Cerberus Subwoofer Build

[Soundslike]

After deciding to build a small subwoofer for my home office, I looked around for a respected design. Noticing the popularity of Roman Bednarek's Cerberus, I decided to go with his design. The sub will be used for music, both in my office and my RV, and probably as a computer sound system.

The Cerberus uses the TangBand W6-1139, a very popular driver as evidenced by the number of builds, and the rather dramatic increase in price. The 11" X 11" X11" enclosure is constructed of 3/4" MDF.

I opted to go with the larger port size, primarily based on advice given on this forum. Roman's original design specifies a 1-1/2" X 4" flared port (PE part number 260-402), a size arrived at for practical reasons, i.e., the limited space within the confines of a small enclosure such as this one. As Roman mentions on his website page (http://www.rjbaudio.com/Cerberus/cerberus.html), the 1-1/2" port can produce audible chuffing, but only when used at very high volume levels. Several recent builds have used a 2" port, 17" long -- the port I elected to use.

As is my habit, I first drew the enclosure, driver, port, and amp, to scale using a CAD program. I take a few liberties that would make my old mechanical drafting teacher shudder, forgoing a maze of dotted lines for the sake of clarity. Here's a bitmapped image of the drawing:

http://i1100.photobucket.com/albums/g414/PCIaint/Cerebus%20Subwoofer/Cerebusasbuilt.jpg

The drawing shows the rabbeted joints (http://en.wikipedia.org/wiki/Rabbet) I will use. I decided to use this technique after reading an early post in what has to be a record length string for the Nano Neos (http://techtalk.parts-express.com/showthread.php?t=218177), and NTN's. Chris Roemer commented that he used a rabbeting technique, and having tried several joint types all having their own disadvantages, I decided to give it a try. I'll be back later to continue with this string. I'm going to make some sawdust...

[50 watt head]

Ron, I just now went back into that thread, to see which rabbet joint you were referring to, and I do remember liking the looks of that particular rabbet joint when I saw those photos the first time. Looks like it would make for easier cab assembly, as well as a good strong glue joint. Looking forward to see how yours turn out :)

[Wolf]

I have to state this since you did all the design graphics, that the design is called the "Cerberus", not Cerebus.
http://en.wikipedia.org/wiki/Cerberus

Just thought you should know.
Later,
Wolf

[Soundslike]

I guess I could claim that I did that intentionally because I put the bigger port in this one, but uh, I doubt that would float. Thanks for letting me know before I got too far into this. I'll be making a few corrections...

[Mikers]

Nice drawings. Here's a bootleg photo of mine, you wanted to see how I routed it. Notice I only did the two top corners, all the rest are 90's. This is to match the ONS. This is down firing so no visable speaker, I like that. There will be 1.5" dia x 1/2" high rubber feet added to the corners.

I chose to go with mitered joints vs. rabbits like I did on the ONS. tough call. Miters are nasty to cut on the table saw as you have so much blade exposed and it wants to pull the work from the fence. You really have to think about you cuts and setup as well so you can avoid running a mitered edge along the fence. Gluing it together isn't too much fun but with some decent corner clamps wouldn't be too bad. I had the advantage of the bottom being dado'ed in to add support while fitting up. On the plus, the part line is right in the middle of the 90's and either method will have a part line on the round over's. So building it is worse with miters but it's better for finishing.


http://farm6.static.flickr.com/5190/5644678179_669a1474e9.jpg (http://www.flickr.com/photos/35875508@N05/5644678179/)

[Randall Kepley]

Ron I cant wait to see this thing going, Cad pix look great to. Have you decided on a finish yet? Paint, Veneer? I know you were taling about this build in another thread and Ive been waiting to see this one take off. Have you made your first cuts yet? R.K.

[Soundslike]

I managed to correct the spelling where Cerberus was misspelled in the narrative, but I may have to live with the title. I also corrected the name on the drawing, but so far, the old uncorrected drawing is still showing on this thread. Probably would load the corrected drawing if I were to reboot.

