After waffling around about which of the million different types of subwoofers to build, I finally decided to design one and actually build it!
The product manager side of me realized that I had some of the usual conflicting challenges as well as a juxtaposition of several different design goals. 🧐
The short (for me!) version:
The product manager side of me realized that I had some of the usual conflicting challenges as well as a juxtaposition of several different design goals. 🧐
The short (for me!) version:
1. I wanted a highly musical sub that could still pressurize the room at significant (but not insane) db down into the teens. I like being able to play the full spectrum of sound including room decay, 16 Hz organ, synthetic sounds, etc.
2. My wife preferred not to have anything there at all, or if something had to be there, then it should be white and innocuous.
3. I wanted to explore evening out the room loading by getting the drivers far enough apart to act as separate sources, while only occupying one "footprint" in the living space.
4. I wanted the sub to look visually clean from the design standpoint, and robust enough to handle having kids around.
5. I wanted to see if the sub could have some value add beyond reproducing low frequencies when music is playing.
The resulting subwoofer:
- Stands 73" (1.85m) tall and is based off of 14" (35.6 cm) inner diameter cardboard concrete form (like sonotube)
- Has two 12" GR Research 4 ohm drivers with servo coils with baffles 60" (1.52m) apart, powered by a Rythmik HX800 servo amp with 2x400W Hypex outputs
- Leverages a 32' (10m) cobb style light strip with a dimmer and 2700K light color
Here's the day mode:
And here's the evening mode:
- Stands 73" (1.85m) tall and is based off of 14" (35.6 cm) inner diameter cardboard concrete form (like sonotube)
- Has two 12" GR Research 4 ohm drivers with servo coils with baffles 60" (1.52m) apart, powered by a Rythmik HX800 servo amp with 2x400W Hypex outputs
- Leverages a 32' (10m) cobb style light strip with a dimmer and 2700K light color
Here's the day mode:
And here's the evening mode:
It was surprisingly hard to find the 14" concrete form, so I was quite triumphant when I finally head home with a 6' piece of it.
Peeling the inside wax layer was super easy to do...
But then I realized that I'd purchased a form with a water resistant outer layer that didn't want to come off... and I probably couldn't trust my outer layer to properly bond to it. This wasn't going to be fun.
Anyway, I worked on a piece of scrap to determine the actual diameter for the internal parts and was super relieved to discover that the same diameter would work at all positions (some are slightly tapered for shipping purposes). Whew!
And got to work cutting the round parts, 6 of them in total. I got one of them right to use as a pattern, sawed the rest to close to dimension, then used the first to duplicate using the router. Quite accurate, but still takes a lot longer than you imagine!
Peeling the inside wax layer was super easy to do...
But then I realized that I'd purchased a form with a water resistant outer layer that didn't want to come off... and I probably couldn't trust my outer layer to properly bond to it. This wasn't going to be fun.
Anyway, I worked on a piece of scrap to determine the actual diameter for the internal parts and was super relieved to discover that the same diameter would work at all positions (some are slightly tapered for shipping purposes). Whew!
And got to work cutting the round parts, 6 of them in total. I got one of them right to use as a pattern, sawed the rest to close to dimension, then used the first to duplicate using the router. Quite accurate, but still takes a lot longer than you imagine!
After this I did a highly precise simulation to determine the best way to fit the amp into the central chamber. This helped find an angle that worked for insertion and mounting, while providing for access to the amp controls later. It was also used to determine how wide the access slot will be--about 6" wide in order to get that really deep transformer through--and guesstimate where the power outlet was going to sit.
I also wanted to make sure that the sub chambers at the ends were tightly enough connected for the force cancellation to work well enough, so I added in some vertical risers into the equation. I ended up making the central panel 16" high... a little more than needed, but since I already had sufficient volume in the sub chambers I was able to give myself a little breathing room in this central amp chamber. Sometimes I try to overoptimize and then make life much harder on myself than necessary!
Here's the collection of parts, followed by the dry fitting and gluing up:
Fortunately it still fit down the tube!
