While walking down the hall around lunch time I walked in on a discussion about surround sound home theater systems. After humbly admitting I watched movies through my two channel stereo I was prodded by my peers to get with the '90s, I started checking out the technology.
The following week I saw a used Carver surround sound processor for sale on the net. Perfect for upgrading my late 80's vintage two channel stereo. However, a suround sound system requires three front speakers (left, center, right), two rear (left and right), and a subwoofer. The five channels provide the spacial realism, while the subwoofer provides the physical realism... that is when Apollo 13 takes off one should be able to feel the vibrations in the chest, and when the horses cross the battle field in braveheart, it's the subwoofer that allows one to feel the horses through the vibrating floor.
While I had other equipment to get by on the five channels, I was short on a subwoofer, so I did some research. The short story is that I built my own powered subwoofer (actually 2) and learned alot in the process.
Here are the details...
Some info on subwoofers...
Difference between a woofer and a subwoofer... subwoofers are usually
only used to recreate a sub portion of what a woofer will. While a typical
woofer may have a frequency response between 20Hz and 2k Hz, a subwoofer
usually has a crossover that limits it's frequency response range from
20Hz to maybe 80-150Hz. Component
wise (the cone/coil thing) woofers and subwoofers are the same.
There are passive subwoofers and powered subwoofers. Passive subwoofers
are powered by your stereo's amplifier. Powered subwoofers is the same
as a passive subwoofer except it has it's own amplifier for driving
the passive subwoofer. Your stereo needs to have a sub woofer out (line out)
for a powered subwoofer. Since most power produced by your stereo's amp
is consumed by driving the low frequencies of your speakers, it's a big
win to have a separate amplifier for driving the low end. This way mid
range won't suffer from the power consumed by the low end.
There are several designs for a subwoofers. Some of the less exotic ones are are ported, sealed, and band pass enclosures. Sealed enclosure's advantage is that they can exist in a much smaller enclosure, however this is at the cost of efficiency and reduced low frequency performance. Ported enclosures are the most common. These have a tube called a port which serves several purposes. It's important that this port be tuned to the driver (woofer) and volume of the box or else the sound may result in unclear, muddy bass. The affect of a properly tuned port with the woofer is similar to blowing a wine jug like an instrument. A band pass enclsure looks like a box with a hole or slit. Bose subwoofers are band pass woofers. The idea is that only certain desired frequencies (bands) end up passing through the hole or slit while the others are absorbed internally.
There's a thing called a crossover to consider. A crossover is like a frequency filter except rather than filtering out frequencies, it redirects frequencies to other channels. A typical powered subwoofer will have a built in crossover that directs low end frequencies (usually < 80-150Hz) to your subwoofer while sending higher frequencies on to another channel (the rest of your speaker system). The high end frequencies rapidly diminish in the channel sent to the subwoofer and the low end frequencies rapidly dimish in the channel going to the higher frequency response speakers. So the tweeters get the high frequencies they are designed to produce and the woofer get the low frequencies they are designed to produce. Almost all multi speaker system will have some sort of crossover. separating frequencies between tweeter, midrange and woofer. In crossovers, there are both passive and active crossovers. Passive crossovers are typical in most speakers. Since they take their power from the incoming signal, they do have an impact on power/responsiveness. Active crossovers get there power externally. However we're starting to go beyond the scope of the subwoofers project.
The prices for powered subwoofer can get pretty expensive. I was looking at models between $200 and $1500. $1500 for a 12 or 15 inch woofer plus amplifier just seemed ridiculous considering the cost of parts to make them... So, I decided to look into making my own. The route I chose to take was not cost effective if time for learning, designing, and implementing is factored in. However what I've learned while doing this makes it all worth while.
Speaker Building Information Sources
I started with two books I picked up at Radio Shack called Speaker Building
and Advanced Speaker building. Both are heavily oriented torwards building
speakers with Radio Shack speaker components, and can almost double
as a Radio Shack speaker catalog. Lot's of shameless Radio Shack marketing
through out. They have a nice intro into building speakers, with a lot of
the theory left out. The first book contains graphs to crossreference
parameters required to make a speaker so that one doesn't need to use
a calculator. The advance speaker book throws in some formulas but is a
little lacking in discussion on how the equations were derived.
In case you don't know how to plug a formula into your calculator, no worries...
the Advanced Radio Shack book will tell you which buttons to press.
Vance Dickason's "The Loudspeaker Design Cookbook" (ISBN 1-882580-10-9) was recommended by several sources. It reads like a collection of white papers on speaker building theory and is heavy on the technical side. Towards the end it contains several commercial advertisements for part sources. Good book.