Roman, if you happen to read this, I apologize for the spelling error...


Ron, I just now went back into that thread, to see which rabbet joint you were referring to, and I do remember liking the looks of that particular rabbet joint when I saw those photos the first time. Looks like it would make for easier cab assembly, as well as a good strong glue joint.

So far, I like the rabbeted joint. It's going to be stronger than a plain b--t joint, and it does help with aligning the panels during assembly. Mitered joints are really difficult to handle -- glue makes things slippery, and as Mikers points out, it complicates the cutting process too. I haven't gotten to the finishing process just yet, but I don't expect to have any problem concealing the glue line. Assembling the box is more confusing than you might expect though. It's a little like a puzzle -- you know it all fits together but get one piece in the wrong place, or turned the wrong way and you'll think you've cut it wrong. Here's another view that show more detail of how the joints are cut -- I should say how I cut them. In this front view, the brown pieces are the top and bottom and the green pieces are sides. The joint looks more complicated than it needs to be -- if rabbets were only cut in the side pieces. But it's actually simpler, if you're using a table saw to cut the rabbets because you only have to set the saw once at 3/8" for the depth of cut, and 3/8" for the width of the cut. All edges of all pieces are rabbeted, and all are cut with the same setting.

The red dotted line shows the length of that particular segment of the port (one of three segments).

I'll add more in the morning...

http://i1100.photobucket.com/albums/g414/PCIaint/Cerebus%20Subwoofer/ScreenHunter_02Apr222148.gif

[Mikers]

Did you leave your edges longer so you can trim them down with a flush bit after assembly?

[Soundslike]

Have you decided on a finish yet? Paint, Veneer? I know you were taling about this build in another thread and Ive been waiting to see this one take off. Have you made your first cuts yet? R.K.

Hi Randall -- I've got the sub assembled, with no decision yet on the finish. It sounds great -- really compliments the Overnight Sensations well. Ultimately, a sub ends up out of sight, so I don't know how much benefit I'd get from veneer. I like the look of Mikers' Cerberus build, with the base he used to get the stand-off distance he needed, so I'm thinking of adding a base to mine for cosmetics.

Mikers -- I didn't leave any "overhang" on the panels, but ended up with perhaps a 1/32nd inch oversize in places anyway (which I did trim with a flush-bit). Trimming with a flush-trim bit is probably a better option to use with the standard b**t joint, but it does complicate things a little if the rabbeted joint is used. I very carefully set the saw cutting height at 3/8" and the fence at 3/8" which allowed me to make two identical cuts in each edge without resetting the saw. I'm sure many will prefer to cut the rabbets with a router -- that should produce a more precise fit providing one has the necessary guides or a router table. I've got a shaper I might use on a future build, but I'll probably have to buy a cutter if I do. I can use router bits with the shaper, but the spindle speed is quite a bit slower.

[Mikers]

If you look at my ONS build I left 1/8", it requires planning. I cut my rabbits with a dado blade. If you have a plywood strip on your saw fence you can over size the dado setting then adjust the fence for you dado width. I like to design for easy (sloppy?) set ups that can be corrected at the end. Well sort of. In this case the inside dimensions have to be dead on, just the overall dado can vary, that's why you trim it.

[Soundslike]

This next photo shows the assembled box. I strongly recommend dry fitting the panels together beforehand, and marking them so you can more easily see which joints go together. Then lay the panels in logical order around the piece you'll start with. Also, think about which surfaces will need glue and mark them so you'll know where to spread the glue. There are three surfaces to apply glue to, on each edge -- note that the side panels are different because they fit inside the other panels..

