The making of: The Two Towers (a 25 driver Full Range line array)

After riding along on a few different line array threads I figured this project deserved it's own thread.
I wanted to get new speakers for my living room for a while but could not quite find something I liked.
The speakers I have right now are just too big to use in a proper setup. In a domestic setting with a girlfriend and a 7 year old son there's just not much room to place anything. After surfing the net in search for an answer I stumbled over several line array's.
Now that was something I thought had potential! Way lower floor space needed than my current speakers with 15" woofers.

So the idea was born to build line array's using 25 full range drivers in each array, long enough to somewhat be floor to ceiling.

Next problem, what should it look like? And what materials to use? I remember seeing a DIY thread with a translam construction. That idea was very appealing to me. You can make it any shape you want on the outside and the inner walls can be used to benefit the design.

I started up my favorite 3D program and started sketching. Pretty soon I had some ideas about the outer and inner shape. Here's a much later version to give an idea:
stacked.jpg


I wanted a round outer shape, round corners to help reduce diffraction. On the inside I choose a wavy shape to hopefully aid in scattering/diffusing the back wave and making the walls uneven in thickness. That last feature should help reduce resonances of the enclosure.

Also obvious are the braces, making it a potentially very stiff enclosure.

The total array design on the left and the final outcome of this project on the right:
Finaldesign.jpg
LineArray.jpg

(not too far off I presume)

Here's a list of (hopefully) useful shortcuts to different milestones and/or projects done with the arrays:

A memorable moment, first sound!
It took a long time to get there, was it really worth it?

13th December 2014, First sound: A memorable moment!

Measurements begin...
My journey to better sound

14th January 2015, Start of first measurements: Measuring my way to better sound quality

Break trough in processing...
stepping outside of the boundaries of DRC-FIR

16th July 2015, Break trough in processing: Getting time coherency to work
(creating a time coherent correction in my listening room)

Proof of time coherency, at last :)

12th January 2016, Proof of time coherency at last! Tested with APL_TDA software

TDA_3D.jpg


Adding Ambience!
Taking control over the room

With installing a virtual Haas Kicker I'm trying to restore some of the energy I "robbed" from the room with my damping panels. I had some clues it might be interesting from my Car Audio days and always wanted to try a proper version in my home. By far one of the most fun tests I have done!
It starts somewhere here... with a later revisit of the theme here! The second part is concentrating on adding reverb to the ambient channels. I can highly recommend playing with this Haas Kicker idea, be it virtual or trough diffusive panels.


Review time!
Printed as is, I hope more will follow

21st January 2016, First official review: Jan Fekkes reviews the Towers

Shuffling the phase, how does that work?
Another interesting, hopefully successful experiment

A thread was started on Fixing the Stereo Phantom Center which inspired me to do some tests with it. Early on it made me do some mid-side EQ, following a paper in that thread. Later on it led to experimenting with a phase only shuffler to brighten up the phantom center. What I think is happening can be read here...

Condensed version of this thread?
For those looking for a more condensed version of this thread, I started a post (a couple actually) on the JRiver forums that's more compact, but it also lacks the detail I put into this thread.
Here's the link: Why I love JRiver, a tale of taming my Line Arrays

Another review, this time by DIYAudio member: Boden
Read it here...
Eelco dropped by for a chat about REW, JRiver and FIR filters and a short listening session. We talked about all things audio and discussed the DSP methods I use in this Monster thread.

Reviews keep on coming in, this time by xrk971
Find his kind words here
We managed to take advantage of a business trip to get both xrk971 and BYRTT together at my home.

It's raining reviews! Another one from BYRTT
You can find it here...
BYRTT was part of the get together. He's the only one to have heard both my lines and the insane effort from B&O, the Beolab 90.
I was very curious about a comparison.