I also wanted to make sure that the sub chambers at the ends were tightly enough connected for the force cancellation to work well enough, so I added in some vertical risers into the equation. I ended up making the central panel 16" high... a little more than needed, but since I already had sufficient volume in the sub chambers I was able to give myself a little breathing room in this central amp chamber. Sometimes I try to overoptimize and then make life much harder on myself than necessary!
Here's the collection of parts, followed by the dry fitting and gluing up:
Fortunately it still fit down the tube!
Then I got serious about making the baffles. I ended deciding to double up on the material to give more sidewall purchase on the cardboard--having never worked with it, I wasn't sure how well a single layer would bond. This is 18 mm plywood, so about 1.4" thick together.
Remember the waterproof layer on the outside of the tube? I decided to sand it off, which was one of the less pleasant half hours of my life. I should have tested it first--maybe it would've held fine?--but I didn't want to do this twice. If I did this again, I'd definitely try it out first!
Speaking of testing, I used a chunk of laminate to see how it would go. 2 layers of contact cement on the cardboard since it absorbed the first layer fairly well, and a single layer on the laminate. Seemed to bend and hold just fine. Note that this is vertical grade laminate, which is a bit thinner and bends a touch more easily.
Trying to accurately cut the tube to length was a bit more finicky than expected, particularly marking it, since neither end was plumb. I ended up putting the tube against a wall and rotated it with a Sharpie held steady.
Remember the waterproof layer on the outside of the tube? I decided to sand it off, which was one of the less pleasant half hours of my life. I should have tested it first--maybe it would've held fine?--but I didn't want to do this twice. If I did this again, I'd definitely try it out first!
Speaking of testing, I used a chunk of laminate to see how it would go. 2 layers of contact cement on the cardboard since it absorbed the first layer fairly well, and a single layer on the laminate. Seemed to bend and hold just fine. Note that this is vertical grade laminate, which is a bit thinner and bends a touch more easily.
Trying to accurately cut the tube to length was a bit more finicky than expected, particularly marking it, since neither end was plumb. I ended up putting the tube against a wall and rotated it with a Sharpie held steady.
Moving along, I purchased some lumber from a fine hardwoods store and rough cut parts for the spine/stand to size. The upper piece shows the deep channel I then cut out to provide a pathway for the power and signal cables, which I didn't want to see. Always a reminder that these kinds of projects just take time, and to accept that it's part of the experience!
Marking out and cutting the amp insertion/access opening on back of the tube. I tried using Forstner bits for the corners, then jigsawed the sides... sorta worked, but maybe all jigsaw would've been better (although using it on the cylindrical surface was "interesting").
I'm a huge proponent of mocking things up and trying them out first before committing, which turned out to once again be a game-saving idea! I dry fit the central chamber and simulated installing the amp and discovered that getting it into place and screwing it tight was a total bugger. However, some additional notches to allow screwdriver access as well as 2 strips of wood to hold the amp in exactly the right location were pretty easy to remedy and much appreciated later.
Then it was "go" time! While I'm happy with Titebond 3 for the wood parts, I decided to use epoxy with some filler to better fill gaps and get these parts bonded in a bombproof fashion. I slid the parts back and forth a little, then refilled and filleted the epoxy on both sides of everything I had access to.
Since the middle chamber ended up very square, I was able to take measurements around the baffle perimeters to get them reasonably square as well. Then I punched a bunch of 15 gauge stainless nails in to the perimeter of each ring to keep things steady as well as provide more mechanical connection. Likely overkill, but they also kept the parts steady when filleting the epoxy, so double utility!
Once the parts had set up I felt very, very confident that things were solid!
Marking out and cutting the amp insertion/access opening on back of the tube. I tried using Forstner bits for the corners, then jigsawed the sides... sorta worked, but maybe all jigsaw would've been better (although using it on the cylindrical surface was "interesting").
