AR 12" Data

[kkantor]

I've placed a great deal of data on repaired AR 12" woofers at the following URL:

http://www.aural.org/ar_hist/data/

Over the next few weeks, I will be uploading more information, and trying to editorialize about what it all means. For now, some notes:

1- Eight woofers were tested for their small and large signal parameters, and for axial response in an infinite baffle.

2- The data is divided into 8 subdirectories, one for each driver.

3- A PDF file in the top directory shows an overlay of all the woofer responses.

4- Unit #1 is the 12" used in the 303a. Others are all some form of refurbished Norwood production, as described. None are stock, unfortunately.

5- Parameters are very accurate, but the acoustical responses seem to suffer from some consistent anomolies related to the jury-rigged baffle used. This will be corrected in future uploads. For now, best not to archive this response data.

6- There is a spread of about 5 dB in overall sensitivity, and a midrange responses do vary somewhat.

-k

[Guest Brian_D]

While sensitivity varies quite a bit, It appears that the woofers are very close in response between 50 and 575hz...

Interesting. Can't wait to see a summary!

-Brian

[tysontom]

>I've placed a great deal of data on repaired AR 12" woofers

>at the following URL:

>

>http://www.aural.org/ar_hist/data/

>

>Over the next few weeks, I will be uploading more information,

>and trying to editorialize about what it all means. For now,

>some notes:

>

>1- Eight woofers were tested for their small and large signal

>parameters, and for axial response in an infinite baffle.

>

>2- The data is divided into 8 subdirectories, one for each

>driver.

>

>3- A PDF file in the top directory shows an overlay of all the

>woofer responses.

>

>4- Unit #1 is the 12" used in the 303a. Others are all some

>form of refurbished Norwood production, as described. None

>are stock, unfortunately.

>

>5- Parameters are very accurate, but the acoustical responses

>seem to suffer from some consistent anomolies related to the

>jury-rigged baffle used. This will be corrected in future

>uploads. For now, best not to archive this response data.

>

>6- There is a spread of about 5 dB in overall sensitivity, and

>a midrange responses do vary somewhat.

>

>-k

Ken,

You have done a lot of work here, and from my standpoint, thanks very much! Obviously, most will not fully understand what is going on, but once you verbalize the data, it will all make sense. The composite overlay of all the woofers is most interesting, and certainly shows the response commonality amongst all the woofers except on the top-end response; there seems to be a lot of variation, but that is to be expected. Also, one can readily see how rough the 12W gets towards the 1kHz range -- clear indication that it needs the lower crossover. Do you plan to mount any of the drivers in the standard 1.7 cu ft. enclosure and measure so as to compare distortion, etc.? Your test setup may indeed be capable of emulating the actual closed-box parameters fairly accurately, but I would wonder about accurately assessing mounted-box, low-frequency harmonic distortion?

At some point (down the road after you have recovered from the time drain of this project) if you want the original AR-1W-type Alnico woofer, even an AR-1W raw cabinet and also the Tonegen service-replacement 12-inch woofer, let me know and I can send them to you. I may even have an original AR-3a 12-inch ceramic woofer that has not completed the inevitable urethane-foam rot.

Very nice, Ken!

--Tom Tyson

[Guest Bret]

>I've placed a great deal of data on repaired AR 12" woofers at the following URL:<

Yes, you have. I am looking forward to reading an explanation of what some of the less obvious measurements mean.

I'll be the the first to admit that I may not understand what I'm seeing and that I haven't had time to download it all and compare every measurement you took yet (it's been a long day) so I may be missing the obvious.

But what I am very surprised by is that these drivers are even close to each other considering the varying degree of repairs. For instance, if I'm reading this correctly, it appears that not-replacing spiders I know were pretty well-worn didn't seem to change much (the Simply Speakers pair) and changing voice-coils didn't change much, and even doing a complete reconing job doesn't seem to have changed much.

After I do print it all so I can see it side-by-side I ask that I may be allowed to revise and extend my remarks.

I have downloaded enough to know that the JPGs of the drivers are mixed-up. Clearly the Simply-Speakers stickers are on the "Edge-It" pair of drivers and the screens are on the 10pi drivers. I don't know how you marked them so I can't say which-is-which of the others.

That doesn't confuse me, particularly, as long as I know the description and test results are matched.

Thank you for doing all this, again. Looks like you have quite a fancy-shmancy piece of lab gear. What a great tool. It's taken almost a year from start to finish (not within your or my control), but it looks like it was worth the wait.