Update! Driver change to Scan Speak 10F
End result here...
In time I'll write more about it after proper DSP has been applied.
I am curious though, will this be the first 25 driver Scan Speak 10F equiped array?
==================================================================================
Line Array Theory, a great read for anyone considering arrays
Infinite line source: analysis by werewolf

An old AES paper on Line Arrays, from speaker dave (David Smith), written at the time he worked at McIntosh:
Constant Beam Width Transducers line arrays

And as long as I'm adding line array papers: This original work of Jim Griffin is worth the read too!
http://www.audioroundtable.com/misc/nflawp.pdf

ISO 226:2003 Equal Loudness Contour:
Interactive link with dB numbers
==================================================================================
Amplifier tests!
The story starts here!
A big special thanks to member koldby and BYRTT. We had a small get together to do some subjective listening.
5 different amplifiers were tested with the arrays.
Vandermill-amp.jpg


==================================================================================

A Picture with the new DIY Scan Speak 30W/4558T00 subs added:

811714d1580045572-towers-25-driver-range-line-array-subinroom-jpg


The actual subwoofer build is scattered all over the thread, but here are some links:
Still making sawdust
Getting it together...
Did I screw it up?
Waiting for paint

744875d1553338967-towers-25-driver-range-line-array-submlv-jpg


Getting ready to paint...
Paint is on!
More info here: www.vandermill-audio.nl outside link (enough to make your own!)

And the new Fetzilla ambient amplifier that followed shortly after that:
817807d1581943721-towers-25-driver-range-line-array-theme-jpg
 
Last edited:
The driver I choose for this project is the Vifa TC9FD-18-08, also known as Vifa 9 BN 119/8 in Europe.
Vifa9BN119-8b.jpg

From the specs this seemed like a very good driver for the price and after looking at reviews and projects with this driver I became more and more convinced this could work.
As 25 drivers in a line array without any EQ would not work I searched the web again for a solution and ended up with a Behringer DEQ 2496.

Imagine my surprise when I found someone on this forum who had used exactly that combination!
Check out: stupid-cheap-line-array page 4, post2648779
Time for some questions obviously :D. It felt good though, to find someone with similar choices and even better when that someone is that enthusiastic about what he created!

After first getting 2 Vifa drivers to get my own opinion on them I ordered another 50 somewhere in 2011. I also ordered the Behringer and expected to be able to start the next summer. But that didn't happen. I never found the time to start this project. I tried to find an economical way to get my layers of wood cut. CNC would be ideal but difficult to find in my area and comes I was worried about the price.
I had read water yet cutting could be used to cut wood, even though that felt like the last thing to do: getting water into the wood would seem like a nightmare to me :scratch1:.

At first I was thinking about MDF as build material for the stack. But to stack that much MDF and worry about expansion due to changing temperature/moisture I switched to Baltic Birch Ply. That should expand less in thickness than MDF and I can't say working with MDF grew on me the last projects I used that.

But a local firm claimed hardly any water would be soaked up by the wood due to the fine spray used. My only left concern was price again. Each array consists of about 125 stacked layers, that meant a lot of cutting!
 
Last edited:
To make a long story short, I didn't get any work on these array's done in 2012.
But that did give me some time to further develop my plans for these array's. Meanwhile the interest in line arrays on the forum continued and another member showed up with a marvelous array based on the same Vifa speakers: stupid-cheap-line-array page 21, post2890230
(a lot more info and discussion went on in that thread, just linking the result)

Member koldby's work peaked my interest for an aluminum baffle for my project. The back mounted baffle looks gorgeous to me so I started playing with that idea. But I wanted to make sure it wouldn't have a negative influence on the output. Again Google search to the rescue I stumbled on some measurements on a German forum. This shows the effect of a baffle of 9mm with fillet used in front of the Vifa TC9: http://www.diy-hifi-forum.eu/forum/showthread.php?t=2157
 
Last edited:
I didn't want to go to an all aluminum enclosure but figured a baffle made from aluminum could have it's plusses. I would use one plate before the drivers with the fillet wave guide like shape and use another plate behind the drivers to mount them.