I'm a huge proponent of mocking things up and trying them out first before committing, which turned out to once again be a game-saving idea! I dry fit the central chamber and simulated installing the amp and discovered that getting it into place and screwing it tight was a total bugger. However, some additional notches to allow screwdriver access as well as 2 strips of wood to hold the amp in exactly the right location were pretty easy to remedy and much appreciated later.
Then it was "go" time! While I'm happy with Titebond 3 for the wood parts, I decided to use epoxy with some filler to better fill gaps and get these parts bonded in a bombproof fashion. I slid the parts back and forth a little, then refilled and filleted the epoxy on both sides of everything I had access to.
Since the middle chamber ended up very square, I was able to take measurements around the baffle perimeters to get them reasonably square as well. Then I punched a bunch of 15 gauge stainless nails in to the perimeter of each ring to keep things steady as well as provide more mechanical connection. Likely overkill, but they also kept the parts steady when filleting the epoxy, so double utility!
Once the parts had set up I felt very, very confident that things were solid!
By the way, let's have NO comments from the peanut gallery on the sheer amount of crap in all directions in my garage! 😉
Outstanding, my kind of DIY sub! I 'hate' that this type has gone out of style as they're so flexible WRT blending into a room or just laid down behind a sofa, etc., though prefer up in a corner loaded at each end to make a BP6 AKA tapped pipe. 
:-D
We used to have a wooden garage door like that.... no matter how much paint went on it, everytime it rained it got soaked and it got very heavy. After several near explosive spring failures we got an insulated metal roll up door.
I was thinking about showing the pictures of your garage to my wife to shut her up. She thinks the garage ought to be tidy and organized... we're running out of room with all of those shelves and tool chests.
Anyhow, did you think of making it passive and taking the plate amp either outside or remote? I'm thinking of making mine with a remote amp since I already have two ICE amps that will fit the bill nicely.
Last edited:
Sometimes projects have those moments where you're really committing to the next step, so I was a bit nervous prepping to glue the laminate around the tube.
Then I managed to blow it before I even could try--while trying to gently precondition the laminate to bend the right direction (since it, of course, arrives rolled the other direction) I managed to crack it! With no way to get around the crack. So, I ordered some more. Interestingly enough, the Wilsonart seemed to be a touch thinner and bend a little more easily? Anyway, for reference I used the "4 ft. x 8 ft. Laminate Sheet in Designer White with Matte Finish" and I'm quite pleased with it.
Back to business... trying to get it on smoothly without capturing air bubbles etc. 2 layers of contact cement rolled on the tube, one layer on the laminate, and a patient wife holding it while I rolled a few degrees down the length, turned a bit, rolled again, etc.
It was hard to get the lead edge straight and centered. Be SUPER careful with contact cement parts--once they touch they'll grab with insane tenacity!! I almost blew it with this step. I later realized I could have put wax paper along my lead line, centered the laminate, and Then slowly let it touch down the length.
Also, don't forget to protect your lungs. I used a 3M mask with appropriate cartridge--even outside it was super intense without it.
Fortunately it worked out reasonably well! Yes, I later bought a good laminate roller but the roll of duct tape did ok.
Then I managed to blow it before I even could try--while trying to gently precondition the laminate to bend the right direction (since it, of course, arrives rolled the other direction) I managed to crack it! With no way to get around the crack. So, I ordered some more. Interestingly enough, the Wilsonart seemed to be a touch thinner and bend a little more easily? Anyway, for reference I used the "4 ft. x 8 ft. Laminate Sheet in Designer White with Matte Finish" and I'm quite pleased with it.
Back to business... trying to get it on smoothly without capturing air bubbles etc. 2 layers of contact cement rolled on the tube, one layer on the laminate, and a patient wife holding it while I rolled a few degrees down the length, turned a bit, rolled again, etc.
It was hard to get the lead edge straight and centered. Be SUPER careful with contact cement parts--once they touch they'll grab with insane tenacity!! I almost blew it with this step. I later realized I could have put wax paper along my lead line, centered the laminate, and Then slowly let it touch down the length.
Also, don't forget to protect your lungs. I used a 3M mask with appropriate cartridge--even outside it was super intense without it.
