Jump to content

AR3a Directivity


Zilch

Recommended Posts

There is generally a whole lexicon of descriptors for this with the majority of them describing frequency response effects. Wooly, muffled, bright sizzle, shouty, thick, boxy, soft, sweet, hard, fast, slow, "aw" sound, "eh" sound, "oo" sound, murky, heavy, projecting, megaphony, nasal, recessed, dark. Most engineers I know whould pretty much know what you were talking about if you used such terms.

Most of these terms seem reasonably self-defining, and I'm not an audio engineer. Maybe if those of you who are in the profession use them when you talk to each other, the rest of us can pick things up.

We could even try making a glossary with links to examples in sound files... :D

Link to comment
Share on other sites

  • Replies 73
  • Created
  • Last Reply
Most of these terms seem reasonably self-defining, and I'm not an audio engineer. Maybe if those of you who are in the profession use them when you talk to each other, the rest of us can pick things up.

We could even try making a glossary with links to examples in sound files... :D

Its not really the talking about them (in issolation), but the practice of listening to the speakers and describing what you hear. I do think that groups that work together train each other a bit in what they hear and how they describe it. That is, they develop a common nomenclature.

I believe Harman does training where a computerized graphic equalizer makes random eq changes while the trainee has to describe the bump or dip frequency. With practice most people can develop an acuity for it. Just playing with a 10 band graphic is also educational.

Sound files would work too as long as you had a perfect speaker to play them back on. Which would be a....

David

Link to comment
Share on other sites

Dismiss it if you like, but I'm about to become a subscriber to Gene's theory that the success of the LvR promotional demos was in part attributable to the festival of phase interferences so clearly in evidence here.

Which leads, inevitably, to my next question. Given that the original drivers are no longer being manufactured and that the audio industry has turned away from the various other design characteristics (cabinet and grille designs, etc.) that may have produced these phase interferences, and without commenting on whether or not they contradict current market preferences or industry practices, how might one go about modding a contemporary speaker to duplicate them? Could it be as simple as taping a picture frame to the front of Spitwad?

Take a speaker that doesn't exhibit the characteristics on these charts, alter it so it does and then conduct a test to see if it takes on the same sound (we obviously can't afford to rent Carnegie Hall, so something more modest would have to suffice). It's the only way I can think of to test the theory.

(The reply to this post might be a better fit in Pete's AR-3a clone thread, even though the 3a wasn't the speaker in the LvR demos).

Link to comment
Share on other sites

Question for Zilch here.

I think I know what you are trying to show, in that the AR3a has messy directivity and most of the CD horns are very good in that regard. Still there are a lot of people here that value wide dispersion, even if it is a little messy. Your data is almost too much to take in. Could you just show something simple like a 5k, 10k, 20k polar curve set for the various types (AR3a and "spitwad" or econowave)?

If the choice is between narrow but controlled vs. wider with less consitency, there will still be proponents for both camps.

David

Link to comment
Share on other sites

I think I know what you are trying to show, in that the AR3a has messy directivity and most of the CD horns are very good in that regard.

It's less about trying to show anything than documenting what it is, which is somewhat other than what many, if not most, believe that to be, and is part and parcel of the "There have been no improvements since the 'Golden Years'" thing.

Can we get to 150° broadband using a single CD waveguide? Dubious; 120°, maybe, but it also took divergent multiples to realistically achieve that back then, as well.

If the choice is between narrow but controlled vs. wider with less consitency, there will still be proponents for both camps.

I don't believe the camps have to be divided along those lines; Voecks et al. accomplish stunningly smooth wide dispersion across multiple drivers up to 8.5 kHz here:

http://www.stereophile.com/floorloudspeake...vel/index5.html

That looks like the LSR elliptical oblate spheroid waveguide delivering the narrowing top octave, an obviously deliberate choice. With this exception, it's apparent that very wide dispersion is both achievable and available in contemporary designs.

[in this example, however, not cheap.... :D ]

Link to comment
Share on other sites

Question for Zilch here.

I think I know what you are trying to show, in that the AR3a has messy directivity and most of the CD horns are very good in that regard. Still there are a lot of people here that value wide dispersion, even if it is a little messy. Your data is almost too much to take in. Could you just show something simple like a 5k, 10k, 20k polar curve set for the various types (AR3a and "spitwad" or econowave)?

