Jump to content

AR3a Directivity


Zilch

Recommended Posts

  • Replies 73
  • Created
  • Last Reply
The Allison tweeter employed in each uses long-life cotton batting under the dome for resonance control and a non-deteriorating foam damping ring at the perimeter....
In any case, when I do full-range measurements (using an averaging technique with an RTA) my two IC-20s measure a tad better through the midrange than they did ten years ago (they are now 19 years old), but with some rolloff above 10 kHz that they did not have when they were newer.
If you are curious enough, get hold of a good test disc and see if you can hear test-tone signals above 10 kHz. If you cannot he[a]r up that high, or higher, then any tweeter deterioration issue will be academic.
This thread might have been 140 posts shorter if people had described real-world test results instead of just hurling dueling theorists at each other...

Done.

[but not the blather part, apparently. :P ]

AR3a: http://www.classicspeakerpages.net/IP.Boar...ost&id=5718

SpitWad: http://www.classicspeakerpages.net/IP.Boar...ost&id=5708

In other words, Zilch, you are publishing information (on the internet) that is, by definition, probably misleading, and for you to make any conclusive remarks about the tweeter's performance was being unfair to your readers and to Roy Allison.

I'm afraid it's up to Howard to verify his claim that Allisons are better than anything and everything.

[i am unable to substantiate this contention, alas.... :rolleyes: ]

Link to comment
Share on other sites

Zilch:

All this testing comparing 30-40 yr old drivers with modern ones seems a bit unfair to say the least. The general consensus among the AR experts here is that, on average, these old drivers just aren't up to original performance standards. So, what's the point?

Link to comment
Share on other sites

Zilch:

All this testing comparing 30-40 yr old drivers with modern ones seems a bit unfair to say the least. The general consensus among the AR experts here is that, on average, these old drivers just aren't up to original performance standards. So, what's the point?

It was you who posted the challenge for me to prove it, as I recall.

We have been blathering in the abstract about AR3a directivity for over two years, now. To the best of my knowledge, I am the first ever to have actually measured it and published the result.

The power response and DI might easily be calculated from this data.

Do we have reason to suspect that the directivity has changed over the years?

Do you see anything in my measurements that is inconsistent with Allison's published 40 years ago suggesting this?

Link to comment
Share on other sites

It was you who posted the challenge for me to prove it, as I recall.

We have been blathering in the abstract about AR3a directivity for over two years, now. To the best of my knowledge, I am the first ever to have actually measured it and published the result.

The power response and DI might easily be calculated from this data.

Do we have reason to suspect that the directivity has changed over the years?

Do you see anything in my measurements that is inconsistent with Allison's published 40 years ago suggesting this?

I'm a poor quoter here, but don't recall posting a challange. If I did, I apologize in restrospect.

There's lots of 'reasons to suspect' that have been already raised.

Don't know the answer to your last question. Perhaps HF can comment.

Comparing apples and oranges has always been poor science.

Some of the stuff I've published here represents one-off trials and experiments and should not be construed as representative of an overall effect or trend. Nor should yours.

Your work with e-wave reclamations is quite unique and admirable. Keep it up. I hope to publish my KLH 17 e-wave results sometime soon. Like Roy, I got bitten by the 'bug'.

Link to comment
Share on other sites

All this testing comparing 30-40 yr old drivers with modern ones seems a bit unfair to say the least. The general consensus among the AR experts here is that, on average, these old drivers just aren't up to original performance standards. So, what's the point?

There are several potential points, some of which have still not yet been fully addressed.

For the classic speaker owner/restorer, one point would be how to choose replacement drivers for repairs or restorations. The comparison between old dome tweeters and new 8" square waveguides doesn't help much here, but knowing that old OEM drivers may no longer have whatever performance might have once made them superior to what can be bought to fit new today might save someone a lot of shopping time and money. Testing and characterizing more old drivers can help create better profiles of real world performance today (especially if we can find old units that are in better working order) that we can compare to new drivers the way you did a while ago with those AR-218v tweeters.

For the classic speaker DIYer who might be interested in building a new speaker or modding an existing one, knowing what modern drivers can duplicate the performance of old, unobtainable drivers (even if they might require crossovers that are unconventional by today's standards to match the voicing of older designs) makes possible the creation of "new" speakers that replicate the signature sound of the old ones even if they don't use the same tech.

