Pacific Stereo, Atlantis Sound, etc.) favored AR at retail. Quite the contrary--none of these retailers carried AR, they only had them on display to disparage them and sell different brands.

Virtually all of AR's sales were mail order--Baltimore Stereo Wholesalers, Illinois Audio, etc. This needs to be kept in mind as we try to evaluate 30 years after the fact why certain engineering changes were made. "Getting more bass in the showroom" wasn't a reason. For AR in the early-mid '70's there was no "showroom."

Marketing and engineering are joined at the hip, and all product changes--in any industry--have to be looked at in that undeniable light.

Steve F.

[RoyC]

>Roy your findings are not inconsistent with my theory. It is

>possible that the effort was a redesign

Pete,

That certainly makes the most sense to me.

Although I have fewer examples, the trend toward larger coils and less stuffing in the AR-4x and AR-2ax is also evident in the early 70's. Later in the mid 70's, AR increased the impedance of the AR-2ax midrange driver by 3 ohms+/- to coincide with the introduction of a ceramic magnet woofer for that model, with no crossover component changes. I wonder if that is documented anywhere.

It may have been a bit difficult for AR to market an "AR-3b" with the "new" ceramic magnet woofer in 1970-ish, when the old one was selling so well.

Roy

[Guest Bret]

Roy, thank you for all the hand washings and itchy nights.

>The answer to me is very clear. #7 inductors are in the minority because 3a's utilizing alnico magnet woofers are in the minority. #7 inductors and alnico magnet woofers were mated...period.<

The evidence is compelling.

>The second question would then be, why did AR switch to the #9 coil, and for that matter, reduce the amount of fiberglass to 20oz when the ceramic woofer was introduced? Letting the woofer operate so high was causing it to be directional and the only way to combat that was by lowering Q into the basement (and I have no idea how or if that works, I’ve just read several places that it does)

Somewhere in all this the crossover specs changed; remember? The old specs crossed the woofer at 575 and the new at 525. We’ve been all around that issue before and opined that this change may not have happened, but been declared for marketing purposes. Well, maybe we are focusing-in on an explanation and it did happen.

The ceramic woofers have higher Q, the stuffing was reduced bringing Q up additionally, and the crossover frequency was lowered. If Ken’s measurement of .71 on a 3a, combined with the literature on the 10pi and 11 models showing Q of .7 mean anything, it would make sense that at some point AR sacrificed extension and changed the meeting of the mid and woofer (#9 inductor) in order to get the woofer to operate as flatly as possible from its crossover frequency to Fc, which just "resulted" in less extension below Fc -and that was deemed acceptable.

Perhaps this was done in concert with the introduction of the ceramic because the ceramic woofer, over-stuffed and crossed-over higher, produced a frequency response “hump” that the Alnico woofer had not exhibited; or alternatively was rolling-off at an unacceptable rate due to the higher Fs of the ceramic driver.

>In listening tests I found that the two versions of 3a's sound somewhat different from each other. The alnico/#7 combo has a more forward midrange and the ceramic/#9 combo more "bassy" with a more reticent midrange.<

Right - the higher Q would *not* have emphasized the bass to Fc, but it would have made it flatter. There would be more bass close to Fc, but this would have been a “correction” and not an attempt to emphasize. At the same time, lowering the crossover frequency and raising the Q (making it less omni-directional) could be heard, comparatively, as having built a “hole” in the crossover region.

>It is tough to believe that the existence of the #7 coil had anything to do with parts availability issues. We have found evidence of other "evolutionary changes" in other models as well. They do not seem haphazard at all..yet are probably not as well documented as we would have hoped for.<

Pure speculation on my part - It “feels” like when AR changed to the ceramic woofer they had to adjust the crossover because the Alnico and ceramic woofers were so different in their Q and Fs. Whatever the original design goal was for the #7 inductor, 30oz of fiberglass, and alnico woofer had to be re-thought when AR was faced with this new woofer. Perhaps the guys opted for better response from the woofer (read: flatter) sacrificing some extension at the bottom and some output at the top (causing the crossover frequency to drop).

Maybe? What did I miss? What's your take?

>Speculation is no substitution for braving the fiberglass<

I couldn’t agree more. Thanks for sharing your observations based on so many examples.

Bret

[Pete B]

>The Truth in Listening phrase applied to the ADD Series, not

>the 3a.

OK thanks didn't know that.

>And remember, by 1973, AR had virtually no "showroom" sales.