Something like this:
BaffleDetail-diy.jpg

The open spaces behind the back plate and in between the two aluminum plates could be filled with mass loaded vinyl. A heavy acoustic rubber, in my mind serving to dampen the aluminum baffle while making it more sound proof.
 
Fast forward to this year. Due to the bad economy I got laid off at work after working for 16 years for that same firm. I was in charge of IT and by boss decided to outsource that part of the business to save some money.

That sped up my plans for this project big time. The economy is still in very bad shape around here so I had to keep in mind I could be out of work for quite some time. I figured it's now or never if I want to actually build this thing! So in between hunting for a new job I started getting everything ready to start.

I bought the needed tools for the job, a Router (never worked with one before) and a drill stand.

As the design work was finished I figured out how to do all this work by hand. I figured I could cut the Birch Ply with a jig saw, a bit oversized and use a template to shape it with a router on a table.

To be able to use the template more than once I had them cut from 8 mm aluminum. (water yet cut) At the same time I had them cut the speaker feet from 2 cm thick aluminum.

A Router table was easy enough to construct so off we go:
P1030822-small.jpg


I started with a cheap € 80,- Router and some good router bit's. From what I had read the bit's were the important part to cut a lot of wood.
Well, that lasted a week with my production. The bearings on the Router were completely gone after that daily abuse. So that didn't go too well. After opening up my wallet a bit more I got a professional Router and what a difference that was! The only thing I had to watch out for is to keep the wood moving or it would burn!

Here's the first stack of a few chambers, also showing the templates used for the layers:
progress.jpg

(you can also see an oversized layer cut with the jig saw on the right below the template)

So this could work! Aside from taking up huge amounts of time that is :D.

After my girlfriend tipped over a stack of finished layers I decided to stack them with some temporary threaded rods (M6 in a 8 mm hole so I could see how big it would become.
Here's the first stack taking shape:
size.jpg
 
Last edited:
Thanks guys, yes it works, if one is crazy enough to put in the time :D.

A close-up of the chamber:
chamber.jpg

The layers used for the chamber are identical, just stacked in opposite order.

After days and days of jig saw and router work I had this:
finalstack.jpg


Still another box to cut. At this time I was happy to be going for 2 channel stereo (lol).

Several weeks later I had everything cut and it was time to work on the braces which will hold the baffle in place. To be able to do it with some consistency I decided to make a simple tool to drill the holes.
This is why I got a drill stand and not a drill press:
threads.jpg

Making use of my workmate to clamp the brace and use the tool to hold everything in place I could now easily drill the holes for the baffle mounts.

The nuts I used are a square piece of ply with a T-nut pressed in. Here you see the tool being fastened to one of the braces:
various.jpg
 
Too easy? :D

Working your *** off making all the needed parts gives you a lot of time to think and rethink parts of the design. That isn't always positive though.
As is reading a lot on this forum. One always gets new ideas and insights but I tried to lock into my plans and try to avoid side tracking.

One of the things that had me wonder is if I should use threaded rods in the glue stage or just use wooden dowels for guidance of the separate layers.

I decided upon the first idea, and have the threaded rod fastened with rubber rings on the bottom and top of the stack to be able to let the wood expand/contract somewhat. In between the layers I have open spaces where I countersunk the wood to attach the template for routing. That space is filled with a rubbery glue that stays elastic.

Here's the setup for the glue stage:
Base.jpg

gluesetup.jpg

The aluminum foot is glued to the floor while keeping everything level and the baffle is fixed to the roof of my garage.

A couple of straps would hold the braces firmly to the baffle. I used some spacers to have the baffle a couple of mm outside the line to be able to strap it down.

