Sort of Compact Line Array

[mante]

Arvo update

Dear Evil Twin,

At the risk of aiming all of the Fountek JP2's at me and destroying the state of Wisconsin, what has happenned to the Arvo project?

Have a productive week off.

Craig

[Evil Twin]

You are most observant, young Padawan.

The same transport delivering the Pau Ferro for the Saint-Saens also carried a hardwood shipment for the Arvo Part MkII prototypes, as well as Baltic Birch ply and other necessary materials. A secondary source is still required for aditional HDF, which is scheduled to be researched today by Jonmarsh after some veneering/gluing operations on the M8ta.

ThomasW will commence construction of his final Arvo cabinets (to replace the "rough" proto system he's using now) in the near future- electrically, his hybrid version is complete.

This will be a busy week....

[mante]

budget line array

Oh Evil Twin,

I'm feeling the force of something.....is it the Arvo or the Line Array?
Any chance of a "modest" Line Array that a young Padawan could afford
(and keep wife.) Gosh, something is making me want to sell wife and build bigger Array, must resist!
:twisted:

[Evil Twin]

Having a line array is like pregnancy- you can't be a little pregnant, and you can't build a "small line array", anymore than you can build a "small DeathStar". I believe "oxymoron" is the term your backwater planet uses to describe the concept.


The Arvo Pärt won't provide the SPL's of a large array and the low net distortion- but at more modest playback levels (< 100 dB) it is strong in the force in delivering a more convincing simulcrum of an original accoustic event than a comparable acoustic output capability box speaker can manage.

This follows from being a three way, reducing the midrange IM distortion, as well as having the more uniform power response and lower room excitation.

We're also looking into the possiblity of another version of the Arvo Ultra, not with 4 twelve inch woofers per side, but with a pair per side of higher excursion 12" drivers not yet announced from that new company being discussed on this forum. Because this driver may not be introduced pending technical evaluations, further discussion would be premature. I'm awaiting reports of further developments in this matter from my master.

We have confirmation of shipment of the initial evaluation set of midbass driver's for the Saint-Saens. It is expected that driver evaluations will be completed within one to two weeks, and construction should commence as planned shortly afterwards.

There are rumors that the competing team for a low cost dipole system to rival the Saint-Saens by using baby Magneplanars has enlisted the cooperation of one Jonmarsh to investigate and update the crossover desgin of the current MMG's. The MMG crossover design reportedly has been downgraded from the first version for "marketing reasons", as the original's performance was embarrassingly good at it's price point. I am skeptical that such an approach will yeild the necessary levels of performance to compete with our new deathspeaker. Only with the power of the dark side can the force be fully mastered...

[mante]

but Master....

"pregnancy, "small DeathStar", "oxymoron" ? Oh please be kind, I was thinking of a "thiswillhurtStar", not DeathStar. Something we could listen to on our small planet and still be able to feed the many children. O.K, I'll sell the wife, but we still need to eat :wink: How about 6-8 HiVi 12's and an RD50?

Always loyal,

Cj

[Evil Twin]

Your proposal is interesting Cj, and is a configuration I've thought about also. I'll run some calcs to see what four M12's per side would be capable of.

I'm also going to be measuring the Ascendant Atlas 12, but I wonder if it will go high enough. There's a chance the TC2+ may work up to 350 to 400 Hz or more; ETC evaluations will be required. To avoid beaming, the crossover shouldn't be any higher than 600.

This simpler configuration may have favorable characteristics for cost/performance characteristics - there's no reason not to evaluate it as well...

[Hank]

Evil Twin, that configuration is what I thought you were heading toward when you started discussing the HiVi 12. Then, you got to The Cafe and started ordering exotic drink concoctions made with super tweeter additions and yet-to-be announced mondo woofs and Saint started sounding like a megabuck adventure. That's when I made the awful mistake of noting the Maggies-on-top low cost array might be interesting. No, Evil, I won't go the Maggie route because I know they wouldn't do justice to a symphonic band or symphony orchestra. ops: But I digress.

[JonMarsh]

Well, who's to say the maggies plus mondo dipole woofers won't due justice? But, as Jethro Tull once said, "Nothing is easy but trying get you worried, my friend, it's OK..."

Maggies tend to have some significant panel resonances in the uper mid bass- and they DID change the crossover in the MMG's to dumb them down... so, who knows? We'll see if there's room for improvement.

This wouldn't be the first time- I did a redesign of the B&W DM6 with a new enclosure while I was a B&W dealer back then; sort of to make a point about some things I didn't like about the original design- of course, they point out my version was bigger and heavier, and would cost more to ship- I counted that my version was 2 dB more efficient, flatter on and off axis, and kicked the original's butt. (modesty was not a strong point of mine in my callow youth).

OTOH, maggies are not particularly sensitive, genearally around 84-86 dB/2.83 VRMS at 500 Hz, and that's considering that most Magneplanars are 4-6 ohm loads. The MMG's are 4 ohms, so they'll have to be wired in series unless you've got Aragon's. 'Course, some of us DO have Aragons, but we have sympathy for the rest of you folks... well, sometimes we do... :rofl:

I know Jim Winey probably won't like me messing around with one of his creations, but heck, when I first met him in the 70's (was also a Magneplanar dealer), he wasn't any spring chicken, so I think he'll have trouble catching me now... :lol:

Maybe this is really a job for Evil Twin....

[Dennis H]

Jon, I'm sure you've seen these but, just in case, MUG has schematics for the "old" and "new" MMG crossovers. I'm sure it's just a coincidence that the "new" one is a few bucks cheaper to build.

http://www.integracoustics.com/MUG/MUG/tweaks/index.html

[JonMarsh]

Quite a bit different, isn't it? I wonder if they changed the panel designs in some way... that 2.2 mH might give something of a long low frequency roll off, something to EQ the bass? VERY different tweeter crossover, too. Jibes with what I've heard.

Hehe. Should be an interesting week coming up...

[mante]

just following orders....

Dear "Evil Twin",

"Your proposal is interesting Cj, and is a configuration I've thought about also. I'll run some calcs to see what four M12's per side would be capable of" was an idea from that Jon Marsh guy. Since I am a young poor Padawan with a R45 ribbon, I was just looking for ideas to improve my current SS bass bin diy project. Also, I wasn't ruling out the Arvo Part V. Seems the longer I absorb ideas from you, the more projects I want to make, so what's another ribbon project.....the power is strong! Shall I take Hanks place? 8O

[Evil Twin]

You seem to be given to much insight, young Padawan.

There are many paths to the dark side- do not be afraid to follow your own unique way.

Together we can convince the skeptics and show them the true power of the dark side of dioples. As light and dark are complementary sides of the force, both needing their experssion in life, so is the front and rear of a speaker driver- both need expression in the listening room. From this principle comes the power of the dipole...

[Evil Twin]

Returning to your proposed configuration with the R45 ribbon and dipole bass, do you have a specific goal for upper volume level at 1 meter for each speaker? This could be modified in the lower ranges, considering that below 60 Hz recordings are often mono. But not always...

To illustrate the issues, consider the possiblities with the M12 HiVi. 7 mm Xmax, an effective Sd of about 530 cm2. For a starting design point, let's assume a path length difference on a straight panel of 3ft total, i.e., 750 mm.

1 acoustic watt is required for 109 dB output. Maybe you don't need quite that much from each side? Let's consider 106 dB for the sake of arguement (I don't know off the top of my helmut what an R45 is capable of). To hit 42 Hz at 106 dB will require 4 of the M12, based on volume displacement calculations for a 3 ft path difference. To hit 106 dB at 30 Hz would require 12 M12 woofers. The limitation is clearly Xmax.

Monopole, two drivers would be required to hit the 42 Hz point at 106 dB, and 4 at 30 Hz. This is a bit easier. If you were to not go dipole, it might make more sense to use a less expensive cone driver line array down to 75 Hz (say, 6 X RS225's per side, at ~$40 ea., or 8 - 7" RS180 ($30)), and a conventional long throw sub below 75 Hz, which could work down to 20 Hz or below.

[mante]

Staying on the same Death Star

"The ScanSpeak drivers John Meyer's uses are some of the best drivers available. That's not where the problem would be. Its the 1000Hz XO frequency. That XO point is high enough to be smack in the middle of the critical part of the midrange, not good IMO.

The transition from a cone to a ribbon/leaf, etc should be as low as possible. realistically this means 250/350Hz is the highest frequency used. This is why I'm considering the Newform for extremely high quality "mid/tweeters" to be used with the B&G RD planars. The B&G's go low enough to use a 300Hz XO point, but they are soft and ragged in the high-end. So XOing them to the Newforms around 2500HZ should create an expensive, but astounding speaker."

The above is a quote from Thomas that was written several years ago. The idea of lowering the xo for the SS8545, raising the xo for the Newform "ribbon" and adding the RD50 sounded like an "easy" solution for my system. A nice 3 way.

As far as 106db, with a 4 and 8 year old and I'm out of high school (ESS Amt 1c days), I tend to play things a little softer. O.K., sometimes Barney rocks.

Course, I could start the ribbon project over from scratch.....and yes, I still plan on expanding my speaker types by building the Arvo Part VII!

[Dennis H]

To me, using an RD with a Newform seems redundant as the RD will only play cleanly about an octave lower. After more testing, Jon and Thomas are recommending 600 Hz for the RD.

The biggest problem with the stock Newforms is the transition from the point source woofer to the line source planar. I'd look toward an array of 5" or 7" midwoofs, similar in concept to the Excelarray (except not so ugly ).

[Evil Twin]

"The ScanSpeak drivers John Meyer's uses are some of the best drivers available. That's not where the problem would be. Its the 1000Hz XO frequency. That XO point is high enough to be smack in the middle of the critical part of the midrange, not good IMO.

The transition from a cone to a ribbon/leaf, etc should be as low as possible. realistically this means 250/350Hz is the highest frequency used. This is why I'm considering the Newform for extremely high quality "mid/tweeters" to be used with the B&G RD planars. The B&G's go low enough to use a 300Hz XO point, but they are soft and ragged in the high-end. So XOing them to the Newforms around 2500HZ should create an expensive, but astounding speaker."

The above is a quote from Thomas that was written several years ago. The idea of lowering the xo for the SS8545, raising the xo for the Newform "ribbon" and adding the RD50 sounded like an "easy" solution for my system. A nice 3 way.

As far as 106db, with a 4 and 8 year old and I'm out of high school (ESS Amt 1c days), I tend to play things a little softer. O.K., sometimes Barney rocks.

Course, I could start the ribbon project over from scratch.....and yes, I still plan on expanding my speaker types by building the Arvo Part VII!

It is difficult to get good comparative data on the top end extension of the Newforms versus the RD50- it's difficult to find any measurements on the Newforms. But you already have them...

Crossing over the SS lower would be more of an advantage than you realize- most of their drivers have a cone profile designed to put an upward slope in the 800 Hz region, to facilitate baffle step compenstion- problem is, they're not pistonic above 800 Hz, due to this technique.

But what is the optimal crossover frequency? That's harder to say... I like using the RD's above 600- if you try to take them lower, the baffle design or enclosure becomes very critical.

The other issue is the difference in radiating behavior- combining a line source midranage with a point source woofer is problematic- it can only be balanced at one distance due to differences in fall off rates. I really have to recommend using a line source midwoofer; especially, as tall a radiating system as you can get in the "power range" of music (150-300 Hz) minimizes response disturbances due to floor bounce.