The baffle covers the entire width and height, as does the back panel, an arrangement that puts the part lines on the top and sides.

http://i1100.photobucket.com/albums/g414/PCIaint/Cerebus%20Subwoofer/P1020224.jpg

Referring back to the drawing in the first post, you'll note that I used the drafting program to determine the port segment lengths, and locations. Working inside the confines of this small box, with such a large and long port, made it necessary to utilize at least two elbows. Although I can't recall the source, I've read that a sufficiently accurate method of determining the distance through an elbow is to simply measure from where the two imaginary centerlines intersect within it. But, being a persistent skeptic, I decided to do a little cipher'in.

http://i1100.photobucket.com/albums/g414/PCIaint/Cerebus%20Subwoofer/P1020225.jpg

A close examination showed that a PVC elbow isn't necessarily perfectly formed, but I came up with the measurements you see on the masking tape. From the tattered appearance you'll probably surmise that I handled it quite a bit flipping it around, measuring and remeasuring. The total straight line distance is 2-9/16" + 2-9/16", or 5-1/8". I also measured the length of a 45 degree path, which you see marked on the tape. This bit of simple measuring convinced me that the straight line method is not accurate. Perhaps it's accurate enough, but knowing that an inaccuracy is introduced with every elbow, was enough to convince me that I should limit the layout to two elbows. I originally considered a different route that would have placed the interior port terminus further from the driver, but rejected it because a third elbow would have been required.

Rounding over the port terminus on the baffle, adds another variable. For the first horizontal segment, I started my measurement in the middle of the baffle, i.e., 3/8" in.

For those who might be interested, I've since calculated the actual distance, using a different method. According to my calculations, the actual length of the path through a 2" elbow is closer to 4.57", as opposed to 5.125". Either calculation, by the way, depends on accurately cut pipe lengths and full insertion into the elbow.

Probably what really matters regarding port length calculations, at least at a theoretical level, is the volume that results from whatever dimensions are used. Even though there is a potentially significant difference between the two methods of calculating elbow length, whatever affect that has on volume could be altered one way or another if the interior dimensions vary through the arc. I'm sure this question has been adequately covered a number of times elsewhere -- if anyone has a link, that would be welcomed...

This last photo shows the port after assembly. I rounded over the edges on the elbows and inside the opening, because it seemed like the thing to do.;) I'm holding the port where it will be inserted through the baffle. I'll be sure to report the acoustical benefits of the paint job.:rolleyes:

http://i1100.photobucket.com/albums/g414/PCIaint/Cerebus%20Subwoofer/P1020230.jpg

[50 watt head]

The rabbets look great, Ron.

[djg]

I made some ONSs and a Cerberus sub for my girlfriend's nephew. He keeps asking for more bass. Your build is very nice.

[Randall Kepley]

Ron, what did you spray that PVC w/? It looks smooth as butter :cool: R.K.

[Randall Kepley]

Ron, I also meant to ask you, how are you making your Rabbets? Are you cutting them w/ a dado blade or Router bit. I can usually tell by the cut , but at the angle of the pix I cant quite make it out. Which ever your using it looks like your dead on. BTW what size of a room are you planning on using this sub in. Thanks bud, R.K.

[Soundslike]

Randall,
I just used a rattle can on the port tubing. It's Rustoleum's "Professional" line, and it actually does a good job on small pieces. I was painting the inside of the port so it would be black when viewed from the outside of the finished enclosure when I thought I might as well paint the outside too. A complete waste of time I'm sure, but I didn't like the looks of the long port with PVC elbows. Even though it would be inside and out of sight it looks like something from underneath your floor, or inside a wall.

I'm using a table saw to make the rabbets. I've got a vertical fence jig I made for making the cuts with the panel in the vertical position. I didn't bother switching over to the dado blade, so two cuts are necessary. I manage to get sufficient accuracy for a project like this. I know you don't need a photo to understand what I'm talking about here, but for the benefit of anyone I may have confused, I'll insert a one here taken by Mark65 during another project we've got going.

http://i1100.photobucket.com/albums/g414/PCIaint/Duellati/DSCF1049.jpg

Regarding the room where I'll use the Cerberus, I have a smallish office about 13 X 13 where I'll use it with whatever speakers I'm listening to, for music. Currently, I'm using it with a pair of Paul's Overnight Sensations and it does a great job. It doesn't produce as much punch as the 12" Dayton subwoofer I made, but it makes more than enough for music played at reasonable levels, and it's about 1/4 the size of the 12" Dayton.

Ron

[Soundslike]

I forgot to acknowledge the comments by 50 watt head, and djd. Thanks for the nice words and encouragement.


I installed the vent inside the box, and it fit as planned in the drafting process, so that it's supported by the bottom and sides. I insulated the box with 1/2" automotive felt carpet padding on both sides and the bottom. There was very little area on the back that could be covered, once the plate amp was installed, so I didn't bother there. I was still concerned that I was using up too much of the little box's volume with the large port and all of the insulation I had added, and to do any good on the back, I would have needed to insulate the cover on the plate amp. So, I opted to leave well enough alone.

In the photo you can see that the driver was moved to the left, and up in the box, in order to make room for the vent. The vent was positioned to take into account the dimensions of the elbows, and a planned roundover of the terminus.

What looks like a ring around the PVC, is actually a repaired goof. I made the hole too large for the PVC, and rather than throw the baffle away, I elected to make a disc from MDF, sized to fill the hole. I was careful to place a pivot hole in the center, before cutting it out. After the glue dried, I went back with my router and circle guide and cut the correctly sized hole.

http://i1100.photobucket.com/albums/g414/PCIaint/Cerebus%20Subwoofer/P1020231.jpg

Here's a view through the cutout for the plate amp. As you can see, I've flush mounted the driver and the plate amp. Pay no attention to the little notch, that will soon be gone.:o

I used a bushing for routing the recess for the plate amp. The easiest method I've found to make this cut is to clamp straight edges for the bushing to follow. It's also necessary to clamp on stops to avoid going to far, and it's important to use a bit with a radius that matches the rounded corners of the plate amp face. To get a bushing with a hole large enough to not interfere with the bit I used, yet shallow enough to for the overall length of the bit, I ended up modifying one. I'll include a longer explanation and a photo, or drawing later, for those who aren't familiar with the process.

http://i1100.photobucket.com/albums/g414/PCIaint/Cerebus%20Subwoofer/P1020234.jpg

Here's a view with the plate amp in place.

http://i1100.photobucket.com/albums/g414/PCIaint/Cerebus%20Subwoofer/P1020235.jpg

[Mikers]

nice work on the flush mount.

button this thing up so we can see some finished photo's.

[Soundslike]

After devoting most of my time to several other build strings, I'm back to the Cerberus. Picking up where I left off - the front two side corners were rounded off, using a 1/2" bit. All other edges were left square, but rounded off softly with light sanding. Earlier, I purchased a 1" roundover cutter for my shaper, planning to use it on this build. But the decision to go with the larger port, made it necessary to move things around, and the necessary real estate was lost. I perhaps could have used a larger radius, but that would mean getting fairly close to the port flare, and I was concerned that it would look odd.

This next photo, is actually another Cerberus I'm building, but it illustrates how things looked after a 70/30 mixture of fiberglass resin and acetone was brushed on (70% resin). The 70% mix tends to leave brush marks that must be sanded out before paint can be applied, and the resin does take several days to cure thoroughly, before it can be sanded without excessive clogging of the sandpaper. The enclosure shown in the photo was coated with a 50/50 mix, which flows out much better, leaving minimal brush strokes. The enclosure in the photo uses the standard 1-1/2" port and the driver is centered.

I like to use resin over the MDF, because it transforms the soft MDF surface into a more durable surface, and it seals very well. The result is a very smooth (after sanding) and relatively tough surface that resists dings much better. It is a messy process, and an extended drying period is necessary.

http://i1100.photobucket.com/albums/g414/PCIaint/Cerebus%20Subwoofer/P1020279.jpg

In this next photo, we're back to the actual enclosure that is the subject of this build. Here, the enclosure has been sanded smooth -- the resin coat was sanded to 150 grit, fiberglass body filler was used to fill all the little inevitable defects, more sanding, then black lacquer based high-build primer was sprayed on, followed by sanding to 220, and finally 400 grit. The reddish stuff is glazing putty used to fill minor defects. You can see the the port required a little smoothing -- that's the result of two little errors. In an earlier post you can see where I had to add a ring around the port tube because I had mistakenly bored the hole too large. Then, when I glued the tube into the resized hole, I didn't insert it fully to the front edge of the baffle, because I had calculated the port length factoring in the port flare. Not too sure how much it matters, but I believe the port length should be measured from the middle of the flare, so I glued it in set back about 1/4". Unfortunately, the roundover didn't reach the tube, so I was left with a little ugly gap to fill.

To make smoothing the port on the second enclosure easier, I used a different installation process (the one shown in the photo above of the resin coated enclosure with the standard port). On the second enclosure, I first located the port and drilled a 1/8" hole through the baffle so it would be possible to determine where to drill/bore/rout from either side. I then bored the hole with a Forstner bit from the back side, to within 1/4" of the front surface. I epoxied the PVC pipe into the hole, which left it positioned 1/4" behind the surface in order to leave a bit of MDF for the roundover. I added a collar made from 3/4" MDF around the PVC on the backside in order to increase the glue surface area. After everything dried/cured I enlarged the center hole to 1/2", and then used a flush-trim bit to create an extended hole through the MDf that matched the inside of the port. That hole was then rounded over.

http://i1100.photobucket.com/albums/g414/PCIaint/Cerebus%20Subwoofer/P1020263.jpg

http://i1100.photobucket.com/albums/g414/PCIaint/Cerebus%20Subwoofer/P1020277.jpg

[fastbike1]

And folks actually like mdf better than plywood. Never ceases to amaze me. ;)

[Soundslike]

And folks actually like mdf better than plywood. Never ceases to amaze me. ;)

Yeah, there are lots of things not to like about the stuff. I should give Baltic Birch a go, I guess.

[Soundslike]

After sanding the primer to 400 grit, I sprayed on a single coat of three-part urethane automotive paint. I managed to produce a nice finish, but there are a few areas that could have been done better. Black is difficult to spray outdoors because it's tough to get the light just where you want it so you can view the build up from an angle to get a smooth, even application. For reasons I can't recall at the moment, I elected to finish all six surfaces which, of course, includes the bottom. Not wanting to hang the box for spraying, I used a method recommended in Jewitt's book on spray finishing -- I drilled four holes in a scrap piece of MDF and inserted four 8 penny finish nails to make a support. I sprayed the bottom first, and then set it on the nails while I sprayed the rest of the box. The method works fairly well, but of course you end up with four tiny marks in the paint -- not too big of an issue if it's on the bottom of the box. I also painted the Duellati speakers described in another build string at the same time.

http://i1100.photobucket.com/albums/g414/PCIaint/Cerebus%20Subwoofer/P1020291.jpg

And here's a photo of the finished subwoofer.

http://i1100.photobucket.com/albums/g414/PCIaint/Cerebus%20Subwoofer/P1020331.jpg

[johnlinvell]

Thats beautiful............I will be building one of these in the am but my finish will not be as good as that...Great Job on this one..............:cool:

[Soundslike]

If you haven't already done so, check post #85 on this (http://techtalk.parts-express.com/showthread.php?t=222437&page=5) build string. Mikers got an outstanding finish using spray cans and color sanding techniques. On the other hand, subs are often out of sight anyway.

[50 watt head]

More outstanding work, Ron! That finish looks terrific :D