Another time consuming job, the glue stage. Here's what I did,
Glue 2 layers together and make them fit the baffle:
glue1.jpg

(here you see the countersunk hole, the layers are glued together at the non-countersunk side)
glue2.jpg

(the thread cutting tool is used from time to time to rid the bolts from residue glue)
glue3.jpg


After preparing 2 of these I glue them to the stack:
stage-1.jpg

Glue the bottom one first, slide down the middle one, again glue and clamp:
stage-2.jpg

(check level while fastening)
 
Last edited:
Do that enough times and you will get to this point:
glue.jpg


And even more days will get you to the top! Yeah!
P1040177-small.jpg

I had sanded the enclosure while being fixed to the roof. More sanding still to be done. The baffle would need to fit all the way down again etc. so even sanding takes days of your time to complete.
Here's a test fit of the baffle:
baffle-fit.jpg

(This is with 2 layers of mass loaded vinyl in between the aluminum plates)
 
Last edited:
One down, one more to go:
secondstack.jpg


Same principal, only learned a couple of tricks from the first one. Still a lot of work.
tower2.jpg

(the straps in action, together with the threaded rods to keep things lined up)

And if you hang in there, you'll get to the top once more:
both-towers.jpg


This is where I am today... of coarse planning ahead the next stages but first more sanding.

One thing on my mind is the final color for the speaker. In some of the pictures you can see some tests I did for staining.
I'm after a dark finish, light won't work in our pre 1930 house.
both-towers-a.jpg


Something like this maybe? It is my preference so far (click the picture for a better view):
finish.jpg

(the material used in this 3D rendering is an actual photo of one of the test pieces,
it's not scaled quite right. just for an impression. Imagine it with a gloss coating)


I think this combination gives it a bit of a classic touch. It would fit in our living room.
We have a dark wooden floor, white walls and lots of black and chrome in between.
Actually, not that much chrome. But a lot of black.
 
Last edited:
My friend, let me just say this: you have a true talent for wood working. This looks like a professional production, not a DIY project made from someone who had never used a router before. Gorgeous cabinetry.

Thanks for that! The planning helped to get to that level. I spend a lot of time thinking things trough to be able to make this work.
Knowing before hand from the earlier examples from members opc and koldby made me confident this kind of array can work good enough to be worth the time and trouble.
 
Yes, if you measure down vertically. But never straight opposing walls. (except for the braces, they will be covered with wool felt)
The shapes are random, but the same in each layer. Totally random in each separate layer is better of coarse but a lot more work. As seen here:
stacked.jpg

The walls will probably be covered with a rubbery coating but even that won't make the difference.
The wall thickness varies from 18 mm minimum to about 30 mm. It's main purpose is to change wall resonance.
I think to properly effect the rear wave the wavy shapes should be a lot taller. But I'd loose a lot of internal volume.

wesayso,

Sorry to hear about the job.

However, your project looks like it's coming together very impressively. The attention to detail and execution is superb, and great job. Look forward to further updates.

Thanks!
 
Last edited:
Very nice thread you have going here. I am sure this speaker will be awesome once finished and this thread will be frequently referenced by potential line array builders in the future. I like your use of mostly small power tools and it shows what one can do with a lot of elbow grease rather than fancy fully-loaded pro workshops.

Keep up the great work and good luck on job hunting! Wouldn't it be great if we all could just be paid to design and build speakers as a job?

I have a question of how many sheets of plywood have you used to make these speakers?

:cheers:
 
I wouldn't mind at all if this kind of work became my job :D.

On the plywood question: 4 sheets of 18 mm plywood, 152,5 x 152,5cm.
And about 1.5 sheet of 15 mm, also 152,5 x 152,5cm. All of them B/BB grade.
From the other half a sheet of 15 mm ply I made my router table.

Are you looking for those numbers to do it in foam core? (lol)

If I wasn't out of a job I'd use CNC though, or at least some kind of pre cutting the layers. The jig saw job was the least pleasant to do. Routing wasn't too bad. I'd have no problem doing that again.
 
Last edited: