Corner Expanding Line Array with KEF Q100 - 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 26th June 2014, 03:03 AM   #1
ra7 is offline ra7  United States
diyAudio Member
 
Join Date: Feb 2009
Blog Entries: 1
Default Corner Expanding Line Array with KEF Q100

Inspiration is from speaker dave's interview here:
[Interview] David Smith [English]

And his expanding line array design:
Snell Acoustics XA Reference Tower loudspeaker | Stereophile.com

Attached is a pic of the speakers. I built some trial cabs just to try out the concept. The goal with this speaker design was to eliminate ceiling and floor reflections. It is supposed to be an expanding array with the array length increasing as the frequency is lowered. The outermost drivers have a LP at 125 Hz. The two central drivers have an LP at 500 Hz. And the KEF coax has an HP at 500 Hz. Then there is a passive crossover between the tweeter and the mid driver in the KEF Q100. There is some shelving of the mid and outer drivers to even the response.

Putting it in the corner gives it 90 degrees horizontal control to the low frequencies. I'm guessing that when you enter the wave region in the room, directivity is meaningless. Also, there should be no front wall reflection from the corner placement. Another big advantage.

There is a quadratic residue diffuser in the center and some thick absorbing panels. With this setup, the front wall completely disappears. There is a great sense of envelopment and a feeling of being in a different acoustical space.

My question is, are there any special rules to follow regarding the crossover and level? Is the approach to low-pass the outer drivers at progressively lower frequencies the correct approach? I've adjusted it to flat response and perceptually, it sounds flat.

Hoping that speaker dave chimes in, like he usually does. Any other thoughts, comments?
Attached Images
File Type: jpg photo (5).JPG (91.6 KB, 378 views)
  Reply With Quote
Old 26th June 2014, 04:42 AM   #2
CLS is offline CLS  Taiwan
diyAudio Member
 
Join Date: Jun 2005
Location: Taiwan
Well done

I have only one question: above 500hz or so, it's no longer quarter space radiation, does that detectable in listening or measurement?
  Reply With Quote
Old 26th June 2014, 06:30 AM   #3
Juhazi is offline Juhazi  Finland
diyAudio Member
 
Juhazi's Avatar
 
Join Date: Mar 2007
Location: Jyväskylä
cls, I figure that mid and tweeter's dispersion is wide enough. Corner placement minimizes early sidewall reflections, and therefore mid and tweeter should be placed deeply in the apex of the corner. With dsp it is possible to use time delay to each driver to get step response right

There has been a sideline discussion about similar construction here at diyaudio (synergy horn thread?) Sort of combination of Klipsch cornerhorn and synergy horn.

Using a line array bass helps to fight room modes and concentric mid-tweeter combo gives sharp imaging. This should be close to optimal configuration, when the room allows it (being symmetrical with straight clean walls)

Please ra7, tell us more, simulations, measurements, construction, everything!

My guess for sound tuning is to get a slightly recessing room response, measure one speaker at time, use rta ( and summing 20-30 cycles) or long gate 500ms. 40-200Hz straight, then gradually down 2dB at 100Hz and 3-5dB down at 5000Hz. I have found that I end to something like that if I tune by ear. The usual recommendation is straight up to 1000Hz, it leaves bass too lean for me!
http://www.prosoundweb.com/article/p..._using_rta_fft
__________________
AES Associate Member / My DIY speaker history: -74 Philips 3-way, -82 Hifi 85B, -07 Zaph L18, XLS10+PR/Hypex, -08 CSS125FR, -08 Hifitalo AW-7, -08 TangBand FR, -09 MarkK ER18DXT, -13 PPSL470, -13 AINOgradient

Last edited by Juhazi; 26th June 2014 at 06:46 AM.
  Reply With Quote
Old 26th June 2014, 09:47 AM   #4
CLS is offline CLS  Taiwan
diyAudio Member
 
Join Date: Jun 2005
Location: Taiwan
I bet it sounds excellent.

The previous question is based on the concern of attenuation along radiating distance, in which the mid to high frequency would be attenuated more than the bass in this setup. Would it a problem?

And, as a fan of 3-front-channel, I'd suggest an additional (wide-dispersing) HF unit at the center in this already excellent system, to compensate the wide separated sources for pinna cue.
  Reply With Quote
Old 26th June 2014, 03:54 PM   #5
ra7 is offline ra7  United States
diyAudio Member
 
Join Date: Feb 2009
Blog Entries: 1
Quote:
Originally Posted by Juhazi View Post
Please ra7, tell us more, simulations, measurements, construction, everything!
No simulations at all. Construction is pretty simple. Nothing spectacular as you can see. The boxes are built out of chip board. The main goal for me was to try the corner and expanding line array config. Intuitively, I knew it would work, so I thought I'd just give it a try. I took a chance with the placement of drivers. The outermost drivers are rolled off pretty early (125 Hz). I could've placed the outermost drivers anywhere on the baffle. Or there could be more drivers too.



The KEF cab and crossover design is documented here:
DIY with KEF Q100 drivers

I really like the KEF Q100 unit. Super smooth response. The crossover and EQ is through a PC-based system that I recently built. It is documented here:
Setting up a PC-based multichannel DSP system

The PC-based system allows full freedom with crossover points, slopes, delay, phase correction. No hardware limits. At this point, the crossover points and slopes are somewhat arbitrary. I haven't spent a lot of time optimizing them. Slopes are LR48. Unfortunately, I'm on the road until Monday, so I won't be able to show measurements until Tuesday. Basically, I set it up flat at a distance of about 6-7 feet.

Quote:
Originally Posted by CLS View Post
I bet it sounds excellent.