Bret

[kkantor]

Tom, Thanks! The reason I posted this early data was to encourage this kind of useful feedback. I will be improving and expanding the results and, more importantly, I will be culling them down as explaining there significance. I realize it is too much to absorb as is.

I would love truely vintage woofers to put into the mix, and I do indeed plan to use typical enclosures in an upcoming suite of tests.

Meanwhile, Bret, I am confused: Isn't the Edge-It done by Simply Speakers? That's what the paperwork I have says.

Machines? Klippel, HP, Audio Precision, GenRad and home made.

-k

[Guest Bret]

>Meanwhile, Bret, I am confused: Isn't the Edge-It done by Simply Speakers? That's what the paperwork I have says.<

Ken, I don't understand the confusion. Yes, the Edge-It is done by Simply Speakers. If if you look at the jpg of the driver in the subdirectory with the Edge-It results you find woofers which are not the Edge-It woofers.

Bret

[kkantor]

Well, just tell me how to sort them out and I will. I have been concerned with keeping the data properly related to each driver, but not attentive to where each UPS box arrived from.

There are four woofers with "Tri-State" stickers on them that I assumed were NOT the Simply Speakers. Rather than being confused, I will just follow your instructions on this.

-k

>>Meanwhile, Bret, I am confused: Isn't the Edge-It done by

>Simply Speakers? That's what the paperwork I have says.<

>

>Ken, I don't understand the confusion. Yes, the Edge-It is

>done by Simply Speakers. If if you look at the jpg of the

>driver in the subdirectory with the Edge-It results you find

>woofers which are not the Edge-It woofers.

>

>Bret

[Guest Bret]

>There are four woofers with "Tri-State" stickers on them that I assumed were NOT the Simply Speakers.<

Ken, that's correct. I'm really, seriously, confused that any of this could be mixed-up or confusing.

The Simply Speakers speakers have Simply Speakers stickers on them. Those are NOT the 10pi woofers. The 10pi woofers have screen on the back (to keep the fiberglass out). Those *better not* have Simply Speakers stickers on them because they were sent to Tri-State.

Okay, let's try again. Clear your mind of all pre-conceptions about what I've been saying. This ought to be really simple.

Edge-Its were from Simply Speakers and have the Simply Speakers stickers on them.

The Simply Speakers drivers were two AR-9 drivers.

In subdirectory "02" we have this description:

"10pi woofer #1, original cone, new coil, surround, spider.

Nearfield response, infinite baffle.

From Tri-State Loudspeakers."

In that same subdirectory there is a jpg - 02_s.jpg.

THAT JPG is of a speaker with a Simply Speakers sticker on it.

Obviously, that's not a Tri-State speaker or a 10pi speaker.

Why is it in the subdirectory with the 10pi results?

Bret

[kkantor]

Yeah, yeah.

Happy now, Mr. Fussbudget?

-k

[Guest Bret]

"Yeah, yeah.

Happy now, Mr. Fussbudget?"

Ken, you're just toying with me, right? Chinese water torture isn't considered fair-play and may not be allowed under the Geneva Convention.

Now you've switched to an AR-9 driver under the 10pi description, but at least the 9-driver pictured is from Tri-State, so things are looking-up!

Ken, uh. . . as you know, better than I, the 10pi drivers are black framed and have screen on their backs.

Fussbudget? FUSSBUDGET!! ("Fussbudget" always calls "Peanuts" and Lucy to mind; "Dog germs! I've been kissed by a dog! Get the iodine! Get the hot water!")

This is just too trivial a thing to get wrong. I can see in my mind's eye someone looking-over the data, pulling-down the jpgs, seeing "screen" on the back side of what's labelled as a 9-driver, looking at their own 9 driver, comparing the two, and then asking, "What's up with this!? My 9 driver doesn't have screen on it. Was this a replacement driver in my 9? Is it supposed to have screen? And why is this pictured 9-driver black?"

Or alternatively, "My 10pi drivers have black frames. These pictured are silver. What's wrong with mine? Are they AR-drivers or replacements? Should I pull the screen off of mine?"

My entire line of fussbudgety remarks is merely intended to ward-off some poor soul's mental anguish and subsequent pace-paths across his/her carpeting.

Fussbudget, indeed. I prefer to think of it as a professional qualification.

I've been called worse for less, but never by a better man. :-)

I think it is *far, far* more important that the test results under each description match that description. (which there is no way for me to test, of course)

The pictures aren't that important -except- as it is they could serve to confuse the uncertain amongst us.