There is quite a bit of wave theory, frequency dampening factors that needs to be taken in to consideration. Everything from cone tensile strength to acoustic frequency dampening of dacron fiber fluff or maybe beach sand, resonant frequency of particle board versus a harder wood, etc., etc. factor into the volume of the box and the optimal length of the port so that it resonates properly. So many factors, it takes some time to become comfortable with weeding out the meaningful parameters from those that are negligible in the quest for making the ultimate speaker. Box dimension I later learned really dosn't matter much for a subwoofer. The main concern in box dimension is standing waves where waves are cancelled out to the tune of +/- 5dB in the worst case. Sufficient dampening material inside to absorb standing waves can cancel this out. In the case of subwoofers, it's not an issue since the box dimensions are so much shorter than the wavelengths (20-100Hz).
Here are some information sights I've found on the web....
Being a neophyte in the audiophile world, my first thought was to get the
biggest driver I could find to produce the lows and push volumes of air
that one could feel.
I ran across some 18" woofers, however I liked the specs on some
15" woofers I later found better.
After doing the calculations I found that for my ported cabinet, I would
need an enclosure that was 14 to 24 cubic feet! I decided that there
is no way I would want to dedicate this much space in my living room to
a box this big. These calculation were based off of an optimal volume
formula I got out of the Radio Shack books. I later discovered that
this optimal volume didn't agree with the manufacturer, several software
design packages, and the loudspeaker cookbook.
An alternative was to go with a sealed enclosure, however this reduces the responsiveness and pushes the F3 upwards. End result is a subwoofer with a reduced ability to produce low end. Since it seems the whole point of having a subwoofer is to produce the low frequencies, a sealed enclosure which raises the low end didn't make much sense to me.
I opted for a 12" woofer with a 3 cubic foot enclosure. I decided to go with Ultimate's AU1290 "Attitude" woofers. They have a frequency response range of 18-1k Hz. They have a dual six ohm, 90oz magnet, 2" kapton coil capable of handling 500 watts sustained, 1000 watts max. Headroom is .44". SPL is 88.9dB, but I don't care about this since it'll be driven by it's own separate amp with it's own gain control.
Test Data from Factory: SPL 88.9dB 1W/1m Fs 23.5Hz Vas 5.3 Qts .38I decided to go with 3 cubic feet ported enclosure since that would give me the lowest frequency within factory recommendations.... With a 3 cubic foot enclosure, this driver has an F3 of about 26Hz (pretty good!). I decided to go with a 4" diameter port tube since that would give me the least amount of port noise. The port for this is about 14" long. I ended up picking up some PVC tube at the hardware store. This worked quite well.
I decided to make a prototype out of particle board since wide oak boards
are expensive and I was afraid of ruining them, especially since I have
very little wood working experience. Particle board cost about 1/8 the price.
Being new to wood working I only had a circular saw which I had learned through experience is not for precision wood working. I was able to pick a used craftsman table saw in perfect shape for about a third of the price of a new one. And this is what I built the prototype with. For the speaker and port hole I used a saber saw.
The prototype came out pretty good. Some things I learned from the prototype is that cross bracing is important so as to cut down on the side board vibrations. Also, all seams should have a good fit and must be well sealed. I used some silicone wood calk I picked up at the hardware store. If the seams aren't sealed, the gaps tend to make a raspy whistling noise when the driver is pushing large volumes of air.
I picked up a used 60's vintage radial arm saw which I found for about 1/5 the price of an equivalent new one, however it needed some work. I spent a Saturday cleaning, oiling, calibrating, and building a new rip fence for it. I discoverd that a radial arm saw is really the way to go for precision cuts. Rather than moving a big block of heavy wood towards the blade, the heavy material stays in place while the blade runs along a track and makes a perfect precision cut. The end result is that the oak boards had a much smaller seam between the boards in the cabinet. I think this will help reduce board vibration as well as make it easier to seal.
I put about 2 pounds of a Dacron fiber called "Acousta-stuff". Supposedly it's better than Dacron fiber that one can get at a craft/fabric shop because the fibers are triangular rather than round..... Don't know about that, but the Loud speaker cookbook had some test results that said it was better, so I paid $8 per pound for this fluff. Fiber glass is suppose to be better, but for other reasons I decided I didn't want to use that. Dampening material is important for the speaker so as reduce standing waves in the enclosure which can mess with sound to the tune of a variable +/- 5dB! I just stapled this to the side of the walls excluding top and near the port tube.
I picked up a gold plated 8 post screw/banana speaker terminal which is intended for a dual voice coil subwoofer with individual crossover for each channel for hook up to additional speakers. Currently I use the active crossover from the subwoofer out in the Carver surround sound processor, so I've delayed on the passive crossover for now.
