Murphy's Corner Line Array project

[ppataki]

After having read many forum threads here and elsewhere I have decided to move forward with a Murphy's Corner Line Array project in my living room
I thought that I would share my progress here in this thread as regularly as possible

The design idea:
The original MCLA design has a 45 degree symmetrical 'toe-in' but that would not be ideal for me since I want to keep the equilateral tringle listening setup in the room so I will build a cabinet that will have a 30/60 degree front instead:

I will use 24 pcs of Dayton Audio ND91-8 drivers in each cabinet. The cabinet will not reach the ceiling (260cm) but it will get close to it (approx. 220cm)
Drivers will be grouped like 6 x 4, meaning four drivers will be wired in series and the six groups will be wired in parallel resulting in an impedance of 5.3 ohm. (4 x 8 = 32 / 6 = 5.3)
The cabinet will be made of 21mm plywood with Qtc=0.5 and the initial design looks like this:

The room with the speakers will look like this:

The listening triangle's size will be approx. 310cm so my head will be approx. 296cm away from the front wall.

I will use Dirac Live 3.x to apply digital room correction and I am also planning to use broadband absorbers on the front wall (with a thickness of approx. 20cm)

Build will start as soon as the drivers arrive, potentially in 4-6 weeks' time. I am planning to update this post with pictures about the process and some listening position measurements once the speakers are in place.

 

[Everett T]

After having read many forum threads here and elsewhere I have decided to move forward with a Murphy's Corner Line Array project in my living room
I thought that I would share my progress here in this thread as regularly as possible

The design idea:
The original MCLA design has a 45 degree symmetrical 'toe-in' but that would not be ideal for me since I want to keep the equilateral tringle listening setup in the room so I will build a cabinet that will have a 30/60 degree front instead:

View attachment 148971

I will use 24 pcs of Dayton Audio ND91-8 drivers in each cabinet. The cabinet will not reach the ceiling (260cm) but it will get close to it (approx. 220cm)
The cabinet will be made of 21mm plywood with Qtc=0.5 and the initial design looks like this:

View attachment 148973

The room with the speakers will look like this:

View attachment 148969

The listening triangle's size will be approx. 310cm so my head will be approx. 296cm away from the front wall.

I will use Dirac Live 3.x to apply digital room correction and I am also planning to use broadband absorbers on the front wall (with a thickness of approx. 20cm)

Build will start as soon as the drivers arrive, potentially in 4-6 weeks' time. I am planning to update this post with pictures about the process and some listening position measurements once the speakers are in place.

How will you wire the drivers, combination circuit?

 

[ppataki]

Just edited the original post to include the answer:
Drivers will be grouped like 6 x 4, meaning four drivers will be wired in series and the six groups will be wired in parallel resulting in an impedance of 5.3 ohm. (4 x 8 = 32 / 6 = 5.3)

 

[Jim Matthews]

How will you wire the drivers, combination circuit?

https://www.trueaudio.com/array/MCLA_project_details.htm

Figure 3.2 shows the original schematic (six parallel banks of four drivers in series)

 

[fluid]

Drivers will be front mounted and not back mounted, no waveguide. The baffle will be just a few mm wider than the speakers with rounded edges
I am not sure about the directivity of the drivers though

You said this in your other thread. I would caution you that tall baffles with lots of holes in are very weak. Even if made of 21mm plywood when the backs of the driver holes have been chamfered there will not be much material left and when moving these around they could be prone to snap. Adding another 10 or 20mm to each side might make physical construction easier.

Maybe you have already considered this but don't underestimate how floppy the material gets when you cut 24 large holes in it.

 

[ppataki]

Thank you @fluid for the hints, I will take them into account

The front baffle will be put together from 3 pieces, the front piece will be 96mm wide (the ND91s are 88.5mm wide, hence I mentioned the few mm difference earlier) and then there will be two adjacent sidewalls with 183 and 219mm width respectively

These 3 pieces will be joint (my neighbor is a carpenter, I leave these things in his capable hands) and I hope it will not break
Then the cabinet will be put together like this (view from above) the three front pieces together mounted to the two back pieces that will be fixed on the wall already

 

[fluid]

It will mainly depend on the size of the cut out for the driver rather than the front frame size, with the neo magnet the cut out might be smaller leaving more material in the baffle and less risk of breakage or weakness.