>They were not a retail-oriented company, and none of the major

>stereo chains at the time (Tech Hi Fi, Tweeter, Pacific

>Stereo, Atlantis Sound, etc.) favored AR at retail. Quite the

>contrary--none of these retailers carried AR, they only had

>them on display to disparage them and sell different brands.

You seem to be generalizing, I saw ARs in show rooms in 1974, and friends bought them from show rooms in the late 70's. It is probably true that mail order was a larger percentage.

>Virtually all of AR's sales were mail order--Baltimore Stereo

>Wholesalers, Illinois Audio, etc. This needs to be kept in

>mind as we try to evaluate 30 years after the fact why certain

>engineering changes were made. "Getting more bass in the

>showroom" wasn't a reason. For AR in the early-mid '70's there

>was no "showroom."

I don't want to argue with you about history, but I believe there were at least limited show rooms in the late 70s.

>Marketing and engineering are joined at the hip, and all

>product changes--in any industry--have to be looked at in that

>undeniable light.

>

>Steve F.

As I've stated before the measurement facts from AR are there to see, the only way to come to another conclusion is to misread them. I've offered my interpretation, seems Roy made the discovery and has done fine work documenting his findings.

Pete B.

[Pete B]

>It may have been a bit difficult for AR to market an "AR-3b"

>with the "new" ceramic magnet woofer in 1970-ish, when the old

>one was selling so well.

>

>Roy

Yes, I agree this all makes sense. Do you know what year the switch from ALNICO/#7 to ceramic/#9 took place?

Pete B.

[RoyC]

Pete,

Tom can probably be more exact, but I'm guessing the switch away from the alnico woofer occurred somewhere around 1970 or late 1969 with a serial number around 30,xxx.

Roy

[Rlowe]

Here's an Excel copy.

956.zip

[tysontom]

>I'm going to have to disagree here based on simulations of

>early impedance and frequency response measurements from AR.

>The only way to get agreement with the early data was to use a

>1.9 mH (Coil #7, rated 1.88 mH) inductor. That there were two

>different inductor values is also suggested by the fact that

>one set of measurements shows the 3a woofer as "flat" under

>half space measurement, and later measurements show it being

>"flat" into full space. Roy Allison's own published

>measurements with the same AR-3a show that this is not

>possible for the same speaker as I discussed here:

>

>http://www.classicspeakerpages.net/dc/dcbo...&mode=full#6202

>

>Allison discusses in that paper design trends regarding 2pi

>and 4pi system analysis and I believe the thinking in this

>paper was the basis for the change in inductor value.

>

>I'd say leave the #7 inductor if people want their system to

>be historically correct.

>

>Pete B.

>

>>

>>We discussed earlier the fact that many early AR-3a cabinets

>>contain an AR #7 coil. If anyone finds this coil in their

>3a,

>>they should replace it with the correct #9. Who knows how

>the

>>#7 got there, but it doesn't agree with information on any

>>early AR-3a document. It is obvious from modeling and

>>listening tests (mostly by Roy C) that the #7 is incorrect.

>

This is basically nonsense. There is little data to support the idea that AR, as a standard, used anything other than the #9, 2.85 mH coil in the AR-3a. Yes, some #7 (AR-2ax) coils have appeared and have been documented here by others, but we don’t have knowledge as to why this has happened, and these coils have appeared in different vintage AR-3as, particularly with the Alnico woofer. Perhaps some AR-3-to-AR-3a conversions contained these coils. However, the first drawing of the "standard" AR-3a crossover clearly depicts the Alnico woofer (#3700) and the #9 coil (see Fig. 1); subsequent drawings also clearly show the #9 coil. All the component charts that I have show the #9 coil; in fact, I haven’t seen any drawings that show anything *but* the #9 coil. If there is such a drawing, I would love to see it. It is misleading for you to suggest that the *only* way to get agreement with early data was to use a #7 coil; I suggest that some input simulation data, calculations or assumptions are inaccurate somewhere.

http://www.classicspeakerpages.net/dc/user_files/958.jpg

Fig.1 First AR-3a Crossover Schematic

AR woofers (nearly all of the AR speakers manufactured up through perhaps the Tower Series) were designed to radiate into a 180-degree solid angle, not a 360-degree solid angle, as you suggest (see Fig. 2). AR speakers (*except for the AR-10Pi*) were designed to be placed with their backs against a wall, on a shelf or bookshelf, or on the floor (on a speaker stand) back against the wall. Never out in the middle of the room. At least until the AR Tower speakers, AR intended to their speakers to radiate uniform bass into a 180-degree angle, not 360.

http://www.classicspeakerpages.net/dc/user_files/959.jpg

Fig.2 AR-12-Inch Woofer (200003) Standard Frequency Response

There was also never any design regarding the so-called baffle-step compensation (BSC or “diffraction compensation”) in the early AR speakers. Can you image how a speaker -- designed with BSC for flat response into 360 degree radiation -- would sound if that speaker were moved back and put in a bookshelf against the wall? It would be unpleasant at best.