06_s.jpg and 06_f.jpg should be in the O2 subdirectory

05_s.jpg and 05_f.jpg should be in the 03 subdirectory

03_s.jpg and 03_f.jpg should be in the 05 subdirectory

subdirectory 07 appears to be correct.

Subdirectory 04 appears to be correct.

I cannot tell, for sure, where the jpgs in 02 belong. I would have said "O6" but the Simply Speakers sticker is missing from 02_sx.jpg.

And to think I should be saddled with the epithet, "fussbudget." ;-)

Bret

[Guest Bret]

>I've placed a great deal of data on repaired AR 12" woofers at the following URL:<

Ken, too much. I can't get my head around the meaning of many of these numbers.

While we wait in extsacatic (compound word) anticipation of your in-depth analysis, can you tell us if you generally approve, generally disapprove, or find merit/fault in any of the various repair methods?

I don't mind the bib if I'm asking to be spoon-fed.

Sincerely,

I. M. Fussbudget

[kkantor]

Bret,

I'm sorry, I'm just joking around. Once you pointed out my mistake, all I did was RETOUCH the sticker off the photo. I promise I will fix everything when I upload the next batch of data. I will put the driver identity onto the photo image, to avoid this kind of mixup.

-k

[Guest Bret]

>I'm sorry, I'm just joking around. Once you pointed out my mistake, all I did was RETOUCH the sticker off the photo.<

It's not nice to fool father Fussbudget!

And you did.

Do I have a big "SUCKER" neon sign over my head now?

Ken. . . when you least expect it, expect it.

Bret

[kkantor]

In a nutshell:

1- The refurbishings do seem to work moderately well.

2- There are some (expected) problems with what is called the "set-height" of the rebuilt drivers. More or less, this the rest position of the cone under no signal conditions. The result of this will be somewhat lower peak output capability for high level sounds.

3- General operating parameters of the bass are preserved well enough for drop-in use. More details to come.

4- Some variation at the high end of woofer operation, but probably OK in most situations. Especially for the models with lower xover points. The exception would be TriState's effort to modify the new cone to match the mass of the old cone,(section 7). This yielded a bit too much break up and resonance at the top end for my taste.

5- Can't tell about the sensitivity issue yet. This might just be to be expected from the range of originals, and not really the results of the repair. A 4 dB spread is not the end of the world, but you would want your L&R matched closer than this.

6- The 303a woofer has more linear deep bass output ability. But it also has a more extended top end, so crossover details might come into play making a drop-in replacement unwise.

-k

[kkantor]

Please Mr. Bret, don't hurt me. I didn't mean to fool you. I was just trying to be funny. I'll pay for the window.

[Guest Bret]

>A 4 dB spread is not the end of the world<

That's the only thing I was smart enough to consider. Seems that you'd want to match those six by output to use them in pairs, then you'd get more like a 1-2db difference.

>More or less, this the rest position of the cone under no signal conditions.<

I wonder why that's an issue for the repair services. You'd think that this would be a parameter they'd routinely deal with. What is a "high level" in this context? 107db? 110? 120?

>The exception would be TriState's effort to modify the new cone to match the mass of the old cone,(section 7). This yielded a bit too much break up and resonance at the top end for my taste. <

Ah. The crux of the matter and the answer to your original pondering a year ago. You have my permission to modify the modifications any way you see fit if you believe it would be at all instructive.

And now to personal matters:

>Please Mr. Bret, don't hurt me. <

Hurt you? Oh, please. But the sarcastic pleading is additional motivation. ;-) The glass house reference wasn't lost on me, maybe I'll moderate. Nah.

Any man who likes embarrassing his M.I.T. professors in front of the class ought to *enjoy* having someone pull a "gotcha" on him once a decade or so.

I might be dead before a really good opportunity presents itself, or senility may set-in before something suitably clever comes to mind, but maybe not. . .

Bret

[kkantor]

Dribbles of data, including small-signal resonance, coil resistance and comparitive impedance curves.

[Guest Sean]

Ken: Thank you for all of the work that you've performed on this project and for sharing it with us. I'd also like to thank Brent who evidently forwarded his drivers to Ken and made all of this possible. Kudos to both of you.

Ken: The one thing that really surprised me about these measurements was the very high Bl that they demonstrated. Even the "worst" one was 14+, which is still quite high. Were you expecting figures like this from these drivers? I sure wasn't. Sean

>

[kkantor]

Sean,

I generally agree. The Klippel system I employed to get the parameters uses a dynamic system identification method that is very different from more traditional mass- or compliance-perturbation techniques. Exactly how these dynamically measured parameters relate to static small signal numbers is controversial.

