Introduction

Cerberus is a character from Greek mythology that is a three-headed monster hound who guards the entrance to the Greek underworld. So despite the small 6.5" driver this woofer has a lot of bark. The project uses the Tang Band W6-1139SC 6.5 subwoofer along with the Dayton SA70 70 watt subwoofer plate amplifier. The enclosure is an 11" vented cube tuned to 36 Hz with an f3 of 35 Hz. At its rated power of 60 watts the Cerberus can reach levels just over 100 dB with no excursion problems until you get below 25 Hz. I designed this subwoofer to be a mate to the Microbe design and the output levels are well matched. The complete design can be built for around $100 which is quite a bargain.

Drivers

Tang Band W6-1139SC 6.5" Subwoofer
Amazing bass in a compact package! This one-of-a-kind 6-1/2" subwoofer has an impressive 11.5 mm of Xmax. A wide roll surround, raised flat spider, and large 1-1/2" 4-layer voice coil ensure impressive bass performance and high power handling. The vented pole and vented spider allow good airflow around the coil to improve heat dissipation and reduce compression. The compact neodymium motor structure saves valuable enclosure space while also allowing closer-than-normal placement to CRT's. The ultimate multimedia subwoofer driver!


Specifications: Power Handling: 50 watts RMS/100 watts max *VCdia: 1-1/2" *Impedance: 4 ohms *Re: 3.6 ohms *Frequency range: 35-800 Hz *Fs: 35 Hz *SPL: 85 dB 1W/1m *Vas: .51 cu.ft. *Qms: 2.81 *Qes: .32 *Qts: .32 *Xmax 11.5 mm *Dimensions: A: 7", B: 5-1/2", C: 3-5/8"

Cabinet Design

The Cerberus cabinet is an 11" cube with about 12 liters of internal volume once you factor in the space taken by the port and the amp. The port is a 1.5" diameter flared port. The port tuning is 36 Hz and results in an f3 around 35 Hz. The cabinet is a bit larger than typically required for this driver but it produces an EBS (Extended Bass Shelf) alignment that lowers the f3 by using a larger cabinet with a lower port tuning frequency. The larger cabinet causes the response to rolloff higher in frequency but at a lower slope. The key with this type of design is to tune the port low enough so that there is little or no peak in the response at that frequency. My design has a slight peak in this region but it is not severe. The power handling for this type of design might actually be better at lower frequencies due to the lower port tuning frequency. You might think that the port diameter is too small for a 6.5" subwoofer and there is a little bit of port noise with bass notes in the 30-40 Hz range at high levels but with music playing the port noise isn't noticeable (I only noticed it when I ran high level bass sweeps in that frequency range). If you wanted to try a larger port you would have to angle it to the left about 6-7" into the cabinet and you might be able to fit a 2" flared port but you would also probably have to move the driver to the up and left to make room for the larger port. I'm wondering if a modified version of this design using one or two 6.5" passive radiators (I like two for symmetry) on the side(s) of the cabinet might sound better because you'll still get a decent f3 but won't have to deal with the port noise although the passive radiators would probably boost the cost slightly.

The cabinet can be made from two pieces of 12" x 48" x 3/4" MDF shelving or a single piece of 2' x 4' x 3/4" MDF. There is a board cutting diagram on the blueprint. A grill is optional but I chose to include one since there is plenty of extra MDF. If you choose to use a grill you'll have to buy some grill posts and grill cloth. The amplifier had to be offset to one side to make room for the port. The 1.5" flared port is available from Parts Express but only in 4" lengths. I was going to use two of them back to back and have flares on both ends but the flared end on the inside interfered with the amp placement so instead I only left the baffle side of the port flared. To get the 7.5" length I took the extra port and cut off just the flared section right up to the edge of the flare "fins". Then since the flared ports are tapered I placed the cut end of the cut port onto the end of the full flared port and it fits snugly with a little bit of overlap with a nice and snug fit. If you want you can glue or tape them together like this. The port hole opening is a tad larger than 2.5" in diameter so I just cut the hole to 2.5" and used a spindle sander to make it just large enough for the port to fit snug.