If the choice is between narrow but controlled vs. wider with less consitency, there will still be proponents for both camps.

David

Another question for Zilch:

Why not, just for the hell of it, take an AR-3a tweeter/crossover and mount it on a flat baffle, measure the output free-field on- and off-axis up to 60 or 75 degrees, and then do the same with the spitwad horn thing -- and compare the two. No need for a gaited measurement here. This will show us two things: (1) it will allow us to see how close your measurements come to the old anechoic/B&K mic measurements made by AR back in the 60s and 70s, and (2) it will show us just how well the spitwad horn device performs measured the old-fashioned way. Who knows, the spitwad might do just fine.

For the up-close, point-blank music listeners, this type of measurement is perhaps meaningless, but for the reverberant-field listeners who listen well back in the sound field, this takes out the element of interference effects and diffraction.

There is no question that your measurements characterize and emphasize the effects of measuring an older speaker like the AR-3a up close with molding in place, and your curves show all sorts of destructive interference effects and lobing, but there isn't agreement here on how much of that sort of measurement is audible back in the reverberant field -- if you accept that there is such a thing.

Why don't you measure the two in this manner, and let the chips fall where they may.

--Tom Tyson

Link to comment
Share on other sites

When SpitWad's woofer and waveguide sat back in the Smaller Advent's 1" recess, there were issues, which were somewhat mitigated by moving the baffle forward to just 1/4" short of flush. The differences are documented in the Tech Talk thread for that project.

I believe there is considerable merit in rounding up a slew of AR3a tweeters and measuring them in the manner you suggest to asses the variability among (and viability of) vintage examples. I am presently doing that with AR4x tweeters for the next "Flex Your PCD Mettle" project.... :D

Link to comment
Share on other sites

The SpitWad summary is now posted on the Parts Express Tech Talk forum:

http://techtalk.parts-express.com/showthre...738#post1650738

They have the full bass extension of AR3a in a package half the size, below; Henry had a secret.

Blue = AR3a, Violet = Large Advent, Red = Smaller Advent.

[build 'em yerself for <$300 the pair.... :D ]

post-102716-1276713749.jpg

Link to comment
Share on other sites

When SpitWad's woofer and waveguide sat back in the Smaller Advent's 1" recess, there were issues, which were somewhat mitigated by moving the baffle forward to just 1/4" short of flush. The differences are documented in the Tech Talk thread for that project.

I believe there is considerable merit in rounding up a slew of AR3a tweeters and measuring them in the manner you suggest to asses the variability among (and viability of) vintage examples. I am presently doing that with AR4x tweeters for the next "Flex Your PCD Mettle" project.... :D

Zilch, what's the difference between the spitwad plastic horn and the Dayton Waveguide? Both seem similar in configuration. Does the spitwad use the industry-standard 1-3/8-inch threads? Any decent compression driver work? How about some of the older E-V compression drivers, or the University T30 or T50, etc.? Is there a link to the spitwad horn website?

How does the waveguide you're promoting differ from the old patented "reciprocating-flare" horn? Who holds the design patents (and what are the patent numbers) for the waveguide and spitwad horns?

Thanks,

--Tom Tyson

Link to comment
Share on other sites

I'll bet that big woofer in the 4430 and the dual 15-inchers, side by side, in the 4435, are rather beamy as the get above, say 400 Hz and reach up to the 1 kHz crossover point. That arrangement makes that of the AR-3 look downright controlled in terms of woofer directivity.

You'd lose again, Howard. In the 4430, the 2235H woofer narrows to ~100° at its 1 kHz crossover frequency for a smooth transition to Keele's 2344(A) 100° axisymmetric Biradial constant-directivity horn. In 4435, the second woofer runs in "0.5" mode, operating only below ~200 Hz to augment the extended bass response of the lower-mass but otherwise identical 2234H, providing greater power handling capacity in the dual-woofer variant:

http://www.audioheritage.org/vbulletin/sho...speaker-Systems

Zilch, what's the difference between the spitwad plastic horn and the Dayton Waveguide? Both seem similar in configuration.