Roy Allison did investigate diffraction loss and boundary effects towards the end of his work at AR, of course, and continued with this once he had formed Allison Acoustics, but no formal work on BSC was done at AR to my knowlege. Designing for 2Pi and mounting a speaker in this environment, particularly flush-mounting in a bookshelf, negated the issue. After the 1980s it is possible that AR designed some speakers with diffraction compensation in mind.

I have asked Roy Allison three or four times now about (1) changes to the crossover chokes and (2) changes to the actual crossover values for the AR-3a. He states repeatedly, that until he left in 1972, there were no changes in the crossover of the AR-3a. Here are his words from just this week:

“Well, My goodness, Tom! Much ado about nothing, I'd say.

I repeat: Unless someone did something I wasn't aware of, which I thoroughly doubt, the AR-3a had the same crossover it had originally up to the time I left. After that I have no knowledge of a change and I would not have expected anyone to tell me if it did indeed happen.

However, if any change were warranted, it surely would not have been a decrease in choke value unless the woofer had been changed significantly also.”

--Tom Tyson

[dynaco_dan]

Hi there;

My 2 1/2 cents worth now.

Is it possible that the wrong inductors were put into the wrong installation area and installed by students or summer workers for example?

I can see someone that is inexperienced, just plucking something and mounting it without even knowing more than that specific job.

I do not know exactly how many individual AR-3A speakers were produced in a day, but maybe hundreds?

We have a local nail manufacturer, and I had a conversation with a former worker, and he said, when they had a bad batch of nails, it went into the ocean.

Sometimes a mistake is discovered, long after the fact, and the cost of recovering these speakers, if they could be identified, if the error is of small proportions, they may have been left alone.

Or else I am wrong.

It would not be the first time.

Have a great day.

[kkantor]

I've attached a PDF of the file that is only about 30K, but still very readable. Due to the particulars of this BBS software, the extension was changed from "PDF" to "ZIP." In order to properly read the file, it will be necessary to change it back after the file is downloaded.

I personally have come to believe that PDF's are generally the best way to share documents and images on-line. If anyone needs a clean, open source application to make these, CutePDF is the best, IMO. There are two files that must be installed: "CutePDF" and

"Ghostwriter." Once this is done, the application works very smoothly, in my experience.

http://www.cutepdf.com/Products/CutePDF/writer.asp

960.zip

[RoyC]

Tom,

I will admit that I am not technically qualified to address the "rightness or wrongness" of the #7 coil. John and Pete have been discussing that..

I do not know, however, how you can dismiss my observations as "nonsense". I'm calling them as I see them...and until you have been into as many cabinets as I have, and open a cabinet containing an alnico woofer to find ol' #7 staring at you (or not) you will never know the truth. (By the way, at last count, 10 post-1971 2ax cabinets with nary a #5 coil!)

Subjectively, the alnico/#7 combo does not sound "bad" to me, but put a ceramic in that #7 cabinet and you have a problem. That is how both John and I independently discovered #7 coil's existence! It simply sounded "bad" with a ceramic magnet woofer. On the other hand they BOTH sound OK with the #9 coil; the inductor of choice. The alnico magnet woofer and ceramic magnet woofer have some different characteristics, measurably and sonically. I know that probably disagrees with another document somewhere.

Who knows what changes were being made? I'll bet somebody at AR knew, regardless of what is being "remembered". If #7 was a 3a error, and we are seeing "some" 2ax's with this or that, well then AR's quality control was pretty crappy, wasn't it?. There were hundreds of thousands of these speakers sold and I am exclusively seeing the exceptions to the "rules"? What are the odds of that being the case?

If what I have observed are NOT evolutionary trends, we are all fools for holding AR in such high esteem.

Roy

[Guest Bret]

>As I've stated before the measurement facts from AR are there to see, the only way to come to another conclusion is to misread them.<

As Tom's chart shows and I can attest, the 10pi has both a #7 and a #9 coil.