I used Parts Express cherry vinyl laminate to finish the cabinet. You can finish one cabinet with about 34" of vinyl, which comes in 24" wide by 10' long rolls. First cut the 24" wide piece of vinyl into two 12" strips. I wrapped the first strip from one side of the cabinet to the top of the cabinet to the other side of the cabinet and then trimmed the edges. Then I wrapped the other piece from the front to the bottom to the back of the cabinet and then trimmed the edges. The tricky part with applying the veneer this way is getting the vinyl tight in the corners when you wrap it over. I didn't do the best job of this and there is a little bit of bubbling near the corners. I've heard that people recommend using a hair dryer to get the corners nice. I might try this but it might make it easier to use the hair dryer while applying it. Anyways its an easy way to finish the cabinets and a roll of PE vinyl costs about $15. Considering the price, durability and ease of application it is a quick and easy way to finish a cabinet and it doesn't look bad either. Personally I think I prefer the less natural colors like textured silver, textured black and black ash because they don't try to simulate the look of a real wood cabinet but a lot of people seem to like the cherry a lot.

For the grill I took a piece of 11"x11"x3/4" MDF and cut an 11" diameter circle out of the left half. Then I left cut out the center leaving a 1" rim and traced the outer port flange size so I could cut around that on the grill. I used a 1/2" roundover bit on the front of the grill to give it a nicer edge. Then I strategically located 4 grill posts for mounting the grill... I used the small PE peg and socket grill posts. I finally covered it with some black grill cloth.

Performance

The plot above includes a set of measurements taken at 10" away which is far enough away to include the output of the port in the response. The various traces represent the frequency response with different subwoofer amplifier low pass filter settings. The adjustment knob has markings spaced similar to a clock layout with the minimum setting of 40 Hz at 7 o'clock and the maximum setting of 180 Hz at 5 o'clock. The traces are as follows: black-5 o'clock (max), navy blue-3 o'clock, blue-1 o'clock, purple-11 o'clock, red-9 o'clock and maroon-7 o'clock (min). According to the measurements shown in the plot the actual crossover frequency for this combination is a bit higher than what is indicated on the knob. For example, the 120 Hz crossover point that is required for good integration with the Microbe is located at 9 o'clock on the knob setting and according to the knob markings this should be 68 Hz but in reality it is a bit higher than this. This subwoofer is better suited to a sealed set of speakers and if you know the f3 of those speakers you should be able to adjust the subwoofer crossover frequency so that the subwoofer response is about 3-6dB down at this frequency by looking at the response plots shown and then just set the knob to the appropriate position.
As shown in the diagram above the Cerberus can provide an output level of 101 dB at its rated power of 60 watts. Since the amp is rated at 70 watts you might be able to squeeze a little bit more if needed. However, the Microbes can only put out about 100 dB at max output so the Cerberus is still more powerful than the Microbes at the driver limits. For use with the Microbes or any other small bookshelf design I would plug the ports of the bookshelf speakers (the f3 for the sealed Microbe is 120 Hz). Then set the low pass filter on the sub to match the f3 of the sealed bookshelf speaker (120 Hz in the case of the Microbe). An even better way to connect them is to use the subwoofer output of the receiver to connect the sub. Then you can set you main speakers to "small" and use the crossover in the receiver. I would set the crossover frequency to around 100-120 Hz maybe higher depending on your design. With this method you shouldn't need to plug the ports and there will be greater power handling for the bookshelf speakers because most of the bass will be filtered out. This is the preferred method.

I've only spent a few days listening to this subwoofer (along with the Microbes) and I would have to say that I'm very impressed overall. The thing that really stands out is how deep this little sub can play and at levels high enough that you can actually feel the bass pretty good. I wouldn't say that the bass is quite as clean as a larger 10"+ sealed subwoofer but it is very good nonetheless. If the relative level is kept at a reasonable level and well matched to the output of the main speakers this sub just disappears and it sounds more like you are listening to a large set of three way speakers as opposed to a tiny subwoofer and two small bookshelf speakers. The one complaint that I have is there is a little bit of port noise. However, this port noise was only detectable when playing a bass sine sweep in the 30-40 Hz range at high levels. I don't notice the port noise when playing music though because the overall sound level is so much higher that it covers up any low level port noise that is present.

