The "Diy" guide to Line Array speakers

So you want to make some line array speakers - woohoo, no stop just a minute and please consider that you might no know enough about speakers to be making line arrays!

Quickly ask yourself the following couple of questions and be honest because you might not actually be ready to tackle a project like an array in this nature just yet, so if you answer ( No ) to any of the following you might want to dial things down just a little and start on a smaller project and work up!

• Do you know how long a wavelength of say 1000hz is and why its length matters?
• Did you know that the crossover point will effect the polar response of a speaker?
• Do you know how to shift a speaker forward or backwards in time using a processor or custom crossover?
• Electrical vs acoustical phase, do you understand the difference?

If you have you have more answers being no or cant answer the above questions a line array might not be for you... just yet.

Why build line array speakers?

Line array speakers are basically a whole bunch of speakers on top of each other right?

Ok wise guy -10 points, sure we can think of an array as a whole bunch of speakers stacked onto one another! Though understanding Pa speaker design a well developed product is often designed with the intent to be used in a defined way, a good example of this would be to look at a loudspeaker horn.

Our first example is the P Audio PH 220 which is a 60x90 degree horn flair, we can assume that the sound plays at 75 degrees by 90 degrees.. right?

Well yes but we also need to understand what the effect of this is, we will be -6dB at 75 degree & -6dB at 90 degree from a defined frequency in which case P audio tells us down to as low as 1200hz.

Line array horns typically have a very narrow vertical dispersion from 3 degree's vertical to as wide as possible, whereas this PH-220 is the polar opposite sending high frequency in all directions. 

So while this might make a great speaker in a large venue where a wide response in all directions is required, for a line array this will not work at all as each box will cause something called comb filtering and the more enclosures we add the worse this will become. This is one of the primary " things " line array speakers try to fix!

Here is a great example by EAW 

Note we have a very wide horizontal angle of dispersion and hardly anything on the vertical, if anything a couple of degree at best to keep filtering to a minimal as best as possible, a second thing to point out is look at how close the lf section are to one another - the horn overlapping both drivers and the drivers tilting in slightly. In general the closer we can get speakers together the better for a wide range of reasons like the mutual coupling benefits and the crossover location point are impacted.

This is one of the key reasons you will never see a line array using just a 2 way using 15 inch speakers, these are often coupled including multiple smaller drivers like 4, 5 or 6" speakers even if the 15" can handle higher frequencies. The distance between the woofers are just to far apart to couple correctly, if this does note make to much sense don't worry I am going to show you a basic drawing and some generic crossover points to further highlight spacing between drivers.

Driver Spacing & wavelength

You will find many crossovers cover this range which are rolled off at different rates depending on the requirement.

• 800hz = 430.8mm
• 1200hz = 287.2mm
• 1600hz = 215.40mm
• 2200hz = 156.7mm
• 3500hz = 98.48mm
• 5000hz = 68.93mm

Knowing the wavelength helps us define how far apart 2 speakers like a woofer and a tweeter can be placed apart from each other, we want to be as close as possible in general or within a quarter of a wave length. 

We can see that using a generic crossover say 3500hz which has a wavelength of 98.48mm, a 15" speaker from the center of the cone to the frame is 190.5 give or take and using something like the Ph220 we are an additional 110mm which defines our total middle of the woofer to the middle of the horn flair distance to 190.5mm + 110mm = 300mm. This means that 1 our crossover point is well not good at all let alone correct to match the phase of our speakers as it was never intended to work with our batch of components but we can say or define that a total distance of 300mm means we should opt for a crossover that is 1200hz to 800hz which we can get as an off the shelf solution..

Though it does not mean to say that the that the woofer or compression driver and horn can work great at those frequencies this is a guide line for noobs, a rule that you should practice. 

So we have defined 2 things really which is important prior to kicking off our Diy line array enclosure, 1 we want to keep our components as close to each other as possible and secondly we need to crossover lower than or as equal to the distance between those components.

I guess the next thing we can ask ourselves is do we use a single woofer and what size or do we use multiple speakers, what are the advantages and disadvantages - which is right for me?

Right off the bat using 2 speakers costs more than 1 speaker.. no brainer :) so the cost factor is a bonus and the crossover consideration is more simplified.

Two drivers we are not just considering the distance between the woofer and the tweeter but the configuration becomes an MTM popularized by D'Appolito a couple moons back presents us with the challenge of the added consideration of the spacing between the woofer to woofer over and above. For those who dont know what this MTM configuration is it means mid tweeter mid in a nutshell 

Reality check in:

This is far and beyond the comprehension of 99.9% of amateur box builders and people making and selling line arrays at least in Southern Africa, sorry boys but you are kind of selling junk and selling something that even "fong kong" speakers get right.

Increased complexity and increased cost are the negatives, the positives are mutual coupling to some degree with increased and improved lobbing characteristics.

Remember line arrays are supposed to help reduce comb filtering over a traditional speaker system and help us direct more sound in a given direction which we could make reference to as lobing or think of it as a polar response in another fashion as we are bringing a bunch of enclosures together which will each have a unique polar response and in some form or another will effect each others polar response characteristic.