I haven't tried to figure-out if these are doing the same thing as in the 3a because I don't know what effect the autotransformer is having on the circuit and their values. But since the 10pi will let you change the "pi" output for different installations (pi, 2pi, and 4pi) it makes some sense that these inductors are both there for that purpose.

Also, this completely blows my Alnico #7 and Ceramic #9 speculation out of the water since the 10pi never had an Alnico woofer.

The most straight-forward answer comes from Pete's investigative work which says that the #7 and #9 coils were changed when "flat" went from being measured into half to full space.

It also accounts for the apparent "lower midrange hump" in the earlier speakers. They need to be against a wall or in a bookshelf for the room reinforcement to the lower frequencies.

I'm afraid that until we get a better explanation, Occam's Razor says we should accept Pete's explanation as being correct.

Bret

[tysontom]

>Tom,

>

>I will admit that I am not technically qualified to address

>the "rightness or wrongness" of the #7 coil. John and Pete

>have been discussing that..

>

>I do not know, however, how you can dismiss my observations as

>"nonsense". I'm calling them as I see them...and until you

>have been into as many cabinets as I have, and open a cabinet

>containing an alnico woofer to find ol' #7 staring at you (or

>not) you will never know the truth. (By the way, at last

>count, 10 post-1971 2ax cabinets with nary a #5 coil!)

>

>

Roy,

Unless I'm losing my mind, I was not directing my message at anything you have said; on the contrary, I felt I was citing your historical data in stating, "Yes, some #7 (AR-2ax) coils have appeared and have been documented here by others, but we don’t have knowledge as to why this has happened, and these coils have appeared in different vintage AR-3as, particularly with the Alnico woofer."

What I challenged in my message was Pete B's statement below:

>I'm going to have to disagree here based on simulations of

>early impedance and frequency response measurements from AR.

>The only way to get agreement with the early data was to use a

>1.9 mH (Coil #7, rated 1.88 mH) inductor. That there were two

>different inductor values is also suggested by the fact that

>one set of measurements shows the 3a woofer as "flat" under

>half space measurement, and later measurements show it being

>"flat" into full space. Roy Allison's own published

>measurements with the same AR-3a show that this is not

>possible for the same speaker as I discussed here:

>

>http://www.classicspeakerpages.net/dc/dcbo...&mode=full#6202

>

>Allison discusses in that paper design trends regarding 2pi

>and 4pi system analysis and I believe the thinking in this

>paper was the basis for the change in inductor value.

>

>I'd say leave the #7 inductor if people want their system to

>be historically correct.

>

>Pete B.

More importantly, I refer to what Roy Allison has repeatedly said about no changes to the the AR-3a crossover. I have no explanation for the #7 choke in the early AR-3as, yet you and others have documented their appearance in early AR-3as. Perhaps AR used the #7 choke in the optional $90 upgrade from the AR-3 to the AR-3a, which was offered for some time during the first year or so of AR-3a production.

I'm sorry if you interpreted my comments as critical to what you have said earlier.

--Tom Tyson

[RoyC]

...and there I was feeling all indignant! Sorry for the misunderstanding, Tom. The fiberglass must be making me irritable.

If it turns out to be an AR-3 to AR-3a transition of some sort, it certainly would make more sense. They both used the same alnico woofer.

It is only when the ceramic magnet woofer is dropped into the #7 cabinet(which AR is not guilty of) that a problem becomes more audible to me. It would be interesting to compare the sound of an AR-3, an AR-3a/#7 coil combo and an AR-3a/#9 coil combo. Hmmm....

OK, I feel better now :-)!

Roy

[johnieo]

>Perhaps AR used the #7 choke in the optional $90 upgrade from

>the AR-3 to the AR-3a, which was offered for some time during

>the first year or so of AR-3a production.

Tom:

My two AR-3a speakers were not identical; I knew that before purchase. One was a factory-upgraded AR-3 to 3a; it contained a #9 coil. The other was an original AR-3a s.n. 7,139; it contained a #7 coil. The inductor used in the AR-3 was a "143" turn inductor, identified in your chart at "#3," so there is no way the #7 could have been left behind in an upgrade.

We found the AR-3a with a #7 coil and Alnico-magnet woofer to be too loud in the upper bass register, when compared to its mate with the #9. The speakers were mounted against a wall.

AR marked its coils with a clearly-scribed numeral, so there would have been no confusion in reading that number. However, if the coil assembly folk were given a handwritten assembly document with a poorly written "9," which they misread as a "7," that error would have gone undetected after cabinet assembly given the nature of the unit testing and the production volume of the early -3a (likely upward of 75-80 units/day.