So the Cerberus subwoofer is a great "music" subwoofer if you are looking to extend the bass response of a small pair of speakers and increase the power handling. If you need more output you can build two or more of them and at around $100 each you can't beat the price. It really is hard to believe how much deep bass can come out of that little 11" cube.

Microbe/Cerberus combination

The Cerberus subwoofer truly is a great match for the Microbe and should reduce the amount of distortion in the sub-120 Hz range. t should increase the power handling capabilities of the Microbe, especially if you use a receiver as a high pass filter. It will extend the f3 from 50 Hz to 34 Hz. I hooked up this combination right next to my Astorius RS speakers to do a comparison and was surprised how good the Microbe/Cerberus combination sounds. I bet if you didn't know what set of speakers was playing and I played the Microbe/Cerberus combo first without playing the Astorius RS you would swear that you were listening to the three ways. The integration is that good and the bass is that deep.

The plot above shows the response of the Microbe/Cerberus combination with the subwoofer filter set at 9 o'clock and the ports sealed in the Microbes. The black trace represents the overall response, the red trace is the Cerberus response and the blue trace is the sealed Microbe response. The purple trace is the original ported Microbe response. The dip in the combined response is a phase error in the combined measurements and doesn't exist in the actual combined response. By looking at the plot you might think that the subwoofer really doesn't extend the response much but going from an f3 of 50 Hz to an f3 of 35 Hz really does make a difference. With the ported Microbes the bass goes deep enough for most music but it just reaches the edge of bass frequencies that you can feel as well as hear. With the Cerberus subwoofer you can definitely feel the bass and there is a lot more impact overall. The improvement is quite significant. Also the bass from the Cerberus is a lot cleaner sounding than that of the ported Microbe design and I suspect that there is less harmonic distortion in the 50-120 Hz range.

In comparing the Astorius RS three way sound to the Microbe/Cerberus combo I noticed a few things. The Microbe speakers have a very warm and musical midrange sound to them. This makes them very easy to listen to and makes them well suited to rock music. The midrange warmth can seem a bit unnatural at times but in general it is a very pleasing sound. The Astorius RS seems like it has a bit more midrange definition which can probably be explained by the response plots and the sound is more natural overall with better midrange balance. However, the Astorius RS three way costs over three times as much as the Microbe/Cerberus combo. Some people may actually prefer the warmer sound of the Microbe/Cerberus combo because it is a bit more musical sounding and not harsh at all. I didn't pay much attention to the imaging and soundstage of the Microbe combo but they still seem pretty focused and have a lot of depth.

So at approximately $300 in parts for a pair of Microbes with a Cerberus subwoofer I think you'll find it very difficult to find a system this small and inexpensive that will perform at this level. They really are quite remarkable considering the parts used. I think of all the designs I've made this combination would be the easiest to market because the performance to price ratio is so high that they can compete with systems several times their cost.

Tang Band W6-1139SG/W6-1139SI driver performance

It seems that Tang Band has once again changed the driver parameters and suffix for the 6.5" subwoofer driver used in this design. The fact is that this design was initially based on the "SB" version of this driver but I found that it still worked for the "SC" version that was sent to me. Tang Band next released the "SG" version and simulations revealed that the original enclosure design still worked well with the newer woofer model, perhaps better than the "SC" version, see below...

I've gone ahead and remodeled the performance of the "SI" version with the original cabinet design and found that it performs more closely to the "SB" version that I based the original design on. Below is a plot of the performance in the current enclosure and it reveals that the new "SI" version of the driver performs quite well in the existing enclosure which is good news.

So as Tang Band continues to change the specifications and suffixes for this driver I'll try to keep on top of things but despite the driver going through 4 revisions (SB, SC, SG, SI) the cabinet has worked well for all of them. The funny thing is that the model that I built using the SC woofer was probably the least optimal combination of the bunch but still worked out well. So if you notice that Tang Band has come up with yet another suffix for this driver please let me know and I'll perform additional simulations to ensure proper performance.