Subwoofer Project - Page two
Design
Since I already knew what spot in my room worked well, it was easy to determine a design that would fit, and blend well into the room. My rather small and less than acoustically ideal room has prompted me to improve the surround field by further processing the surround channels into even more channels. My system uses ten speakers (not counting the sub) crammed into my small 13.5 x 14.5' room, seven of which are visible. I didn't want yet another black box in this room, especially since the sub is in such a visually prominent location. Since my room already had some "gothic" accents, I chose to make the sub look like an architectural column. It is built from 3/4" MDF, with some internal bracing. The exterior dimensions are 15 x 15.5 x 37", not counting the trim. By the time you subract out the thickness of the walls, the internal bracing, drivers, and divider panel, that results in roughly 1.75 cu. ft. of internal volume per driver. That falls at the high end of the recommended size range for the servo kits of 1.6-2.0 cu. ft (note their volume specs INCLUDE the space taken up by the driver and amp). Because the sub will be mounted high, and difficult to get to once installed, I did not provide holes to mount the amplifiers in the sub box. They are mounted below the shelf the sub sits on. This also allowed the box to be slightly smaller for the same internal volume. It should also be noted that since I'm using two separate amps for one sub, I chose to modify the amps to allow one amp to be the "master", and the other to be the "slave". This allows me to set phase, crossover, volume, etc., on one amp. The pre-amp section of the second amp is bypassed. Because this requires soldering and using a dremel tool on the PCB of both amps, this is NOT for the faint of heart, but if you do want to consider a dual servo sub, and want this convenience, you may want to talk to Brian Ding at Rythmik about it. It also requires a modified servo board in one of the amps.
Below you can see the cut panels merely propped up against each other, so you can see the interior layout.
The blocks of wood around the circumference of the driver holes are pieces of red oak that I cut to give the drivers screws something tougher to bite in to. MDF isn't known for holding screws very reliably. There are a couple other cross braces that aren't shown in this picture. Below are views of the back of the sub after all the panels were screwed and glued together:
Here are some more shots after the sub box was painted, showing the inside after it was lined with fiber fill, and the back after the drivers were installed:
Since I was building a dual driver unit, Brian Ding of Rythmik Audio suggested the push-pull driver arrangement, meaning one driver is mounted backwards (magnet facing outwards), while the other is mounted normally. One of the drivers is wired out of phase relative to the other, so the net result is they operate IN phase (otherwise, the output would be mostly canceled out). The reasoning behind this set-up is that non-linearities, primarily from the driver's suspension, can cause distortion due to the differences between the inward motion and the outward motion. For example, it may be easier for the driver to be pushed away from the magnet than pulled towards it, so the resulting motion may not perfectly mirror a sine wave signal it is fed. Reversing one driver causes the NET output from the two to once again be linear, as the deficiencies in a particular direction are cancelled out. There is more information on this type of set-up at the diyAudio WiKi. Some of the information I've seen has suggested that the actual audible benefits of such an arrangement are minor, if detectable at all. And in this case, the servo system should already be correcting for non-linearities, and therefore reducing any improvements that the push-pull arrangement can offer. There is also a potential downside to push-pull mounting: because the magnet assembly of one of the drivers is open to the listening space, any noise generated by air turbulence through the motor assembly can be audible, since it won't be contained within the box. But fortunately, the TCSounds TC2+ drivers are exceptionally quiet in this regard. And because of the way my sub will be mounted, the drivers won't be visible. So there was no downside to this orientation, in my case. I may never know if the push-pull arrangement actually played a part in the clean sound the subs provide, but there's certainly no way it should have hurt. The connectors shown above the back driver in the above picture are for the voice coil connection, and the servo feedback coil connection, for the top driver. Obviously, since the bottom driver is mounted magnet-out, those connections are accessible directly on the driver.
The small clear tube you see to the upper left of the driver is a "just in case" feature I added, to ensure that pressure within the box remained equalized to ambient pressure. For a noise free system, a sub box must be completely air tight. Small air leaks can cause very audible noise when a sub is operating, due to high velocity air finding its way through tiny cracks and holes. If the leak is big enough, it can work towards canceling out the sound from the drivers. But with a perfectly sealed box, barometric pressure changes, as well as internal heat generated from sub operation, can cause a pressure differential between the inside and outside of the box, pushing or pulling the drivers out of a centered rest position. This reduces their output and increases their distortion, because extension in one direction is compromised. While small, undetectable air leaks likely provide the needed equalization, participation in this thread convinced me that it was better to be safe than sorry. The trick is providing the needed bleed hole that DOESN'T cause a performance compromising air leak. I chose to use a small air tube (used in fish tanks) as the equalization port. The tube is 10 feet long, and only has a 1/8" internal diameter. This prevents ANY airflow due to driver operation (based on the same principle that ports use in a ported sub), yet allows a constant pressure differential to be equalized. I probably could have gotten by with 1 foot completely internal to the sub, but since the tube wouldn't be visible, it didn't hurt to use the entire package of tubing. :-)
As can be seen from the above pictures, I didn't waste any energy worrying about recessing the screw holes or excessive glue drippage on the sides of the sub that wouldn't be visible. Because of where this is being mounted, I only had to worry about the appearance of three sides. Below is a close up view that shows it from it's "pretty" side (taken before the drivers were mounted, but they wouldn't have shown anyway). I had originally planned on lining it with granite, but after I learned how difficult it is to miter the edges for the corners, I chose to just sponge paint it instead:
And finally, after lifting this 100+ lb beast up onto it's final resting place, a view of the sub looking in to the den from the living room:
And at night with the system in operation:
May not be everyone's cup of tea, but it at least looks better than a typical black box sitting up there. :-)
Continue on to the performance