DAW Techniques

Building your own PA cabinets

Part One

Published in PM March 2009
Printer-friendly version Printer-friendly version

Technique : DIY
In this, the first part of a short series on DIY PA cabinets, we explore why you would want to self-build, and give you tips and advice on how to set about doing it.
David Mellor
Photo by David Mellor
There are some things that you just have to do at least once in your life — for example, see a total eclipse, perform a parachute jump, and make your own set of loudspeakers. Actually, I might give that parachute jump a miss. But I’ve done the total eclipse. I remember the immense sense of collective disappointment from the thousands of would-be viewers gathered together at Parc l’Eclipse (really!) near Cherbourg, when a cloud obscured the event from view at the last moment. I can also remember the massive sense of disappointment when I heard what my first self-built loudspeaker sounded like. More on that later.
But why would anyone want to build their own loudspeakers when there are so many excellent products on the market these days? Well, the answer to that is probably a desire to really get to grips with the technology we use to make music. Buying stuff and using it is all very well, but somehow there is an inner urge that isn’t fulfilled. And although making a musical instrument would satisfy that urge perfectly well, it’s a whole lot more practical to make a loudspeaker. I could have said it’s a whole lot easier. Well, to design and build a really good, world-class loudspeaker would probably be as hard as making an acoustic guitar, and it is obvious how difficult that would be. But it is quite easy to make a simple speaker, and there is every chance that it actually will sound quite good.
My moment of inspiration for writing this article was when, one day last Summer, I passed by a field near the village where I live. I noticed a small sign, pointing into a field. ‘Arcane Festival’, it said. “What could that possibly be?”, I thought. Well, I ventured into the not-too-muddy field and at the far end I could see a small crowd, and I could hear the beat of music. As I got closer, the music became more defined, and by the time I had made my way to the front it sounded rather good.
There was only one speaker stack that I could see, which I fortunately had the presence of mind to snap a photo of, which you can see above. As you can see from the photo, it doesn’t exactly look like the product of some slick manufacturing operation. Indeed, it has a very ‘home-spun’ look. There is an air of randomness about how the boxes are piled up, with not a great deal of science involved. And that certainly is an interesting way of enjoying the bass.
That got me thinking. Over the years I had become accustomed to the idea that speaker designers are really clever people. That their boxes and systems really are very high-tech, and it wouldn’t be possible to self-build loudspeakers that could sound even remotely as good. Yet this stack sounded great! I wouldn’t go so far as to say that it was anywhere near as accurate as a good hi-fi or ‘designer’ PA system, but it just sounded really good in the context in which it was set.
So although the principles of loudspeaker design are now very well known to the experts, and are applied with precision, there is still a strong case for building loudspeakers ‘by feel’, and stacking up the boxes any way you like to get a sound that pleases you and your audience. Rule-book loudspeakers are great to have available, but home-brew loudspeakers have just as valid a place in the universe of sound textures.
In the beginning...
Now, getting back to my own first loudspeaker, the first of quite a number It was back in the hazy days of the 1970s, when the rule book of loudspeaker design was still some way from completion. It was then the fashion for guitarists to have a ‘stack’, comprising an amplifier and two 4 x 12 loudspeaker cabinets. I would have liked to have a Marshall stack, just like Jimi, but I couldn’t afford one. So I bought eight 12-inch drive units, the cheapest I could find, and several sheets of chipboard. I bought the chipboard because it was the cheapest material I could get, but I later learned that it is actually quite a good material for loudspeaker cabinets.
So I sketched out a design for a cabinet that was rather larger than the Marshall equivalent. I don’t know why I did that Actually, yes I do — I just wanted my stack to be bigger than anyone else’s! Then, over a couple of evenings, I put that first cabinet together, covered it with a cheap Rexine imitation, and fitted the drive units and handles. Finally, I wired up the drive units to the jack socket and screwed on the back. I have to say that it looked great. Of course I couldn’t wait to hear it, so I plugged in my amp and guitar and performed my best Pete Townsend power chord.
Er there was something wrong. It didn’t sound good at all. All the drive units were working but the sound was just wrong. Over the next few days I came to the conclusion that I didn’t know as much about loudspeakers as I needed to, and decided to cut my losses and sell the cabinet. So I advertised it at a price that just about covered the cost of the materials. I soon got an enquiry from a local working mens’ club, as they were called in those days. Since I’m an honest trader (with 100-percent positive feedback on eBay!) I needed to give them a demonstration so that they could properly consider what they were buying — a speaker that worked but wasn’t all that good. So I took the cabinet round to the club and played some music through it. The committee were satisfied and said they would have the pair. “The pair?” I don’t know how that happened but, since I had the parts for the other speaker, I realised that if I wanted to get my money back, I would have to make another one.
So we shook hands, although they did take the opportunity of exploiting my youth and innocence and knocked down the price to about two-thirds of cost. I set to work building the other speaker. Once finished, I had to test it of course. And it sounded great! It was exactly what I wanted a 4 x 12 cabinet to sound like, and I reckon it would have given the Marshall a run for its money. So I looked at the first speaker again to see what had gone wrong. I spotted the error straight away, when hours of looking hadn’t helped just a few days ago. I had wired the four drive units in the conventional series-parallel way (I’ll explain that in a later installment), but two of the drivers were wired in reverse phase. So effectively at any instant when two of the drivers were pushing at the air, the other two were pulling. I corrected my error, and I now at last I had my brilliant twin 4 x 12 stack. Except I had agreed to the sale and later that day the club sent a van round. I never did make any more 4 x 12 cabinets, but at least I knew that I could if I wanted to.
The science of speakers
An X1 folded-horn design during construction (above) and completed (left), by Rog Mogale of Void Acoustics.
An X1 folded-horn design during construction (above) and completed (left), by Rog Mogale of Void Acoustics.
Since this mini-series is all about home-brew loudspeakers, I’ll keep the science on the ‘lite’ side. Until comparatively recently there wasn’t a great deal of science involved in cabinet design, just a lot of ‘try it and see’. Firstly we need to go back to the reason why loudspeakers need cabinets in the first place. After all, why wouldn’t a bare drive unit work just as well? The answer to this is that a drive unit radiates just as much sound from the back as it does from the front. No one has found a way around that, yet. And since the sound from the back is inverted in phase compared to the sound from the front, it creeps around to the front and partially cancels out the sound that you want to hear. So the drive unit can be working really hard, but you only get to experience a fraction of the sound power that it is producing. The solution is to contain the output from the rear of the drive unit in a cabinet of some sort and not let it escape into the outside world. But this creates a new problem. If the cabinet is sealed, then the air inside will act as a spring that opposes the motion of the diaphragm of the drive unit. As the drive unit pushes out, the air inside will be at a lower pressure than the air outside, therefore the diaphragm will experience a force pushing it back in. The reverse is true for when the diaphragm moves inwards.
Fortunately this problem can be lessened by tuning the size of the cabinet to the various physical parameters of the drive unit, which we will look at later. If you want to be precise about this, you can plug the numbers into a set of equations that will tell you what volume the cabinet should have. Alternatively, you can just build a big cabinet. That works almost as well.
This ‘closed-box’ cabinet will work well. But it isn’t as efficient as it could be. An alternative design is the bass reflex. In this there is a tube that connects the inside of the cabinet to the outside air. If the dimensions of this tube are precisely calculated, then the cabinet can be made to resonate at a frequency that is cunningly placed where the output of the driver is falling off. Thus the bass-reflex design produces more bass. Do you have to calculate the dimensions of the tube precisely? Well, if you are going to use a tube (called a ‘port’ in the business), then probably you should calculate, otherwise the result might be unpredictable in a less than satisfactory way.
There is also the open-back cabinet. There have been plenty of open-back cabinets through loudspeaker history, and they can sound quite pleasant. Partially opening the rear of the cabinet to the outside air can relieve some of the pressure on the diaphragm, although not producing the same benefit to low-frequency output as a bass-reflex design. Many guitar cabinets have a partially open back. Why? Because they seem to sound good on guitar, and the guitar doesn’t have enough in the way of bass content to positively demand a bass-reflex design.
Closed-box, open-back and bass-reflex cabinets are part of a class of loudspeakers called ‘direct radiator’, where the diaphragm pushes directly against the air. But the diaphragm/air interface is a boundary that sound does not cross with perfect ease. So to make it easier for the sound wave to enter the air with its full might and vigour intact, we can couple the diaphragm to the air by means of a horn. The gradual taper of the horn smoothes the transition between the acoustic characteristics in the area that is being actively vibrated by the diaphragm, to the wide open spaces of the air in the venue.
A high-frequency drive unit doesn’t need a huge horn to connect to the air, so the horn can easily be incorporated into the drive unit itself, or mounted on the front. Similarly, a mid-range loudspeaker can manage with a small-ish horn crafted into the cabinet in front of the drive unit. A low-frequency loudspeaker, however, benefits most from a big horn (don’t we all?) so this is most commonly built into the internal structure of the cabinet itself, in some cases taking advantages of loopholes in the laws of physics to get benefit from the rear radiation from the drive unit, rather than it being nothing but a problem as described before.
Types of cabinet
Scoop bins, such as this copyright design by Rog Mogale, use a horn-like structure to impose a delay on the sound waves emanating from the rear of the bass driver, so that they reinforce those coming from the front.
Scoop bins, such as this copyright design by Rog Mogale, use a horn-like structure to impose a delay on the sound waves emanating from the rear of the bass driver, so that they reinforce those coming from the front.
So let’s take a look at the types of cabinet you might consider for a self-build. Firstly there is the open back. The advantage of this is that it is so rough-and-ready that you can throw one together with virtually no design involved at all. There’s also a very good chance that it will sound excellent on electric guitar. If you use a large drive unit and make the cabinet bigger, it may well show a decent quality of tone on bass guitar, although this design is not really appropriate for low bass due to cancellation caused by sound leaking from the partially open back. I have heard excellent sound quality from cabinets of this type using 15cm (six-inch) drive units all the way up to 37cm (15-inch). I did try a 45cm (18-inch) drive unit once, but its lack of definition in the upper frequencies meant that, for me, it was unsuitable for use on its own. However, thinking back, it was certainly an interesting tone colour and might have been nice to hang onto for an unusual sound quality every now and then.
If you aspire to build a bass-reflex cabinet, then the back of an envelope sketch doesn’t work so well. If you simply guess the dimensions of the port, then it will reinforce some random range of frequencies, possibly well below the drive unit’s ability to produce anything at all, or in an area that is already strong. But start flexing your maths muscles now because I will be giving you the formula next month.
In PA, we normally think of low, mid-range and high frequencies as separate entities. The reason for this is that a low-frequency drive unit positively needs a large cabinet to work well. A mid-frequency drive unit benefits from a cabinet, but it does not need to be all that large. A high-frequency drive unit doesn’t need a cabinet at all. The box is simply there to provide a mounting and a convenient means of stacking, transportation and storage. Often in PA, the mid-range and high-frequency bands are combined into one cabinet, but the cabinet is designed for the mids — the HF driver is just along for the ride.
One of the most popular designs for LF cabinets is the folded horn. The drive unit sits inside the cabinet with its rear radiation contained by an enclosed space. The front of the drive unit radiates into a duct crafted from the material of the cabinet, which at this point is narrow. However, the duct opens out gradually, eventually reaching its maximum dimensions at the opening of the cabinet. In this way, an approximation to the ideal horn shape is formed, albeit made from a series of flat panels. It is possible to bend the panels if your woodworking skills are up to the job.
An alternative to the folded horn is the scoop bin. Here, the front of the diaphragm is open to the air. The output from the rear is channelled through a horn-like structure within the cabinet, emerging from the front. The time delay involved in traversing the length of the horn means that the output from the rear of the diaphragm becomes useful, rather than a problem as in other designs.
The sound quality of the scoop bin is interesting though. I remember once visiting the Notting Hill Carnival where you are likely to hear the sounds of several systems blending from different directions and distances. But on this occasion I could also hear a mysterious rumbling sound, but I could not tell where it was coming from. As I walked, the sound got louder and louder until it eventually reached a peak and started to diminish. I looked back, and saw a speaker stack I hadn’t noticed at first. When I went really close up, I could be sure that this was where the rumble was coming from. But honestly, there was so little definition in the sound I could hardly make out that it was meant to be music. It was, however, mightily impressive, and this is one of the factors that leads me to realise today that loudspeakers don’t all have to aspire to be ‘perfect’ — ‘interesting’ can be just as good, and often better.
Another cabinet design
A DIY PA stack comprising scoop bass bins and mid-range horns, topped off with some small cone drivers and piezos.
A DIY PA stack comprising scoop bass bins and mid-range horns, topped off with some small cone drivers and piezos.
Photo by Rog Mogale
suppose you really want to focus in on a range of frequencies and blast them out into the room. You need a band-pass cabinet. Now this is an odd one. It looks like a speaker cabinet but it doesn’t have an opening for the drive unit. Instead, it is completely enclosed apart from a port, similar to the port of a bass reflex cabinet. Inside it’s like a bass-reflex cabinet, but instead of the front of the drive unit facing the open air, it fires into another cabinet! This front cabinet can be sealed, or it can have a port. Basically, the object of this enterprise is to make a cabinet that is really efficient over a narrow range of frequencies. A typical use for the band-pass cabinet would be low bass, where there is another separate cabinet to handle higher bass. As we can see, people take bass very seriously indeed.
The philosophy of self-build
Adopting my best Kevin McCloud posture, I would say that the philosophy of self-build is firstly to learn what loudspeakers can do. OK, you can buy them and learn something of what they can do. But it’s not until you have got to grips with the details of drive unit selection and cabinet design that you can really link the components and construction with what the loudspeaker will actually sound like. Add to that the premise that good loudspeaker design doesn’t necessarily mean a faithful recreation of the sound source. It can simply mean something that sounds good in its own right. Perfection is nice, but it can be boring. It’s a micro-universe in comparison with a much greater universe of loudspeaker designs, which are much more interesting to explore.
In next month’s installment, I will be looking at drive units and their parameters. There will be more about cabinets, and I will be talking to a real expert in loudspeaker design.
Thanks to Rog Mogale of Void Acoustics for his insights, which were very helpful in the preparation of this article. I will be interviewing Rog next month, and you can see a photo of his motorbike!
Published in PM March 2009