We don't like distubances in the force or in the radiation pattern versus frequency...

Below 500 or 600 Hz, a short line array of RS180 or RS225 would not be too expensive. But they would be heavy.

Consider these points carefully, padawan.

[PMazz]

combining a line source midranage with a point source woofer is problematic

As soon as I moved from a "bass box" to line of 7" drivers I was convinced. IMO, don't waste your time. The difference was immediate and, believe it or not, night and day .

Pete

[mante]

ideas, ideas....

Dear fellow members,

Thank you for your excellent ideas. Yes Master, this is a lot of information for a young Padawan to consider.

I enjoy the detail and large sound-stage the R45 offers. I haven't been off my planet enough to hear what else is out there, though, that could be a benefit. So, how about an array of woofers with a cross over around 1500hz-2000hz (this eliminates the HiViM12). Starting out as a two way, then with time it could become a three way.

The Excelarry uses six Seas W18EX woofers. Any suggestions on a "better" speaker to use? Does anyone know Jon's favorite 8" woofer? Finally, the force is strong to dipole.

Many thanks,

Young Padawan

[Evil Twin]

This presents a challenge, young padawan, considering your need for a 1500 Hz crossover. The least expensive driver I can recommend with above average performance would be the Dayton RS180- 1400 or so is where I would recommend crossing over. This is too high for 8" drivers, due to both energy storage and wavelength issues.

Given the low cost of the drivers, it would be reasonable to plan a line array of at least 8 drivers...

My next choice based on low non-linear distortion, good Xmax, and low energy storage, even up to 1800 Hz, is the Peerless 850439. But at about $75 each, the price, though much lower than ScanSpeak, is a marked step up from the Dayton. This is a driver that doesn't get a lot of attention in the DIY market, but is strong in the force for clean, neutral reproduction, neither favoring the light side or the dark side. It has a somewhat more open basket, but isn't sheilded, if that is a factor in your plans. A satisfactory crossover for the Peerless would be simpler and lest costly.

[mante]

Hello Evil Twin,

Thank you for your excellent advice. I did some further research and learned that the R45 can be crossed over as low as 880hz (1000hz recommended.) Does this xo change allow for other speaker choices that you would recommend; say six M8a's, or six RS225's ? The speakers do not have to be "cheap", just affordable for a young Padawan who is also saving for four Avalanches 8O

As always, your loyal follower.

[Evil Twin]

A 1000 Hz or lower crossover changes the picture substantially. In this case, crossing over around 900-1000 Hz is completely practical for the M8, and should be realizable for the Dayton RS225. The latter has about 2 mm more Xmax, so it would be stronger in the force for low frequency dipole output. It also has somewhat lower non-linear distortion in the range from the upper bass on down- an important consideration.

But the crossover design is more critical, as the RS225 has a cone mode (dip) at 1500 Hz. You will achieve the results you desire if you stay with a crossover in the 900 to 1 kHz region, and keep it steep. In fact, this might be the best way to optimize the LF extension of your R45- perhaps we can seduce Jonmarsh into designing an elliptic cauer LR-8 equivalent crossover for your application, if you choose to go passive in the crossover execution.

[mante]

Don't throw me off the Death Star

Currently, I'm running full active crossovers which can be modified. Are there any disadvantages when combining a line source monopole midrange with a dipole bass? Does one of your woofer recommendations sound "better" than the other?

So many questions and such little time to learn.

Thank you,

Young Padawan

[Evil Twin]

The first dipole speaker we reverse engineered as an experiment was the Legacy Whisper- this system was dipole in the range from 75 Hz through 300 Hz- no higher. though not dipole in the midrange (it used a control directivity array of Eton 7"s, still, it had in that range a naturalness of reproduction that well engineered monopole bass speakers in the same room struggled and failed to match. I prefer to maintain diple well up into the 1 kHz range, so that the radiation pattern is fairly constant- this should be possible with the design you're contemplating. The desired goal is uniformity of power response with respect to frequency, without being "hot" or recessed in the on axis behavior. Plus, some immunity to room modes, which seems to come from a dipole bass setup.