Alright which brings me to another point I want to highlight about cost, sure I can build you a line array cab for R2000 but I wont & I never will because I wont waist my time building something junk and neither should you buy a line array enclosure and throw in speakers and would be better off buying some good quality ready made gear from a respectable brand. If you are going to commit into building your own line array it is important to be able to  capture real life data such as Frd & Zma data which is the electrical and acoustical response of the system, without basic tools and information to get it right it's not something  you should attempt or be willing to buy if the other person cant provide it to you.

Right the sad part out the way, this wont stop us from kick starting a cool project off for a customer who whish's to remain unknown, his objective is for us to obviously build a line array lol

I'm going to use this as part of our tutorial covering a wide bunch of aspects about product selecting  and the process we will follow in a guide format.

So in order to meet our customers wish's I need to define a couple of questions with some additional extra's because of the application which start like this!

1.  What size speakers are we going to use?
2. How many enclosures are we going to build per side?
3. Budget?

10 inch speakers, 2 per enclosure with a total of 8 enclosures per side as good as we can make it without costing the same as a branded product.

knowing this information allows me to define how much sound can I control in the direction it needs to go, the more enclosures we have the lower in frequency we can send forward but it can also present us with the solution as a problem in reverse as balancing highs can also be tricky. The semi open budget means the customer wants to get it done right and is willing to work with us to define and refine the solution to achieve the end result.

Though true expectations is to go beyond what is being made by local builders and in reality this is a simple task simply due to the fact we can make custom crossovers as measure phase and you know design a little better to an end product. I naturally leaned into the Nexus Evo range of speakers which we had designed to be competitive with the best of the best leading brands in performance but without the premium price tag that is unfortunately associated to our international counterparts. The Evo 10" mid bass driver is a 16 ohm 3" inside outside voice coil neodymium driver that has to be one of the coolest little drivers ever, with low levels of distortion thanks to the demodulation ring it has a wonderful dynamic through the vocal range which it exerts with ease. A usable range from 80hz to 3000hz we are going to attempt to exploit from 110hz to around 2.2khz, I want to keep cone displacement to a minimum overall in the tightest and most compact enclosure possible and lucky for us the driver itself wants to be loaded like a mid into a micro enclosure. The huge problem is can we fit it all together and what challenge are we facing to get to work with what is pre defined?

Customer has outlined that he would like to use an existing horn flair - not ideal but this is a starting to point. a typical shape of a generic 10" array from the side view is truncated with the face height rolling in around 300mm with a depth of 400mm and the rear give or take 230mm or so, this shape is almost pre determined because of the brackets itself. So we have some dimensional data to start off, remember that driver spacing we spoke about earlier or 10" driver roughly equates to 254mm is size + our material on the top and bottom 15mm birch plywood we already using 285mm worth of our height. In an ideal world world we would want a single horn for the whole array which is not practical in the living world so we very much want to get our horn flair to the very top and bottom end of our enclosure.  Ideally we want to be around 290 to 300mm to get as close as possible to the next enclosure. I have seen and continue to see very poor examples of people using the wrong size horns!

The existing horn has a dimension of 235mm tall by 240mm so we can now estimate our enclosure volume  in fusion 360

Remember guys proper planning prevents poor performance so to more we lay a foundation the easier this will all become in the end, so our rough sketch needs to include the thickness of our materials and include potential items like driver volume displacement which must include things like the horns act. These items remove volume from the enclosure and we want to try nail the tuning as best as possible, you will see in the rough sketch I have pushed the face backwards to accommodate the frame of the driver the and the horn mounting flange, its not absolute but again a basis of our start to work from as it's almost  guarantee we are going to make changes !! 

Right now we can see a generic sketch, the face is pushed into the enclosure by 20mm with our plywood being 15mm in thickness we can see if our speaker will even fit, in this instance we have 266mm to work with and we know our driver is 261mm and our horn in total diameter with our horn being 235mm. Good stuff so far the next step is to add the widths of the components and turn our basic sketch into something that is 3d in nature.  I know I have not checked the horn flair + compression driver depth yet which could bight me in the rear end but for now it's ok ;) 

Driver total diameter is 266mm but I want to give myself a little wiggle room so we can make it 270mm this gives me 4mm spacing, our horn flair is 335mm so lets add things up and keeo a 4mm space between each part!

Side wall  4mm 

Speaker 270mm

Horn 335mm

Speaker 270

Side wall = 4mm

grand total of 883mm internal width, so lets get going back into fusion and turn it into a shape from our sketch and I'm going to leave the top off and the one side of the enclosure off for this example

Next thing I want to do is extrude the internal volume which we are going to turn it into a body to get the internal volume, with an estimate volume I know we are going to have some parts take up space but I can already start to estimate the enclosure tuning to a point which is helpful. So creating the body looks like this:

Our volume is 0.072 meters cubic or we can also say it is 72 liters roughly without anything loaded, looking at the specs of our driver.... Huston we have a problem!

For all that and more check out  part 2 of this series but for now I will leave you with a couple pictures of our little 10" - thanks for following The Sound Guy folks, love peace and respect to you all.