The key to making this effort pay off is for everyone to avoid injecting their personal opinions of what represents "good or bad" or taking potshots at each others' preferences to the point where the threads keep getting closed down.

Link to comment
Share on other sites

I'm a poor quoter here, but don't recall posting a challange. If I did, I apologize in restrospect.

The challenge was for me to build what I suggested might get the job done, and for Howard to review it. Howard categorically opted out.

I did my part anyway, in public, on a mainstream website, for all to see and build for themselves. Anyone with minimal resources can replicate it. The Part # for the 120° waveguide is 365124-001, $12 apiece. There is a small change in the value of one component in the crossover, which I will post over there along with the system measurements....

Link to comment
Share on other sites

While the Scan-Speak and Vifas may have their virtues, it does not appear that constant directivity is among them.

...

Indeed. Over the course of the original design effort, as well as subsequent design revisions, CD was removed from the list of goals for these tweeters.

In fact, the D-S-T (ne: Tymphany, etc.) has tested many, many waveguide designs over the years. Sometimes these were done for well-know customers, sometimes these were done as R&D projects, sometimes these were tests of 3-rd party designs that were offered up for license.

There is a huge body of work in this field. The names often mentioned here are but the tip of the iceberg. It's not the people don't know how to do CD horns, at least not in the real world. I remember lots of young engineers who worked for me at various places over the years, (AR included), who loved to spend their weekends making these. Defining a clear business proposition and sonic advantage is always the catch.

Speaking of which, I gotta catch a plane soon... bye.

-k

-k

Link to comment
Share on other sites

For direct comparison, here is the AR3a high frequency (above 1 kHz) processed at the same resolution and smoothing:

Thank you Zilch, you proved just what an outstanding tweeter AR3a is with your measurements and nearly confirmed Tom's assertion about its dispersion 60 degrees off axis. It is important to keep in mind that the tweeter crosses over at 5 khz. Therefore the comparison is between the output at 5 khz and 15 khz. look at the distance between the yellow line on one hand and the purple line on the other at 55 degrees off axis. about 5 db on both the left and right side and close at about every angle in between. You have proven exactly what you set out to disprove. Congratulations. BTW, remind me never to hire you as my lawyer! :rolleyes:

Link to comment
Share on other sites

It is important to keep in mind that the tweeter crosses over at 5 khz. Therefore the comparison is between the output at 5 khz and 15 khz.

There isn't a curve for 5 kHz and I wasn't sure how far below that the tweeter output rolls off to inaudibility, but here's what it looks like with 1600 and 2000 Hz deleted. Is there any audible tweeter output on a 3a down around 3 kHz?

post-102742-1276911425.jpg

Link to comment
Share on other sites

Thank you Zilch, you proved just what an outstanding tweeter AR3a is with your measurements and nearly confirmed Tom's assertion about its dispersion 60 degrees off axis. It is important to keep in mind that the tweeter crosses over at 5 khz. Therefore the comparison is between the output at 5 khz and 15 khz. look at the distance between the yellow line on one hand and the purple line on the other at 55 degrees off axis. about 5 db on both the left and right side and close at about every angle in between. You have proven exactly what you set out to disprove. Congratulations. BTW, remind me never to hire you as my lawyer! :rolleyes:

You presume to ascribe motive where none exists, other than to document what IS.

The Grn, Yel, Org, and Blu, 4 kHz through 10 kHz, all converge at ~55°, -6 dB, which defines a beamwidth of 110° in the inboard (+°) direction.

Outboard, there is no such convergence, and the beamwidth therefore actually varies with frequency between 120° at 4 kHz and 100° at 6.3 kHz, with the other frequencies intermediate between these values.

Violet is 12.5 kHz, actually, and symmetrical inboard and out, at 37.5° each direction, for a beamwidth of 75°. Cyan is 16 kHz, a bit narrower with 70° beamwidth. At these higher frequencies, the beamwidth has fallen to half the system nominal 150° in the woofer/midrange region.

OTOH, SpitWad holds steady at 120° through 8 kHz, then narrows to 90° at 10 kHz, 75° at 12.5 khz, 60° at 16 kHz, and 50° at 20 kHz.

There isn't a curve for 5 kHz and I wasn't sure how far below that the tweeter output rolls off to inaudibility, but here's what it looks like with 1600 and 2000 Hz deleted. Is there any audible tweeter output on a 3a down around 3 kHz?

AR3a's mid and tweet both play through the region between 5 kHz and 7 kHz....

Link to comment
Share on other sites

You presume to ascribe motive where none exists,

Hmmm, other than to visit a place called The "Classic Speaker" Pages to enlighten the congregation to the fact that they worship false gods, and are likely placing too much faith in pagan relics...based, of course, strictly on empirical measurements that you are making on their behalf. Sorry, couldn't resist :rolleyes:... Z, I do find your measurements interesting (love the colors), and, as you know, appreciate the information on waveguide systems you have been providing.

I'm guessing most popular non-waveguide speakers of ALL eras do not show the orderly dispersion characteristics of the modern waveguide systems. Personally, after experimenting with a variety of replacement tweeter candidates for the 3a, I subjectively agree that the dispersion attributes of the old AR systems have been mostly overstated, and is not the primary reason for their popularity.

...SO, with all this activity going on in the kitchen, I decided to drag out my Avid 102 E'waves and a pair of AR-2axs for some very unscientific, haphazard comparisons...

The 2axs are certainly more subdued in the higher end of things, and the E'waves are more detailed. Interestingly, however, as the volume is raised, the ARs seem better behaved, and become preferable, as the E'waves begin to exhibit a somewhat harsh midrange character. It is worse on axis, but still apparent as I walk around the room. The E'wave has an l-pad, but taming the midrange at higher volume takes the highs with it.

Other random opinions:

-Moving the speakers into a considerably smaller room seemed to affect the E'waves less than the ARs. Since that room dictates less volume, I actually preferred the E'waves in that room.

-For TV/DVD sound the E'waves were noticeably better for dialog...

-I seemed to prefer cymbals through the ARs on a number of recordings in both rooms.

-The results were pretty much the same as the last time I did E'wave comparisons (with Avid 103s and AR-3as).

The E'wave is an easy way to breathe new life into a tired old system...and/or a viable alternative for the adventurous to explore. I do not see it as a threat to those folks currently enjoying the sonic and visual personalities of older "stereo" systems. It is, however, significantly more worthwhile than exotic cables, boutique capacitors, and other dead-end efforts often pursued by those afflicted with "audiophilia neurotica" to "improve" old systems.

It would seem the orderly dispersion characteristics of waveguide systems would be effective for managing modern audio/video systems comprised of multiple speakers and much signal processing.

Roy

post-101150-1276928895.jpg

Link to comment
Share on other sites

Hi, Roy, and thank you for the eWave review. :P

Have you tried toeing them in at 45° cross-fired in front of your listening position? That will engage the "enhanced image rendition zone" mode, and no listener will hear on-axis direct, as that will be generating the first reflection off the opposite wall, and the early one from the near wall will be significantly attenuated. Constant directivity allows that to be done, as the spectral balance off-axis and on is substantially the same.

It's a bit disconcerting at first, not being able to find a precise "sweet spot" in the center, but after a bit, quite engaging, as the phantom center stays put when you move side to side.

[We've discussed image stabilization in another thread here in the Kitchen... :rolleyes: ]

Link to comment
Share on other sites

Hmmm, other than to visit a place called The "Classic Speaker" Pages to enlighten the congregation to the fact that they worship false gods, and are likely placing too much faith in pagan relics...based, of course, strictly on empirical measurements that you are making on their behalf. Sorry, couldn't resist :rolleyes:... Z, I do find your measurements interesting (love the colors), and, as you know, appreciate the information on waveguide systems you have been providing.

I'm guessing most popular non-waveguide speakers of ALL eras do not show the orderly dispersion characteristics of the modern waveguide systems. Personally, after experimenting with a variety of replacement tweeter candidates for the 3a, I subjectively agree that the dispersion attributes of the old AR systems have been mostly overstated, and is not the primary reason for their popularity.

...SO, with all this activity going on in the kitchen, I decided to drag out my Avid 102 E'waves and a pair of AR-2axs for some very unscientific, haphazard comparisons...

The 2axs are certainly more subdued in the higher end of things, and the E'waves are more detailed. Interestingly, however, as the volume is raised, the ARs seem better behaved, and become preferable, as the E'waves begin to exhibit a somewhat harsh midrange character. It is worse on axis, but still apparent as I walk around the room. The E'wave has an l-pad, but taming the midrange at higher volume takes the highs with it.

Other random opinions:

-Moving the speakers into a considerably smaller room seemed to affect the E'waves less than the ARs. Since that room dictates less volume, I actually preferred the E'waves in that room.

-For TV/DVD sound the E'waves were noticeably better for dialog...

-I seemed to prefer cymbals through the ARs on a number of recordings in both rooms.

-The results were pretty much the same as the last time I did E'wave comparisons (with Avid 103s and AR-3as).

The E'wave is an easy way to breathe new life into a tired old system...and/or a viable alternative for the adventurous to explore. I do not see it as a threat to those folks currently enjoying the sonic and visual personalities of older "stereo" systems. It is, however, significantly more worthwhile than exotic cables, boutique capacitors, and other dead-end efforts often pursued by those afflicted with "audiophilia neurotica" to "improve" old systems.

It would seem the orderly dispersion characteristics of waveguide systems would be effective for managing modern audio/video systems comprised of multiple speakers and much signal processing.

Roy

post-101150-1276928895.jpg

"The 2axs are certainly more subdued in the higher end of things"

As I reported several times elsewhere, after restoring a pair of AR2axs I was amazed at the uniformity of HF dispersion. Placed about 1 1/2 to2 feet off the ground in the corners of a 12 x 15" room with large archways on both side walls in a room of average liveness and toed in about 30 to 45 degrees, hf balance seems the same everywhere in the room. Once compensated for the hf falloff and with adequate bass boost, the speaker is an outstanding performer. It makes me wonder if the 3 1/2 midrange driver isn't a better match for the tweeter than the 1 1/2 dome in AR3a. I've had very little experience with AR3a and then only a very long time ago. I'm not sure I understand why what appears to be lambs wool was placed behind a grill in front of the midrange cone. So far except for its ultimate LF power handling capacity (one AR2a woofer had previously been damaged badly by an 80 wpc pioneer receiver but has been repaired) there seems little to fault for a direct firing speaker. Yet even AR engineers were not satisfied with this tweeter's dispersion and before going off on his own, Allison's attempt to widen HF dispersion even further led to LST. LST could have been manufactured more cheaply without the angled side panels. I do not agree with Ken that this was a marketing ploy to compete head to head with Bose 901 but was an engineering objective based on a consistent philosophy. I have to wonder if LST would not have been an even better performer and better able to produce successful LvR demonstrations had there been midrange drivers on the front panel comparable to the tweeter placement.

I have not yet experimented with further augmentation of AR2ax's hf dispersion, partly because I expect only marginal gains at best for my efforts. I haven't even remotely decided what to do with AR2a. Perhaps I'll experiment with some Tonegen ribbon tweeters I've had laying around since forever. I think the midrange drivers will have to go also. I saw a pair waiting for repair in Bill DeGall's shop when I was there and they suffered the same warped midrange cones mine have. His customer wanted an accurate restoration and so he actually fabricated new cones for them. Amazing what people can do.

Link to comment
Share on other sites

It makes me wonder if the 3 1/2 midrange driver isn't a better match for the tweeter than the 1 1/2 dome in AR3a. I've had very little experience with AR3a and then only a very long time ago. I'm not sure I understand why what appears to be lambs wool was placed behind a grill in front of the midrange cone.

Sounds like it's time for somebody to run a polar of a 2ax MR. :rolleyes: I have a spare I can donate if anyone wants to do it.

Many years I listened to a 2ax whose MR grille and pad had been knocked off accidentally for a bit before reattaching them, and even with a lot of fiddling with the MR control there was a boost somewhere in the MR that couldn't be leveled off. With all the EQing you're doing, though, you may have eliminated that issue. Have you tried taking the pad off?

Link to comment
Share on other sites

Zilch and Soundminded,

I will set up each pair per your suggestions today, and do more listening. With a little luck I can get some second opinions.

SM,

I agree...Bill LeGall of Millersound is a magician.. http://www.millersound.net/about.htm

The material under the 3a mid grille is a felt lined masonite ring, and dot, over a pad of fiberglass (same kind of fiberglass found on the 2ax cone midrange). I guess the 3a dome midrange is supposed to provide better dispersion than the 2ax cone. My quibble with it is its transition to the big 3a woofer. The 8 ohm version of the 3a style midrange found in the AR-5 actually seems to handle the transition to the 5's smaller 10 inch woofer more smoothly, imo. Another likely factor is the more complex 3-way crossover of the 3a (and 5) as compared to the 2ax, which acts more like a 2-way system.

My comments about the dispersion of the old AR systems pertain to my experience with tweeter experimentation, not the systems as a whole. Those old tweeters drop rather quickly below 5kHz. I think AR's tendency to allow those big, reverberant woofers to operate rather high into the midrange frequencies, while no doubt causing some negative issues, could be another reason wide dispersion is considered to be an attribute of the old timers.

Roy

post-101150-1276965460.jpg

Link to comment
Share on other sites

-For TV/DVD sound the E'waves were noticeably better for dialog...

Roy, any thoughts on whether this was directivity at work or just frequency response? I'm using an AR-1ms as my center channel speaker, and even after after putting a 1/4" layer of FG in front of the tweeter to bring its HFs down closer to the 3a's it's sitting in between, it also seems to do a better job with dialog than the 3a or 2ax did. This speaker has a dome tweeter, but it's an 80's model and in addition to being brighter than the classic models (before the FG), also seemed a bit "beamier" than the 3a and 2ax when I tested them as a pair on my second system. They were actually closer to the spread of the cone/dome combo tweeters on my AR-6s (which I haven't tried for this). But I wonder if that characteristic might actually be a good thing on a center channel speaker.

Link to comment
Share on other sites

Zilch and Soundminded,

I will set up each pair per your suggestions today, and do more listening. With a little luck I can get some second opinions.

SM,

I agree...Bill LeGall of Millersound is a magician.. http://www.millersound.net/about.htm

The material under the 3a mid grille is a felt lined masonite ring, and dot, over a pad of fiberglass (same kind of fiberglass found on the 2ax cone midrange). I guess the 3a dome midrange is supposed to provide better dispersion than the 2ax cone. My quibble with it is its transition to the big 3a woofer. The 8 ohm version of the 3a style midrange found in the AR-5 actually seems to handle the transition to the 5's smaller 10 inch woofer more smoothly, imo. Another complication is the more complex 3-way crossover of the 3a (and 5) as compared to the 2ax, which acts more like a 2-way system.

My comments about the dispersion of the old AR systems pertained to my experience with tweeter experimentation, not the systems as a whole. Those old tweeters drop rather quickly below 5kHz. I think AR's tendency to allow those big, reverberant woofers to operate rather high into the midrange frequencies, while no doubt causing some negative issues, could be another reason wide dispersion is considered to be an attribute of the old timers.

Roy

post-101150-1276965460.jpg

"I guess the 3a dome midrange is supposed to provide better dispersion than the 2ax cone."

I'm virtually certain it does. That's one reason I think the 2ax mid is a better match to the tweeter, it's dispersion is probably a lot closer.

"My quibble with it is its transition to the big 3a woofer."

That's one reason I think the 2ax mid is a better match to the woofer, lower fs.

"The 8 ohm version of the 3a style midrange found in the AR-5 actually seems to handle the transition to the 5's smaller 10 inch woofer more smoothly, imo."

I think the advantage is that the 10" woofer performs better at the high end of its range than the 12" woofer.

"Another complication is the more complex 3-way crossover of the 3a (and 5) as compared to the 2ax, which acts more like a 2-way system."

That's a problem with the 2ax midrange, there is no limit to its high end. As a consequence it will beam its hf forward reducing the overall hf dispersion of the system. A two way low pass filter for it might be the answer to that problem....if it were a problem. But if it is, I don't seem to hear it...yet.

Link to comment
Share on other sites

Roy, any thoughts on whether this was directivity at work or just frequency response? I'm using an AR-1ms as my center channel speaker, and even after after putting a 1/4" layer of FG in front of the tweeter to bring its HFs down closer to the 3a's it's sitting in between, it also seems to do a better job with dialog than the 3a or 2ax did. This speaker has a dome tweeter, but it's an 80's model and in addition to being brighter than the classic models (before the FG), also seemed a bit "beamier" than the 3a and 2ax when I tested them as a pair on my second system. They were actually closer to the spread of the cone/dome combo tweeters on my AR-6s (which I haven't tried for this). But I wonder if that characteristic might actually be a good thing on a center channel speaker.

Hey Gene,

In the case of the E'waves, probably both. Along with constant directivity, Z's graphs show a rather flat frequency response.

In contrast, some of the charm of the old AR character, often described as "warm", "musical" or "non-fatiguing", has something to do with a slightly depressed, less detailed upper midrange response. It made for friendlier speakers with regard to a variety of issues such as tape hiss, vinyl noise, radio broadcast issues, and harsh musical recordings (still a problem) of the day they were designed. Even Henry Kloss made a comment about depressing the center frequencies of an equalizer to make other speakers sound more like KLH speakers way back when. I believe that attribute may work against the old timers when it comes to modern video sound.

...just my seat of the pants opinion.

Roy

Link to comment
Share on other sites

That's a problem with the 2ax midrange, there is no limit to its high end. As a consequence it will beam its hf forward reducing the overall hf dispersion of the system. A two way low pass filter for it might be the answer to that problem....if it were a problem. But if it is, I don't seem to hear it...yet.

Tried taking the pad out of the MR yet? :rolleyes:

I suspect the pad is a really low-tech/low-cost low-pass filter.

Link to comment
Share on other sites

In the case of the E'waves, probably both. Along with constant directivity, Z's graphs show a rather flat frequency response.

In contrast, some of the charm of the old AR character, often described as "warm", "musical" or "non-fatiguing", has something to do with a slightly depressed, less detailed upper midrange response. It made for friendlier speakers with regard to a variety of issues such as tape hiss, vinyl noise, radio broadcast issues, and harsh musical recordings (still a problem) of the day they were designed. Even Henry Kloss made a comment about depressing the center frequencies of an equalizer to make other speakers sound more like KLH speakers way back when. I believe that attribute may work against the old timers when it comes to modern video sound.

Makes sense to me. In test reports of the period, my 1ms measure very flat between 200Hz and 10kHz (very boundary-dependent below 200 and rolling off above 10-12k), which is about all the range I need for a center. It's just that the all-black look doesn't fit in very well with the rest of the system. :rolleyes:

Link to comment
Share on other sites

My comments about the dispersion of the old AR systems pertain to my experience with tweeter experimentation, not the systems as a whole. Those old tweeters drop rather quickly below 5kHz. I think AR's tendency to allow those big, reverberant woofers to operate rather high into the midrange frequencies, while no doubt causing some negative issues, could be another reason wide dispersion is considered to be an attribute of the old timers.

Speaker dave has AR2ax and the requisite tools and skills to get the answers.

Anyone can set up to do this for less than $150.

[Assuming they have the desire to know.... :rolleyes: ]

http://www.classicspeakerpages.net/IP.Boar...?showtopic=5282

Link to comment
Share on other sites

Zilch, as far as I am concerned you are backbiting little creep who has a vested psychological and financial interest in becoming some kind of internet audio guru.

Don't hold back now, Howard; let it ALL out. :rolleyes:

Acutally, your cause would be better served by acquiring some decent measurement gear and presenting actual directivity data on your own speakers here....

Link to comment
Share on other sites

That's a problem with the 2ax midrange, there is no limit to its high end. As a consequence it will beam its hf forward reducing the overall hf dispersion of the system. A two way low pass filter for it might be the answer to that problem....if it were a problem. But if it is, I don't seem to hear it...yet.

Agreed...I don't think it is a problem. It is probably what the fiberglass pad is for. In the early days AR seemed to try to design drivers, and mechanically modify them, to blend naturally as much as possible. Electrical contouring seems to have been a last resort, and kept as simple as possible.

Roy

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.


×
×
  • Create New...