[mante]

"Perhaps we can seduce Jonmarsh into designing an elliptic cauer LR-8 equivalent crossover for your application". Thank you for the very kind offer. Jon is so busy I think we'll have to clone him :wink: Which woofers would you recommend for "best" sound and any suggestions on dipole dimensions would be greatly appreciated. I'm already way over my head and still having fun.

Young Padawan

[mante]

More Jon

Dear Evil Twin,

Have you had time to clone Jon yet? With so many exciting projects, me thinks we need at least three of him!

Young Padawan

[Evil Twin]

Hello Young Padawan,

Your project has not escaped my notice, but as you might imagine, holding a far flung DIY empire together
can be a daunting tasks, considering all the projects under way.

What I believe you may end up wanting is something like a reduced size version of the Saint-Saens, instead of a scaled up Arvo Part. The Saint-Saens wil be built with a trapezoidal panel design, possibly with the tweeter-mid line array edge at 90 degree and the outside angled, possibly with both angled. At this time, investigations using Baffle Diffraction Simulator are under way.

Do you have any impedance data and FR for the Newform ribbon? It would be possible to design for it just making an assumption of a pure resistive load and a nominally flat response at stated sensitivity within the passband, but this would be more "guesstimating" than designing. OTOH, I haven't seen measurements for these, and if they really work best as an very narrow baffle radiator, (just the ribbon driver in the air) this implies that they radiate full space below the half wave dimension for the driver width, and if you put them on a baffle, this will alter the response- boosting the forward output below about 3200 Hz. So, some caution should be used.

This could be worked around, but making unfounded assumptions would complicate the design process and require corrections down the road.

Interestingly, one of the few measured results I've seen on the web (Secrets of High Fidelity and Home Theater) would seem to indicate that the drivers are probably pressure amplitude flat in a large baffle, but as used in the Newform systems, have the expected rise in output up to about 12 kHz, then start rolling off above that. Hmmm. Level at 3-12 kHz is definitely up about 6 dB from 1 kHz. About what I would expect, if baffle mounting were to produce flat response.

But the response curve available at LDSG indicates otherwise, that the response is elevated below 3 kHz, in a way which would compensate for the lack of baffle.




This curve would indicate that if used on a larger baffle, a shelving function should be implmented below 3 kHz, so that the forward radiation is balanced. If used with a dipole line array, this would be mandatory. It holds the possibility of a flatter net power response with varying distance than the current Newform design seems capable of, as it mates a line source midrange and treble element with point source low midrange and bass.


Midwoofers are an interesting matter. In the case of the Saint-Saens, a goal I have been persuing is to try to find midwoofers with the greatest output for least weight and size, with reasonably low distortion. This has led me to the Extremis 6, for example, as extremism in the persuit of SPL is no vice, within reason.

But cost must be a factor at some point, and on this basis, the RS225 looks attractive; higher SD, half the cost, lower Xmax; overall, about the same output capability, but is larger and weighs much more, due to conventional magnet structure with sheilding. So, are you willing to use 8-10 RS225 per side, minimum, for a semi-wide range (to 60 Hz or so) dipole line array? This will test your commitment to the dark side, young Padawan... :uhoh:

[mante]

on track

Dear Evil Twin,

"So, are you willing to use 8-10 RS225 per side, minimum, for a semi-wide range (to 60 Hz or so) dipole line array? This will test your commitment to the dark side, young Padawan..." Commitment, or lack there of, 10 RS225s would be twice the height of my lightsabre, and taller than my ceilings :E That would really send my wife's back to the surgeon. Must stay strong!

I sent a letter to John Meyer regarding your questions so am awaiting his reply. The response curve above is also shown on Newform's web site. I was initially visualizing a ExceLarry (only less ugly) with an open back and tapered sides kind of thing. The R45 would not have an added baffle. OTOH, I will follow your excellent advice.

Again, thank you.

Young Padawan :T

[Evil Twin]

Do you have subs?

The problem with the diople array at lower frequencies is the unavoidable front to back cancellation. This reduces LF output, and requires more driver area and swept volume for the same SPL output compared with a sealed speaker. Rough guide is 4:1. That means, for the SPL you might be happy with from two sealed drivers, you'll probably need at least 8 in a dipole configuration.

One can run these calculations using Linkwitz's spreadsheet- for example, for a single dipole speaker using Extremis 6 in the midrange and midbass area, to get to 107 dB output down to 60 Hz requires 5 drivers at full excursion, assuming a net baffle width of ~ 20", and symmetric location. (40 Hz would require 14 drivers!) Since one would rather not push the drivers to their full limits, increasing the driver cout to 8 is prudent. Then, to get an asymmetrical baffle layout (smoother frequency response), the baffle size needs to be increased so that the minimum path difference remains at 50 cm.

With more conventional capability midbass drivers, for example, Dayton 7" midwoofers, between 12 and 16 would be required. Double the cone area (RS225), then halve the number of drivers.

Raise the low end from 60 Hz to 85 Hz, and the number of drivers drops by 3/5. Are you getting a feel for the tradeoffs? This is part of why the dual 8" Hivi midwoofers are crossed nominally at 185 Hz. They don't have to work very hard, then, and midrange purity and distortion are quite low as a result.

Consider that there are reasons to put the midwoofers and the Newform on the same baffle. For good horizontal dispersion for line source crossovers, there are some basic criteria for crossover frequency to keep in mind- (as for all speakers)- don't have the radiating area at crossover larger than 1 wavelength, and if possible, keep it to 1/2 wavelength. This means keeping the driver centers (vertical axis) within about 6-8" for 1 kHz crossover; no more than 10-12 inches at the outside.

One baffle to rule them all.

One other point to ponder- there's no rule that the drivers have to be in one column only, as long as the net width at the highest frequency is in the range of 1/2 to 1 wavelength. Preferrably toward the former.

For example, a staggered Array of 12 Extremis, six vertically, with some offset on vertical height and intermelding, might work fairly well... and make a 45 Hz cutoff feasible. (for the Saint-Saens). A similar approach with RS225 might be feasible, with a reduced number? Perhaps two rows of 4...

[mante]

To Sub

Hi Evil Twin,

I mailed JonMarsh some R45 stuff and sketches of a Padawan's dipole array. I forgot to mention that this system could be mated with a new diy sub. I guess we would have to compare advantages of a sub verses more speakers in the array. I understand from some of your other posts how difficult it is to go low in a dipole system. Also, Chad's business is not that far from me if a sub is needed. Finally, this is a music only system, no ht.

Young Padawan

[JonMarsh]

Hello Mante,

Reviewing your sketch and uploading is on my to-do's. Just for fun I'll run an SPL evaluation for the RS270's in the configuration you show, vs. frequency. I think getting the xover as low as 75 Hz at least is a good idea, if a sub is used, which means shooting for 60 Hz usable output near max SPL. OTOH, if no floor shaking HT is involved, a more reasonable target is feasible...

Doing distortion measurements this afternoon, so I won't have a detailed overlook before Monday.

[mante]

Jon,

No hurry, your plate is very full, and I'm still a young Padawan. John Meyer's suggestion was to put a back on the dipole and avoid the baffle for the ribbon. Wouldn't putting 6-8 RS225s or 8 RS180s in a long coffin style ported box be the simple way? Slap the R45 on the side and we're done :T

On the other hand, although I know what like, I'm certainly the beginner at making speakers speak well. Also, when are you headed this way, as in MPLS, because there is a warm front on the way.

