Shooting for SOTA dipole line array -- seeking comment (long) - diyAudio
Go Back   Home > Forums > Loudspeakers > Multi-Way

Multi-Way Conventional loudspeakers with crossovers

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 21st February 2011, 06:51 PM   #1
diyAudio Member
 
Join Date: Oct 2010
Default Shooting for SOTA dipole line array -- seeking comment (long)

The speakers I have enjoyed the most are large planar/ribbon dipoles – e.g. the Infinity IRSV and the speakers I have owned (Magnepans and Apogee Divas and Duetta Sigs). The goal of this project is to create a speaker that builds on the strengths of the speaker topology I like while improving on some of the major weaknesses of the classic Apogees (impedance, sensitivity, dynamics…all related, of course) and minimizing unavoidable compromises and tradeoffs.

Below I have laid out my main goals as well as my first draft design, including some of my reasoning behind the design. I'm a newb speaker designer so I'd appreciate any thoughts on obvious design problems, better solutions, etc. That said, I'm not looking for suggestions to do something entirely different (unless you've heard something just like what I'm proposing, liked it, but found something else much better). Engineering is all about tradeoffs, and I don't imagine this design or any other is 'perfect.' I'm trying to get lots of what I want, with as little of what I don't want as possible. I'd like to avoid first-order mistakes/problems, minimize second-order issues, and recognize/understand third-order tradeoffs. To the extent you can indicate whether issues/flaws are first, second, or third order that would be great.

Also, in the short term I need to decide three main things that I'd appreciate feedback on:
-Whether to go as open-baffle as possible for max directivity or run with narrow baffles.

-Whether the woofer array will be dipole or cardioid (related to the first point). I have spent some time looking at John Kreskovsky’s site re dipoles vs cardioids but have never heard either a big dipole cone woofer (I have heard the Linkwitz Orions, but they don’t qualify as big) or a cardioid. The cardioid would be more work, but I like the ability to go lower, and to pressurize the room below the fundamental room resonance.

-What baffle shape to use for the mid-tweeter array. The three obvious options are: parallel-side simple baffle board with half rounds on the edges; narrow trapezoidal baffle similar in shape (though not width) to the classic Apogees; or a narrow baffle with ‘wings’ attached to the back like the Wisdom Audio Rush speakers. I haven’t really figured out how to use the Edge program and so I’m not sure yet what the tradeoffs are.

Key design goals of the project
- Nearfield performance at 500Hz at ~3m seating distance
- 120db max spl/dynamic range
- High (>95db) sensitivity for all drivers/arrays so as to bring the max SPL within reach of ‘reasonable’ amp power outputs (<200W)
- Very low distortion levels for all drivers, including F4 and F5.
- Very low moving mass for all drivers, esp try to achieve compatible/similar moving masses for mid and tweeter drivers.
- Smooth power response with minimal off-axis crud (how badly do I want to go open dipole all the way/higher directivity...??)
- Not more than 3 way design
- Crossover points at favorable frequencies (e.g. 1000-1500Hz for mid-tweeter, where the ear has a local sensitivity minimum).
- Relatively narrow front panels to minimize visual impact

Note -- crossovers will be digitally generated by Acourate software (can use Linkwitz, Butterworth, Bessel, or Neville Thiele, at any filter order >= 1). Inter-driver sensitivity differences will be handled either via Acourate or by the 80-bit equalization in the Metric Halo LIO-8 D/A. System will be actively driven with 6 amp channels.

Also note that my ceiling height is 9’3”. The room is relatively small. (I’m not going to say how small, because if I do, everyone will tell me it’s *too* small...and I'll ignore those comments anyway. Not least because these speakers will need to last me for the next 20 years+ and could end up in much larger rooms.)

And finally, the overall system appearance I'm thinking would be similar to the Wisdom Audio Rush speakers, i.e. a relatively narrow mid-tweeter tower plus a slightly shorter bass tower. Or like the Transmission Audio Megatrend with a narrow woofer tower.

First draft of speaker per the above goals

1. Mid-tweeter line array ca 90 inches tall.

-90 inch array puts the near field at 400Hz at a 3.1 meter listening distance excluding any ‘array mirroring’ by the floor/ceiling.

-90” mid-tweeter array comprises (on paper...) 9x BG Neo 10 plus 90 inches worth of Mundorf AMTs designed for line arrays (not in the Mundorf catalog, but available on an OEM basis). Could save money by running a shorter tweeter line – not sure how much negative impact that would have. Biggest expense in the project is in the tweeters.

-Estimated sensitivity of the Neo 10 array is:
500Hz = ~86db plus 9.5db of array gain plus 0.7db sensitivity gain = ~96db
1000Hz = ~101db
1500Hz = ~105db
Could end up needing some EQ on the low end, or cuts/pads on the woofer & tweeter

-Estimated sensitivity of the AMT array is 106db at 1500Hz and 111db at 4000Hz.

-Measurements for the Neo 10 were done by John at Zaph, see Zaph|Audio. See also the Neo 10 spec sheet at: http://meniscusaudio.com/images/Neo1...whitepaper.pdf.

-Pending measurements, assume the line array version of the AMT is similar to the 2540 unit shown here: http://www.mundorf.com/english/speak...ransformer.pdf

2. Woofer line array ~84 inches tall using 7x AE Dipole-12 drivers per array.

-Estimated array sensitivity is 90db single driver plus 8.5db array gain (plus a bit of sensivity gain) for a total of 99-100db.

-12in C2C spacing implies that the array shouldn’t run into half-lambda lobing issues up to 500-600Hz, an octave and a half above the proposed crossover frequency.

-Distortion plots for the AE Dipole 15 (should be very similar to the 12) are available at Drivervault. Distortion is very low. Mms is also low at 55g.
AE Speakers LO15 4ohm - drivervault

3. Crossovers at ~1500Hz and ~200Hz.

-Dispersions of Neo 10 and AMT are both ca -6db at 120deg at 2kHz… should still be similar at 1500Hz (TBD)

-Off-axis behavior of Neo 10 is better at around 1500Hz than at 2000Hz. Lowering the crossover point to 1500Hz will also reduce off-axis lobing from the center to center distance of the Neo 10 and AMT. At 1500Hz, the half-lambda distance is 4.5 inches, and the C2C distance should be equal to or less than this. This spacing is achievable with the Neo 10 and AMT.

-Per the Fletcher Munson curves, the ear has a local minimum of sensitivity at ~1250Hz at listening levels of 80-90db. This should be the best spot for a crossover till one reaches ~150Hz on the low side and ~8000Hz on the high side.

-200Hz crossover for the woofers means that at 1 octave above xover (400Hz) the woofer line at ~84 inches is still well coupled to the ceiling boundary at 9ft 3in (i.e. separation is less than 1 lambda) – an 84inch array implies 1 lambda or better boundary coupling to ceiling up to ~500Hz.

-While it would be preferable to cross the woofers at 80Hz or below (i.e. at or below the localization limit), the Neo 10 won’t go that low. Furthermore, the classic Apogees crossed over to the bass panels in the 400-500Hz range (Grand, Full Range, Diva, Scintilla, etc.) so 200Hz doesn’t seem very aggressive given the low mass, low distortion, and high sensitivity of the AE Dipole 12 drivers.

-12in drivers hit ka=1 around 180Hz and ka=2 around 360Hz, so should still be ~nondirectional at 200Hz. The Neo 10 has 180deg dispersion below 1kHz, so this xover point should not result in significant power response problems between bass and midrange drivers.

4. Room treatment.

-Want damping at ~500Hz and below, and diffusion above that (though I may want to damp rather than diffuse the first reflection points). The RealTraps Diffusor product appears to accomplish these goals.
  Reply With Quote
Old 21st February 2011, 07:33 PM   #2
ScottG is offline ScottG  United States
diyAudio Member
 
Join Date: Feb 2003
Location: US
I think we have gone over this before, but the few things I'd point out this time are:

1. http://www.systemsolutions.co.za/Tec...te%20Paper.pdf

Page 8 has the information on line length for your tweeter line vs. listener distance.. Typically listeners to this type of system listen between 3 and 4 meters from their loudspeakers. At 4 meters that puts the required tweeter line at around a minimum 1.25 meters in length.

2. Make sure you get horizontal polars on the Mundorf drivers BEFORE making a purchase.. many of their earlier models were directive above 3 kHz (..though this is NOT the case with their newer non-dipole AMT's). IF it's more directive then you generally want a rising response on axis, with a fairly flat target off-axis (..usually "aimed straight ahead" into the room).
__________________
perspective is everything
  Reply With Quote
Old 21st February 2011, 07:54 PM   #3
diyAudio Member
 
Join Date: Oct 2010
Thanks Scott. I have worked through a bunch of (simple) math to get to this point, and much of it came from Griffin's paper. As you point out, I could make the tweeter line modestly shorter, but not a lot.

I'm waiting for data from Mundorf, as they're preparing a revision to their OEM line array AMT. Will certainly get more data before buying anything, including more detailed polar plots.

Any other comments? Open baffle vs narrow baffle? Does this design make sense to you or does it run off the rails in obvious ways?

cheers,

Brandt
  Reply With Quote
Old 21st February 2011, 10:10 PM   #4
ScottG is offline ScottG  United States
diyAudio Member
 
Join Date: Feb 2003
Location: US
Quote:
Originally Posted by seaspeak View Post

Any other comments? Open baffle vs narrow baffle? Does this design make sense to you or does it run off the rails in obvious ways?

cheers,

Brandt

With this type of design I'd probably go wide-baffle patterned on an Infinity IRS V basic design. In fact you could always go with cheaper closed back tweeters for your tweeter array but double-up on them (front in phase, rear out of phase). Something you would NOT generally want to do with a small baffle. (..and yes, small light domes in a dipole array sound *very* similar to planars and heils that are dipole, in some respects better, and in others worse - often depending on the quality (not the expense) of the tweeters.)

I personally don't see the need for a line of subwoofers (dipole or not). A mid-bass line, yes. But not something that extends much below 60Hz near the average. From there I'd pick other options for a subwoofer.. less expense would be a monopole subwoofer just behind the listener (with full phase adjustement), greater expense = 4 monopoles with one at each corner of your room with full digital eq. and phase adjustment for each driver. IMO (and experience), both are better.
__________________
perspective is everything

Last edited by ScottG; 21st February 2011 at 10:12 PM.
  Reply With Quote
Old 21st February 2011, 11:42 PM   #5
expert in tautology
diyAudio Member
 
bear's Avatar
 
Join Date: Apr 2002
Location: New York State USA
Whoa... what is ur budget for this thing? Pretty big?

Also this sounds like a pretty large speaker...

And so, why not a real ribbon of some sort and why not an ESL??

_-_-bear
__________________
_-_-bear
http://www.bearlabs.com -- Btw, I don't actually know anything, FYI -- [...2SJ74 Toshiba bogus asian parts - beware! ]
  Reply With Quote
Old 22nd February 2011, 12:10 AM   #6
diyAudio Member
 
Join Date: Oct 2010
@ Bear -- this project won't be cheap, but it should be a factor of 6-10x less expensive than it would be if it were a branded, dealer-sold product of comparable performance.

The ESLs I have heard over the years have mostly been Martin Logans and have mostly been very unsatisfactory to me -- just don't like the presentation. From a technical perspective I prefer narrower drivers than stats can typically do. I have heard ESL 57s and the midrange was staggering. But to get performance like I want I'd need something like Picquet's stacked quads (which are an ancient technology and prone to arcing), and then I'd need subs too, and pretty soon we're talking about real money on a not totally reliable product.

As for ribbons, Scott persuaded me (in a thread a few months back) that a midrange ribbon would be inferior to an array of Neo 10s. And for the tweeter, I'd consider a ribbon, but I'd prefer not to design and fab it myself. I just don't have time, and the odds of screwing up a DIY ribbon the first time, vs screwing up an array of professionally designed and produced drivers (like AMTs) are much larger.
  Reply With Quote
Old 22nd February 2011, 12:29 AM   #7
diyAudio Member
 
Join Date: Oct 2010
hey Scott,

Ok, re the mid/tweet panel, I understand that a wider baffle will get me louder/lower. Will play with the baffle diffraction simulator as a next step.

As for the bass drivers... I don't want more than a 3 way. I understand that 3+ monopole subs in the room will take me down to 10Hz and will flatten out the response from point to point in the room, but to me that becomes a 4-way design as I don't want to use those subs above 60Hz and no ribbon/narrow planar is going to get down that low...so a cone array is going to be necessary, and I might as well take that array as low as it can go. Plus my neighbors wouldn't appreciate the monopoles...

I could simply buy some used Pipedreams/Nearfields or Megalines and add subs as necessary. But then I'd have non-dipole cones in the midrange (all the way with the pipedreams). And the quality of drivers would be lower than what I'm proposing. But that would certainly be the quick way to get a tall 3-ch array with deep bass.
  Reply With Quote
Old 22nd February 2011, 01:48 AM   #8
ScottG is offline ScottG  United States
diyAudio Member
 
Join Date: Feb 2003
Location: US
Quote:
Originally Posted by seaspeak View Post
hey Scott,

Ok, re the mid/tweet panel, I understand that a wider baffle will get me louder/lower. Will play with the baffle diffraction simulator as a next step.

As for the bass drivers... I don't want more than a 3 way. I understand that 3+ monopole subs in the room will take me down to 10Hz and will flatten out the response from point to point in the room, but to me that becomes a 4-way design as I don't want to use those subs above 60Hz and no ribbon/narrow planar is going to get down that low...so a cone array is going to be necessary, and I might as well take that array as low as it can go. Plus my neighbors wouldn't appreciate the monopoles...

I could simply buy some used Pipedreams/Nearfields or Megalines and add subs as necessary. But then I'd have non-dipole cones in the midrange (all the way with the pipedreams). And the quality of drivers would be lower than what I'm proposing. But that would certainly be the quick way to get a tall 3-ch array with deep bass.

The "might as well take that array as low as it can go" is a mistake. (..and yes, something I've experimented with - as have others.)


There is added non-linear distortion, but perhaps more importantly you don't get quite the "heft" of a good monopole (..or "kick" or "punch"). Ultimately its desirable to have a mid-bass "blend" of open baffle and a "sub" monopole - you get the clarity of the open baffle config and the tactile sensation of the monopole. If you read through Kimmo Saunisto's various projects you should start to "see" this character.

DIY archive of Kimmo Saunisto


BTW, when I suggested a mid-bass line - I was think more along the lines of a small driver array to keep mass low - more in keeping with the planars. I was specifically thinking of the Peerless 830656 - inexpensive, but not at all cheap (..and nearly perfect for the application - provided the baffle is very large). On a very large baffle next to your planar line these should provide excellent clarity, and with the high fs they wouldn't need a high-pass filter.)

If you wanted a somewhat smaller panel but with a line beside it (more like a smaller IRS V with panel AND sub array) ..the Exodus Anarchy wouldn't be a bad pairing (..though more costly).


A single sub near-field is a fair bit louder, with effectively less distortion, and MUCH better tactile sensation. Because it's louder for you it's less loud for your neighbors. It *should* be low-passed to where there is significant output up to 200 Hz - that way it provides an upper bass "punch".


Anyway.. as always, just suggestions.
__________________
perspective is everything

Last edited by ScottG; 22nd February 2011 at 01:51 AM.
  Reply With Quote
Old 22nd February 2011, 02:30 AM   #9
diyAudio Member
 
Join Date: Oct 2010
hey Scott, to clarify, as I understand it you're saying that a dipole or cardioid 3-way is going to have first order bass problems (within the context of trying to create a personal 'ultimate' system).

Is that a fair summary -- i.e. 3 way no, 4 way (with a dipole line of 6"-8" cone midbass drivers) yes?
  Reply With Quote
Old 22nd February 2011, 02:49 AM   #10
expert in tautology
diyAudio Member
 
bear's Avatar
 
Join Date: Apr 2002
Location: New York State USA
Well, I don't want to meddle with ur plans...

Looking at the AMT specs I'd be worried about the HF vertical response... if they make one "baffleless" that stacks, maybe, but I'd want to see proof in the form of a plot showing the response above 10kHz.

I got scared looking at the polar response curves above 10kHz... also I was somewhat concerned about the distortion graphs too... dunno.

And the BG driver, does that flatten out in an array?? The way it looks on their curve it is climbing at 6db/oct.... makes me want to drop it into a waveguide!

But it is an interesting project...

Oh, I would NOT judge ESLs based upon Martin Logans at all... but they will in general not meet your sensitivity and SPL goals... possibly a tweeter alone might.

_-_-bear
__________________
_-_-bear
http://www.bearlabs.com -- Btw, I don't actually know anything, FYI -- [...2SJ74 Toshiba bogus asian parts - beware! ]
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Dipole line-array idea StigErik Multi-Way 105 14th September 2012 02:18 AM
Line array dilemma for 7.1 HT (Long) exipnos Multi-Way 29 17th May 2005 10:23 PM
Long, narrow electrostatic tweeter for line array Bill Fitzpatrick Planars & Exotics 13 29th October 2004 08:33 PM
Line Array Ruminations and Questions (long) cytokine Multi-Way 28 21st October 2004 12:13 AM
line array? dipole? Shpoop Multi-Way 4 19th July 2004 12:18 PM


New To Site? Need Help?

All times are GMT. The time now is 09:13 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2