I have some small Polk speakers (4b?) I'm using as the satelite front channel speakers now. They have a 6.25" woofer. I don't have any performance spcse on these speakers.... so I'm guestimating by looking at other 6.25" of comparable quality that they will begin their roll off around 100Hz. I'm considering putting together a 2nd order low pass filter at 100Hz for the subwoofer. I had underestimated the cost for a crossover for this subwoofer. Doing some rough calculations for this low 100Hz filter looks like I'll need a 12mH inductor coil and ~200uF capacitor. Because of the high power nature of this subwoofer, I'm looking at $15-$30 for the inductor, and ~$50 for the capacitor.... Since this is a dual coil driver, everything will be mulitplied by two.
I needed a mono amplifier to drive this subwoofer. While a cheap and easy
way to do this would be to buy a bridgeable automobile amp, I had been looking
for an excuse to build my own amplifier.
Acoustic Labs has designed several high end amplifiers. Their designs
are then marketed as "high end audio" project kits. Mark V Electronics, Inc.
out of California markets these (800) 521-MARK.
A mono amplifier(s) has a lot of uses in a component stereo system,
especially in a 5.1 channel AC-3
or Dolby Pro Logic (THX) surround sound
system that I can see using it for the forseeable future. As such,
my thinking was... if I'm going to put all the time into making this, why
not start with the best possible design.
I decided on the AF-3 design which is a 300 Watt constant current MOSFET amp.
MOSFET amps are known for having very similar characteristics to pentode
vacume tubes without the drawbacks of being a vacume tube. The design
has a built in speaker protection cicuit relay for protecting the speaker
from power surge during power up. The design, PCB, mega heatsinks,
and components that go on the PCB sell for $165.
Not bad I thought... that is until I picked up
everthing else I needed. The design calls for 4 filter 80-100 volt 10k uF
capacitors which sell for about $25 each with mounting hardware. I picked
up an 8 amp toroidal transformer (low hum, low EMF) for $66. Also I needed
a cabinet of which I found a nice black anodized aluminum vented case with
rack handles for $44. Other miscelaneous components such as an optional
(quiet 23dB) fan, gold plated mega 4AWG capable speaker binding post,
wire, rectifier bridge, fuse holder, switch, some extra capacitor and high
wattage resistors, etc. brought the total price up to roughly $400 for this
mono channel amp. Had I known this would be the price ahead of time, I
probably wouldn't have done it. However, I now rationalize that I've
learned alot making it and that I have the best mono amp for my forseeable
I built two amps (Mom wanted one). Each amp took approximatly 50 hours assembly time. The end result is a nice sounding extremely powerful amp. 300 watts for this subwoofer is more than adequate. I have the subwoofer line out on the surround sound processor turned down to about as low as it will go. The subwoofer itself will handle 350 watts sustained, 500 watts peak. Watching the movie "Braveheart" during some of the battle scenes tends to set the whole house vibrating. One can feel the horses running across the wooden floors in the house.
After a couple weeks, I noticed the first amp I built would run much hotter than the other. I opened up the amp and recalibrated. There are two potentiometers that need to be adjusted. One is adjusted by reading the current to the amplifier board, the other is adjusted by zeroing the ouput voltage to the speaker when no input signal is present. I had adjusted these after assembling the amplifiers. However after running the amps for a while, one of the two potentiometers needed recalibrating. Now they seem to run at a reasonable temperature with my typical use.
I had picked up a 120 volt cooling fan at Radio Shack for cooling the amp. At the time, I didn't know how loud 38dB was which is what the fan noise was rated at. After trying it out, I decided this was too loud for an audio amplifier application. I found a 23dB fan + fan guard in a Jameco Electronics catalog. It's quiet enough, however, cutting the hole in the metal case has become a problem. I found a 3" circular saw, however this requires a drill press that one can set at 100rpm and can apply 150 lbs of pressure. I don't own a drill press, and haven't found one that can operate as slow as 100 rpms. I just finished a welding class, where in addition to stick welding, MIG and TIG, we learned to use a plasma cutter which is as easy as using an exacto knife. I think if a circular wood template was made ahead of time, the plasma cutter would be the easiest. I think the local welding shops rent a 110 volt plasma cutter for about $40. ...an expensive hole. I also needed to cut the case to install the oversized speaker binding posts. I used a coping saw for this. It was hard frustrating work and the results were not very nice. Fortunately, the cut is hidden by the speaker post mount, and so doesn't matter. A plasma cutter would have been great for this too.
Even though components in the amplifier would probably last longer running cooler, I've decided to wait on installing a fan until I have a better understanding of how hot it will run and can justify all the trouble of installing one. This also gives me more time to think up some way to wire up a thermostat controller.