Young Padawan
(another CJ)

[Dennis H]

In other words, Meyer wants you to klone the Excelarray. Boooooo. Something about the looks of that dildo standing next to a box thing really bugs me. If it were me, I'd go open baffle and build the ribbon into the baffle, crowding it as close as possible to the woofs. More work but it should both look and sound better.

[JonMarsh]

Ooh, put a back on a dipole? Uh uh. :nonod:

That may sound odd, coming from me, but let's face it, I'm a little schizophrenic at times...

So, I'm still designing some box speakers for folks and situation where really nothing else will do.

BUT!

If you ever hear a good dynamic dipole, you really don't want to go back. Really.

Example 1. Anyone who's heard Orions.

Example 2. ThomasW had a new acquaintance over to his place a month or so ago, who was interested and seeing his Acoustat + RD75 rig (obviously, diople...) Guys a doc, and interested and determined to put something together for himself sometime in the not too distant future. (I met him also while out there recently- but this wasn't the same visit).

He heard Thomas's Arvo's, as well as the big system, since I tweaked up both while out in December. He expected to be impressed by the 'stat's and BG's, but he was almost flabergasted at the way the Arvo's project the same kind of floating in space sound, with all those "ordinary looking drivers". His first reaction was that we ought to be building and selling these on the 'net.

So, my recommendation is to keep up the dialog re this project, and see if we can't put a concept together mating these which meets your goals while retaining the dipole behavior. I can design a crossover easily which will provide whatever "inverse baffle step" is required to compensate for the natural response curve of the Newform.

Using your original concept, I will calulate what some driver options and extension ranges are- then I'd propose a little BDS simulation to work out a good baffle design. I'd just consider it a warm up for the Saint-Saens...




Dipoles speakers are like CRT projectors- once you've seen or heard a good one, it's hard to go back to anything else.

~Jon

[mante]

ouch?

Jon,

Was that a laser that just hit me? I haven't given up on the Dark Side. I was really trying to ease your very busy schedule. Can't speak for John Meyer, though he does design and recommend monopole speakers, so it might seem obvious that he would suggest putting a back on the dipole.

Of course I'll keep up the dialog. I've invested two visits per day since June to this Mission Possible. It's too late to walk away, lets do it 8O

Craig

p.s. don't forget, I'm the one writing you every week to finish the Arvo's

[JonMarsh]

I've been doing some additional testing and configuration evaluation of the TC2+, including single VC drive in dipole- that may be the way I go, but it would change the crossover- AGAIN! So I'm going to evaluate carefully before proceeding with any additional changes.

For reference, here's the nearfield dipole FR and distortion plots on the TC2+ with 1 VC only driven- changes the Q, in the right direction for dipole use, of course. Drops the sensitivity above 80 Hz, but maybe not enough to be a problem. Note that the distortion is raised 30 dB in plotted level. This is an attractive response profile for the bottom woofer on the Arvo Part.





Here's what the TC2+ looks like in normal dual voice coil mode, coils in series. The response and level for single voice coil is shown in yellow for comparison.





You know, it's been a little painful doing this work on these other projects, and looking over and seeing the Arvo's in the corner, waiting for me to put new tweeters in and test... :roll:

~Jon

[mante]

dipole design with Newform

Back from the dead.......

How about two RSS315HF and four RS225 per side. Crossovers will be at 150hz and 1000hz, asymmetrical baffle layout ......

Cj

[tutter]

Well, who's to say the maggies plus mondo dipole woofers won't due justice? But, as Jethro Tull once said, "Nothing is easy but trying get you worried, my friend, it's OK..."

Maggies tend to have some significant panel resonances in the uper mid bass- and they DID change the crossover in the MMG's to dumb them down... so, who knows? We'll see if there's room for improvement.