 

[John L Murphy]

After having read many forum threads here and elsewhere I have decided to move forward with a Murphy's Corner Line Array project in my living room
I thought that I would share my progress here in this thread as regularly as possible

I look forward to following your progress. Your proposed system should produce excellent results in your room. The increased size (top view) of your enclosure will move the corner reflections further apart than the original design but that would seem unlikely to compromise the total system response. I do like the modification to place the listener on-axis. I currently allow my arrays to cross in front of the listening position and am pleased with the result. Once constructed the biggest issue will be the precise voicing ("House Curve") you dial up. My current voicing is my "modified 10 dB tilt" and looks something like this:

It might be a good starting point for your voicing if your music library resembles mine (vintage rock, jazz, country, orchestral).
I have attached a .txt file with frequency and dB data.

Best wishes for a successful project.
John

 

[ppataki]

Thank you @John L Murphy, I really appreciate your comments - I keep my fingers crossed that all will be fine
I will keep updating this thread as I progress (still waiting for the drivers...)
Also thank you for the proposed house curve; I will give it a try

 

[ppataki]

Hi @John L Murphy I was playing around with the design a little bit and if I increase the Qtc to 0.577 I can have a cabinet that is considerably smaller (42 liters vs 100 liters) and I would lose only 2.2dB at 20Hz (that can be compensated for with the amp)
Would that move the corner reflections closer? Would that be more desirable with this design? I would be happy to lose that 2.2dB if I would gain other benefits

This is how it would look like:

Thank you

 

[kipman725]

If your amp can supply enough voltage to move the drivers to xmax you're not really loosing 2dB. I wouldn't run the array to 20Hz regardless as the available output will be small and dominated by distortion even from a large quantity of small drivers, you will also cause doppler distortion of mids/high. Idealy you would want to cross to subwoofers around 100Hz.

 

[ppataki]

I will include distortion (and other) measurements once the project is finished
I did some simulations with BassBox Pro: at 20Hz SPL shall be -27dB (as seen in my previous post), my room compensates +15dB at 24Hz (I know since I have measured that) so that will leave me with approx. -12dB to compensate with the amp (that equals to 16x amplification)
I hope that shall be OK with my Hypex NC250MPs

Cone displacement will be 4x with the amplification of 12dB. Xmax of these drivers are 5.1mm so I shall be safe till 1.275mm which is reached at around 30 watts

Now at 30 watts this setup shall produce 111dB of SPL at 1 metre

I believe that even when listening at approx 3 metres this shall be fine for me
But again, measurements will tell at the end of the day

 

[abdo123]

If your amp can supply enough voltage to move the drivers to xmax you're not really loosing 2dB. I wouldn't run the array to 20Hz regardless as the available output will be small and dominated by distortion even from a large quantity of small drivers, you will also cause doppler distortion of mids/high. Idealy you would want to cross to subwoofers around 100Hz.

if he crosses at 100Hz then this project is pointless because one of the main points of this sort of project is to eliminate the effect of ceiling/floor bounce at all frequencies.

 

[John L Murphy]

Hi @John L Murphy I was playing around with the design a little bit and if I increase the Qtc to 0.577 I can have a cabinet that is considerably smaller (42 liters vs 100 liters) and I would lose only 2.2dB at 20Hz (that can be compensated for with the amp)
Would that move the corner reflections closer? Would that be more desirable with this design? I would be happy to lose that 2.2dB if I would gain other benefits...<snip>

Hello ppataki,

Yes, shrinking the enclosure will definitely reduce the spacing of the wall reflections as you can see in these two views from above:

versus

...or consider the (one) paper speaker with two reflecting surfaces (mirrors).
(my crude photo of a single paper speaker with two mirrors at 90 degrees.)