They are both JBL-design progressive transition waveguides. The standard EconoWave uses a 12" x 6.5" version, whereas in SpitWad, with a 10.5" baffle width, I built using the smaller 8" square one, which does not play as low. Both are 90° x 50° dispersion, but the 8" is also available in 120° x 50°.

You'll see in the Tech Talk thread that I started out with the even smaller 6" square one, which I particularly like, but it does not maintain pattern control low enough to blend smoothly with the Smaller Advent woofer.

Does the spitwad use the industry-standard 1-3/8-inch threads?

They are both industry standard thread mounts, yes.

Any decent compression driver work?

Yes.

How about some of the older E-V compression drivers, or the University T30 or T50, etc.?

Yes, very likely.

Is there a link to the spitwad horn website?

We have to extrapolate from the JBL products which use them:

http://www.jblpro.com/catalog/support/getf...&docid=1080

SpitWad directivity is very much in accord with those factory specs.

Here's the 120° variant:

http://www.jblpro.com/catalog/support/getf...&docid=1079

There is a tech note detailing the design principles:

http://www.jblpro.com/catalog/support/getf...7&doctype=3

How does the waveguide you're promoting differ from the old patented "reciprocating-flare" horn?

I am not familiar with that term. Perhaps Dave knows.

Who holds the design patents (and what are the patent numbers) for the waveguide and spitwad horns?

I have not seen patents for the PT waveguides. The number for the Keele Biradial used in 4430 is listed in the product brochure....

Link to comment
Share on other sites

They are both JBL-design progressive transition waveguides. The standard EconoWave uses a 12" x 6.5" version, whereas in SpitWad, with a 10.5" baffle width, I built using the smaller 8" square one, which does not play as low. Both are 90° x 50° dispersion, but the 8" is also available in 120° x 50°.

Why does a "square" waveguide have different X and Y axis spreads?

Link to comment
Share on other sites

Why does a "square" waveguide have different X and Y axis spreads?

The contours of the flare define the pattern. There are smooth diffraction devices in the horizontal flares to spread the dispersion wider than the vertical. They are effectively Biradials without a diffraction slit.

Two fifteen inchers side by side will be beamy down to a pretty low frequency. And so would the single driver in the 4430 (probably no better than the AR-3 woofer).

Read what I said above. It's only one woofer playing above 200 Hz in 4435.

If your definition of "beamy" includes a 90° beamwidth, then, yes. The industry would not agree with that characterization, however.

I do not have the requisite expertise to participate in the Allison Dip debate. Perhaps Ken will weigh in on that one.

P.S.: The Allisons I have are Allison Acoustic 110s with the Allison-design tweeters. If they work, I will measure their directivity....

Link to comment
Share on other sites

Quick comment I've lost track of the key points of this thread. But, much research and field experience at AR, (consistent with tests I made and posted here recently in communication with Zilch), suggests that silicone grease is not a useful coolant in tweeters. It neither stays in the gap, nor does it conduct heat as well as ferrofluid. It does add some mechanical damping, until it migrates.

-k

http://picasaweb.google.com/kkantor/Ar#

... Silicone grease would spatter out of the gap that far down, but above 3.5 kHz silicone grease cools better than ferrofluid....
Link to comment
Share on other sites

Conclusion: The sample Allison 110 tweeters exhibit a nominal 160° beamwidth with a broadband stepped narrowing to 60° in the octave between 8 kHz and 16 kHz in situ. The response in this region is optimized for maximum flatness when listening at 20° - 25° off-axis.

Link to comment
Share on other sites

I think it was about 1987 at a trade show in NYC (possibly sponsored by Stereophile magazine) where I asked Peter Snell's mother about the polyurethane glued to the front of the tweeter on the AIIIi (Peter Snell had died a few months or a couple of years eariler.) She told me it was to improve high frequency dispersion. My first reaction was ????? But then after thinking about it, it made sense. By making the forward launched wave less efficient, the off axis launched wave by comparison was relatively louder. I presume the crossover network was adjusted to equalize the FR to compensate. It was so effective that it created the peculiar effect I'd heard with the AII at an acquaintance's house a number of times before he'd gotten rid of them. I inadvertently duplicated the effect while experimenting with some lateral firing auxilliary tweeters. I recognizd that sound instantly and I liked its coloration. I listened to it for a day or so and then wiped it out by reducing the laterally launched peak around 7 or 8 khz with a higher crossover frequency. Now that I know how it's done, I have no desire to hear it again.