This wouldn't be the first time- I did a redesign of the B&W DM6 with a new enclosure while I was a B&W dealer back then; sort of to make a point about some things I didn't like about the original design- of course, they point out my version was bigger and heavier, and would cost more to ship- I counted that my version was 2 dB more efficient, flatter on and off axis, and kicked the original's butt. (modesty was not a strong point of mine in my callow youth).

OTOH, maggies are not particularly sensitive, genearally around 84-86 dB/2.83 VRMS at 500 Hz, and that's considering that most Magneplanars are 4-6 ohm loads. The MMG's are 4 ohms, so they'll have to be wired in series unless you've got Aragon's. 'Course, some of us DO have Aragons, but we have sympathy for the rest of you folks... well, sometimes we do... :rofl:

I know Jim Winey probably won't like me messing around with one of his creations, but heck, when I first met him in the 70's (was also a Magneplanar dealer), he wasn't any spring chicken, so I think he'll have trouble catching me now... :lol:

Maybe this is really a job for Evil Twin....

Hello,
I just purchased a pair of B&W DM6 speakers from a distant location; cabinets arrived damaged. The sound is quite nice and I will have new cabinets built to order to improve appearance.

What I would really like to do is something on the order of what you did a few years ago - build a larger enclosure and hopefully improve the overall efficiency of the speaker system.

Do you mind sharing what you did to accomplish the gains that you wrote about?
Regards,

[JonMarsh]

Well, as you might imagine, I don't have any documentation lying around for that project these days (let's see, that was 30 years ago!), but here's the gist of it.

First/main problem is the whole low frequency alignment and the use of an autoformer to try to adjust the low frequency level/Q, in an enclosure that is too small in practice. I make that last comment because in those days most speaker designers were stil going for a cabinet Q of 0.707 or a little higher. This measures maximally flat anechoically, and if we listended to speakers that way, that would be fine, well, almost. Problem wtih a Q of 0.707 is the transient response is not critically damped. But a Q of 0.5 is, and works much better in room with typical boundary re-inforcement (see CARDAS website for info about speaker placement, as well as Avalon owners manuals).

So, the original crossover went out the window, the woofer went in a significantly larger LF enclosure, net Q stuffed was about 0.57, a Bessel alignment, technically, and the driver crossover networks were design by first cleaning up the impedance curves with zobel networks, using the same approximate offset positioning as B&W, and then designing new first order networks optimizing pulse response (1 msec pulse) and near field on axis response using an RTA, shooting for a slight droop in the power response (which is what RTA measures because there's no windowing to exclude room response) above 1 kHz.

Cabinets were 1-1/2" MDF.

If you don't have test equipment and some crossover design experience, that probably doesn't help much.

These days I'd probably shoot for slightly higher slopes (3rd order all pass, like the Modula MT) and forget the nominal time alignment except for making the crossover easier to setup.

You know, there's always more than one way to skin a cat.

In the room I listended to these in, the way they were setup, they sounded good, VERY good. They had holographic like imaging, but that was because they were essentially at the interface between two rooms, hand little side wall reflection, and had over ten feet clear behind them- no early reflections off the rear wall. They were really spooky on things like Holtz's "The Planets" in that room, in being able to reproduce the sense of an acoustic extending many tens of feet behind them. They were also pretty pychedelic (literally) on things like Boston and Santana. (system was an Audionics Preamp, my own custom built Dynaco 400's with enormous power supplies and double the nubmer of output devices), a Rabco linear tone arm and Denon DL103 cartridge.

If I was doing that again I'd want to look real closely at the cone modes in the kevlar midrange, just because I've been "burned" by similar Scanspeak drivers 10 years ago.

Hope this helps a little- but seriously, if you dont' have test equipment or someone familiar with crossover design, I'd go for upping the woofer enclosure volume by 50%, and crank the autoformers up to full (maybe you can just get rid of them), and get the whole mess a bit higher off the floor.

Good luck with your penguins. If it's of any help, I have a manual in PDF format, including crossover description/circuit.





~Jon