In our particular situation, the Klippel seems to give somewhat high BL numbers, and related low Qes. Not out of the question for a beefy 12", and it is hard to know what "reality" is without doing destructive (or risky) testing. Still, this matter is on my list for double checking when there is time to test the drivers using more conventional and labor-intensive methods.

Also, I should have a test box ready for use shortly after CES in January, so we can see which parameters best relate to actual in-box performance.

[Guest Sean]

>In our particular situation, the Klippel seems to give

>somewhat high BL numbers, and related low Qes.

Glad to hear that you'll be doing some other follow-ups on this. Out of curiosity, have you tested some other "well known" drivers or current models that some manufacturers might still have spec's posted for? I'd be curious to see how the Klippel measurements for other drivers compare to the AR's and to other manufacturer's method of rating such things.

Were there any specific reasons that you went with the Klippel over the Dumax? I've read a bit about the two, but can't really recall the specifics. Sean

>

[kkantor]

These are the first drivers of this size that I have tested on the Klippel (which I've only had for a couple of months). Results on smaller drivers have been, perhaps, on the high side of BL, but not as noticably.

I plan to run some 1259's and TC's on the Klippel for a benchmark. I also plan to run the AR's using my tried and true impedance methods. (I trust them, and they generally relate well to real-world performance.)

Why the Klippel? Well, the group I am working with now bought one because it is very well suited to the kind of work on driver linearity (and novel driver design) that we are doing. For example, the Klippel generates polynomial expansions for the diff-eq's of the woofers that are useful in our work.

I have sent Dave Clark about 10 drivers for Dumax testing recently, just to see what could be learned. The Dumax perturbs the cone pneumatically, and measures an acoustic probe tone, while the Klippel perturbs the cone electrically and uses a displacement laser. As such, each method better simulates different aspects of actual use, and each has its short-comings. For our purposes the Klippel is more appropriate, though we are working on our own in-house system.

[Guest Sean]

Thanks for taking the time to explain the differences between the two systems. I'll be curious to see how the two systems compare to each other when measuring identical drivers, let alone how they compare to more conventional methods of measurement. If you can post a brief synopsis of your findings, that would be highly appreciated. Sean

>

[kkantor]

Hi all.

I have retested the seven 12" woofers I have on hand, using "traditional" parameter techniques that I think will be easier to understand. These are at:

www.aural.org/ar_hist/data/feb04_data/

Start with summary_02.xls. This is original spreadsheet, but with the new data added in red. It is easy to see how much higher the EdgeIt efficiency and Fs is, for example.

ar12_01.gif compares the simulated response of six of the woofers (not the 303a woofer), in a 1.5 ft^3 cabinet. (Sorry, I can only do 6 overlays at once.)

ar12_01.xls is a comprehensive interactive spreadsheet containing all the relevant data for the drivers, with means to try them with any kind of amp, crossover and enclosure you want. Different drivers can be brought up in the "database" field, and don't miss the various analysis tabs at the bottom of the worksheet.

Hope this is useful, and understandable. Not sure how much more data can be extracted from these units, until I have some real originals to compare to.

[Guest Bret]

Ken! I thought you had been forced to temporarily abandon us. Good to see you back.

>I have retested the seven 12" woofers I have on hand, using "traditional" parameter techniques that I think will be easier to understand.<

Thank you, thank you, thank you.

>Start with summary_02.xls. This is original spreadsheet, but with the new data added in red. It is easy to see how much higher the EdgeIt efficiency and Fs is, for example. <

Thank you thank you thank you thank you.

The interactive spreadsheet is just "way cool." Or in the vernacular of my day, "Far-out,man; that's really heavy." Obviously, I should have studied, you get the neatest toys.

I've got hours and hours of playing to do now.

I've got hours of digesting to do of the pop-up description of each of the measured parameters. It's like. . . it's like. . . sudden clarity, total consciousness, there's something Tibetan about it.

You do keep mentioning an "original." Which version of "original" do you want most?

Bret

[Guest Sean]

Ken, thanks for following up on our previous conversation and providing further test results. The only problem on my end is that i don't have Excel on my puter, so i'll have to see what i can do to correct that. I was able to view the gif though and it looked as if a couple of the woofers were markedly up in output at lower frequencies as compared to the others. I'll go back and take a look at that once i can view the other data that you presented. In the mean-time, are there any specifics that really stand out to you between these "conventional" measurements and those taken with the Klippel? Sean

>