Link to comment
Share on other sites

Is it the grille?

Nope, VHF is simply MIA.

Speakers become increasingly directional at high frequencies. A family of curves taken from 0° to 90° off axis reveals the severity of the effect.

[Directivity issues cannot be resolved with EQ.... :rolleyes: ]

post-102716-1276818120.jpg

post-102716-1276818159.jpg

post-102716-1276818176.jpg

post-102716-1276818194.jpg

post-102716-1276820382.jpg

post-102716-1276820393.jpg

post-102716-1276820406.jpg

Link to comment
Share on other sites

I thought you had an AL-110, with the 6.5-inch woofer, in combination with a tweeter having a rounded screen-guard protector over the dome.

Like the one in the spec sheet you posted at #52, right?

It's not about midrange or bass extension, rather, the allegedly superior (and eternal) performance of the Allison design tweeter. Allison states there's a directivity issue in the higher frequencies, and I have confirmed and quatified that here. Yes, it's "wide dispersion" to 8.5 kHz, but there it crashes. AR3a does, too, but more gracefully.

The question has arisen several times as to why, if the Allison tweeter design were all it is claimed to be, nobody has ever taken a similar approach. Until proven otherwise, I believe we have the answer.

Bottom line: Revel does wide dispersion better using a narrow-dispersion EOS waveguide in the top octave, even. I'm not familiar with their other product offerings, but perhaps they have something more "affordable" than Ultima Salon II that would be suitable.

PS: I never have said anything about directivity issues being controllable with equalization. Maybe you were referring to something somebody else said.

Yeah, you did; you didn't know it, is all.... :rolleyes:

Link to comment
Share on other sites

...

The question has arisen several times as to why, if the Allison tweeter design were all it is claimed to be, nobody has ever taken a similar approach. Until proven otherwise, I believe we have the answer.

...

1- The first "if" is a loop of circular reasoning. It sounds like you are trying to prove it doesn't work by saying it doesn't sell because it doesn't work. Does AR's popularity, (such as it was), prove that all of the company's technical assertions were correct? Of course not.

2- Power handling, cost, ease of manufacturing, all contribute to a tweeter's commercial viability. (eg- Defense against errant fingers is one the the >real< reasons that soft domes got popular. )

3- The "Allison" approach is not without legacy. (eg- some ScanSpeak and Vifa models, etc.)

4- I have a casual opinion, (no more than that), that this approach to tweeters has a relationship to the use of diaphragm flexure to assist the hf directivity of woofers.

-k

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.

I don't know about everybody else, but to me, a tweeter's supposed to get to 20 kHz these days, with a finite -6 dB beamwidth up there.... :rolleyes:

Link to comment
Share on other sites

Have fun measuring, Zilch.

Thank you, Howard; since I had them on hand I installed a $12, 8" square JBL PT-D26HF-1 120° waveguide into SpitWad, as it has the the best potential for approximating AR3a's dispersion pattern. Indeed, it holds wide dispersion out to 8.0 kHz and then smoothly transitions down to 50° at 20 kHz and beyond:

post-102716-1276842743.jpg

post-102716-1276842763.jpg

post-102716-1276843152.jpg

post-102716-1276843203.jpg

post-102716-1276843257.jpg

post-102716-1276843271.jpg

Link to comment
Share on other sites

I thought you had an AL-110, with the 6.5-inch woofer, in combination with a tweeter having a rounded screen-guard protector over the dome. What you have there is an LC-110, which is a considerably older model - and the one in that photo looks pretty old too, and pretty beat up as well. Lots of rust on that woofer, and one can only guess what the environment that generated that rust has done to the tweeter. One glance at the system and anybody can tell that it has not been well taken care of at all. I would be suspicious of the tweeter performance, Zilch. The previous owner may have used it for a automobile jackstand for all I can tell.

Howard Ferstler

I'm currently repairing a tweeter out of an AL-120 speaker. If that's the one of interest, I'll try and get some polars when it's fixed and re-installed in the cabinet. Let me know. Otherwise I won't bother.

Link to comment
Share on other sites

Archived

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


×
×
  • Create New...