Moving the single speaker out of the corner increases the spacing between all 4 sources.
Image analyses works great for sound sources.

Ref: J. B. Allen and D. A. Berkley, "Image method for efficiently simulating small-room
acoustics." J. Acoust. Soc. Am., vol. 65, no. 4, pp. 943–950 (1979).

Best Regards,
John

 

[kipman725]

mmm but look what happens to THD for small drivers:
https://www.erinsaudiocorner.com/driveunits/purifi_4/

 

[abdo123]

mmm but look what happens to THD for small drivers:
https://www.erinsaudiocorner.com/driveunits/purifi_4/

He can cross them to big subwoofers below the first room mode in the future. but definitely not 100Hz.

 

[fluid]

I wouldn't run the array to 20Hz regardless as the available output will be small and dominated by distortion even from a large quantity of small drivers, you will also cause doppler distortion of mids/high. Idealy you would want to cross to subwoofers around 100Hz.

In reality this is not the problem you make it out to be, subwoofers may well be useful and beneficial particularly if very high SPL's are desired, but for music there is enough output for many peoples tastes.

mmm but look what happens to THD for small drivers:
https://www.erinsaudiocorner.com/driveunits/purifi_4/

When you array a lot of small drivers the load is shared between them and it is like having a large Sd with a small x max.

I don't keep a lot of HD plots because I don't place much value on them other than showing that the design is functioning as intended

Here is outdoor with no EQ

Then with some basic EQ, the distortion is still 25 dB below the fundamental at 30Hz outside, the EQ improved the signal to noise and increased the HD less than the raise in signal level

Above 200Hz this is as good as most speakers

 

[John L Murphy]

mmm but look what happens to THD for small drivers:
https://www.erinsaudiocorner.com/driveunits/purifi_4/

Hello kipman725,

Yes, Erin's review of the Purifi Audio PTT4.0W04 is very thorough indeed. At $290 each those drivers should self-assemble into whatever speaker you want! The humble ND-90 (or ND-91) are much more suitable for arrays if your budget is less than the $14,500.00 that 50 of those would cost!

The good news is that "magic" (advanced technology...) happens when you gang up 24 small (and relatively inexpensive) drivers as seen in my test results on a single corner array using 24 ND90s:

Ref: https://www.trueaudio.com/array/MCLA_array_test_results.htm

Even with a single corner array the distortion at 50 Hz is inaudible at the highest sound levels a person (or room) can stand.
Now that's magic.

Thanks to Erin for the glowing review of DATS V3. (

)

Best Regards,
John

 

[kipman725]

works better than I expected. now show me 20Hz

 

[John L Murphy]

if he crosses at 100Hz then this project is pointless because one of the main points of this sort of project is to eliminate the effect of ceiling/floor bounce at all frequencies.

Don't want to be a butt head...but actually the point of the floor-to-ceiling (almost) line array is to gainfully employ the floor and ceiling (and side-wall) reflections in an attempt to approximate an infinite line array and all its seemingly "magical" attributes. Sort of like this...but infinite.

Maybe your eyes don't see those reflected images but your ears do.

The idea of the MCLA is to marry the speaker to the room not to make the room go away. As I see it, the speaker-room interface is the "final frontier" of the whole loudspeakers-in-rooms problem. Now, if only that pesky back wall and the secondary side wall reflections would just go away we'd have a complete solution!

When you consider the above image Is there any wonder that no subwoofer is needed? The constant nature of the sound field around the room is an indication that the listener is in the presence of an effectively infinite line array. Consider this, you cannot escape the near-field of an infinite line array no matter how far you move away from it! Well, you can leave the room and go outside but short of that you are in the grip of a wonderfully controlled sound field. Just ask speaker legend Roger Russell.

Roger Russell's Sound Column History
http://www.roger-russell.com/columns/columns.htm

Have a great weekend all, hope I wasn't a butt head.
John