The previous question is based on the concern of attenuation along radiating distance, in which the mid to high frequency would be attenuated more than the bass in this setup. Would it a problem?

And, as a fan of 3-front-channel, I'd suggest an additional (wide-dispersing) HF unit at the center in this already excellent system, to compensate the wide separated sources for pinna cue.
This is a good question. I'm not sure what the answer is. Maybe speaker dave can enlighten us. Certainly, what I hear is not bass heavy. It does sound pretty good.

Last edited by ra7; 26th June 2014 at 03:57 PM.
  Reply With Quote
Old 26th June 2014, 05:07 PM   #6
diyAudio Member
 
speaker dave's Avatar
 
Join Date: Nov 2009
Location: The Mountain, Framingham
Quote:
Originally Posted by ra7 View Post

My question is, are there any special rules to follow regarding the crossover and level? Is the approach to low-pass the outer drivers at progressively lower frequencies the correct approach? I've adjusted it to flat response and perceptually, it sounds flat.

Hoping that speaker dave chimes in, like he usually does. Any other thoughts, comments?
I'm all for it!

Progressive low passes is the way to go if you have a line array of similar units, say a row of 4" full ranges. Then you would shoot for letting the effective array length stay in proportion to the wavelength.

Since you are using a variety of drivers sizes (therefore bandwidths) then you have to crossover high and low in the conventional way. For the XA series that is what I did. For the most part I empirically derived the designs using XOPT as the system simulation. I could track on axis, 15 degrees and 40 degrees up and down responses, and shoot for minimal change at 15 and lobe free constant drop at 40 degrees.

Geometry is key to this. You would love to have units spaced to less than a 1/2 wavelength apart at the tops of their respective bandpass (say, from M to M of the central MTM) but that is very difficult to achieve in practice. The greatest difficulty is getting a tight enough spacing between the tweeter and the 2nd-to-top elements. You have circumvented that by using a Uni-Q unit for the top crossover of 2 sections.

Based on that it would be nice to get a good match to the Uni-Q directivity for the lower sections. Your geometry is already fixed so the remaining variable is crossover points. If you don't have a simulator you can measure where your nulls and lobes come in by measuring up and down from the mid line and see if you can practically cross over before the worst off axis dips. Note that you can do some simple array modeling, good enough for your particular geometry, with Tolvin "The Edge".

I started with LR type crossovers but in the end had slightly better performance with non-symmetrical crossovers.

It will be interesting to see some measurements.

David
  Reply With Quote
Old 26th June 2014, 05:45 PM   #7
diyAudio Member
 
speaker dave's Avatar
 
Join Date: Nov 2009
Location: The Mountain, Framingham
Quote:
Originally Posted by CLS View Post
I bet it sounds excellent.

The previous question is based on the concern of attenuation along radiating distance, in which the mid to high frequency would be attenuated more than the bass in this setup. Would it a problem?
I think you are saying that the bass is more of a line array so it falls more slowly with distance.

If the XA (expanding array) is done right then it is a constant height with regard to wavelength, so it should drop with distance very uniformly.

In reality, when you get to the bass frequencies sound doesn't drop with distance at all since you are strongly into modal action of the room's standing waves. When standing waves occur then the pattern is balanced all over the room, meaning that peak level at a boundary is the same at the speaker end as at the far end.

I have worked with systems where different radiation patterns meant that varying drop with distance was a significant issue. The Mac XRT24 had a tweeter array and conventional woofer and mid. The line array falls off more slowly with distance. The only practical effect is that you should go out to your typical listening distance and do the response balancing there. Not a big deal.

David
  Reply With Quote
Old 26th June 2014, 06:06 PM   #8
ra7 is offline ra7  United States
diyAudio Member
 
Join Date: Feb 2009
Blog Entries: 1
Thanks Dave! I do plan to post measurements. Will do next week when I'm back home.

The 8" MM's are 21 inches apart (center to center). At 500 Hz, wavelength = 27 inches. So, they're a bit more than what is needed. A crossover at 330 Hz would put them 1/2 wavelength apart. That would eliminate the lobes or atleast make the central lobe very wide. The other option is to build a new cab for the KEF Q100. The cab right now is 11 inches tall. The driver itself is just 5 or six inches wide. I could also put the KEF cab on its side. That would result in a CTC of 18.5 inches.
  Reply With Quote
Old 26th June 2014, 11:23 PM   #9
diyAudio Member
 
Join Date: Jan 2011
Location: Willamette Valley
Now that's what DIY looks like right there, and I'm thinking it sounds real good. Dang, those newer KEFs aren't in the Goodwill store yet.
  Reply With Quote
Old 27th June 2014, 02:35 AM   #10
ra7 is offline ra7  United States
diyAudio Member
 
Join Date: Feb 2009
Blog Entries: 1
Quote:
Originally Posted by phivates View Post
Dang, those newer KEFs aren't in the Goodwill store yet.

You can buy the Q100 bookshelves and then harvest the drivers. Or even use the same cab and upgrade the crossover. I think the bookshelf model uses a simple first order network.
  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
DIY with KEF Q100 drivers ra7 Multi-Way 18 21st November 2013 06:08 PM
MCLA (Murphy Corner Line Array) project Dmontella Full Range 12 27th March 2012 03:47 PM
Line array corner sub Hezz Subwoofers 76 1st May 2006 05:37 AM
Line array behavior with corner placement johninCR Multi-Way 8 19th August 2005 02:52 PM
Line Array Corner Horn for any 4" driver johninCR Multi-Way 42 1st December 2004 05:13 PM


New To Site? Need Help?

All times are GMT. The time now is 07:49 PM.


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