Fortunately it worked out reasonably well! Yes, I later bought a good laminate roller but the roll of duct tape did ok.
Trimming the brittle laminate around the access zone was harder than expected. I finally landed on using a multitool to rough cut it, followed by a trim router. It was again weird using it on the curved surface, but I slowly got to a decent result.
Why is there a gap in the laminate? Well, I knew it would be almost impossible for me to get the edges perfectly mated together, so I deliberately made sure that the ends wouldn't overlap and designed the spine/stand to cover up the misfit. Thinking through the full process earlier helped me choose this approach--what works in sketches or in CAD needs to also be implementable!
Anyway, the next steps were drilling the holes for the 4 mounting bolts, marking the spine, then bonding the threaded rod into place.
Dry-fitting the spine:
Why is there a gap in the laminate? Well, I knew it would be almost impossible for me to get the edges perfectly mated together, so I deliberately made sure that the ends wouldn't overlap and designed the spine/stand to cover up the misfit. Thinking through the full process earlier helped me choose this approach--what works in sketches or in CAD needs to also be implementable!
Anyway, the next steps were drilling the holes for the 4 mounting bolts, marking the spine, then bonding the threaded rod into place.
Dry-fitting the spine:
Somewhere along the line I got to thinking about the end chambers and someone having said this kind of sub could sound dull. For some reason I thought it would be worth my time to put in a layer of 6 oz fiberglass into each of the end chambers. I actually did this before the laminate.
Not only was that probably totally unnecessary, but I now know that I hope to never have to fiberglass the inside of a cylinder again for the rest of my life! 😡 That was absolutely agonizing. However, the end chambers could now survive most things short of a nuclear bomb blast.
Then I slowly, carefully trimmed the tube to its final length. I left the rim to hide both the driver and the light strip I'd decided to integrate earlier in the design process... 1.5" at the bottom, around 2" at the top.
Note to Always give yourself a little extra for these kinds of things! It was essential for this project.
Not only was that probably totally unnecessary, but I now know that I hope to never have to fiberglass the inside of a cylinder again for the rest of my life! 😡 That was absolutely agonizing. However, the end chambers could now survive most things short of a nuclear bomb blast.
Then I slowly, carefully trimmed the tube to its final length. I left the rim to hide both the driver and the light strip I'd decided to integrate earlier in the design process... 1.5" at the bottom, around 2" at the top.
Note to Always give yourself a little extra for these kinds of things! It was essential for this project.
Ok, time to work on the stand parts. I'd very scientifically studied different angles on the shop floor and eyeballed what looked right.
Made through-holes for the power and signal cables in the spine and the tube, then enlarged them twice once I tried to actually feed things through.
Ready to glue, then glued it up. I don't think I roughed up the surface well enough, and didn't get enough clamping pressure on it. It was hard with the weird angles and non-specialized gear that I have for that kind of thing.
Shaping the parts with a mixture of router, sander, and hand file.
Then we went outside and put on some Minwax. The first coat was semigloss and made it look horribly cheap. The satin matte I put on after that was much better.
You know when you anticipate a problem, ignore it, and it bites you immediately?
The wind had picked up a bit and I thought, gee, that thing might get knocked over!
Yep. Came back a bit later and it had hit the ground hard enough to crack the upper channel and mess it up a bit. Lost a day gluing it back into place and trying to blend the errors.
Made through-holes for the power and signal cables in the spine and the tube, then enlarged them twice once I tried to actually feed things through.
Ready to glue, then glued it up. I don't think I roughed up the surface well enough, and didn't get enough clamping pressure on it. It was hard with the weird angles and non-specialized gear that I have for that kind of thing.
Shaping the parts with a mixture of router, sander, and hand file.
Then we went outside and put on some Minwax. The first coat was semigloss and made it look horribly cheap. The satin matte I put on after that was much better.
You know when you anticipate a problem, ignore it, and it bites you immediately?
The wind had picked up a bit and I thought, gee, that thing might get knocked over!
Yep. Came back a bit later and it had hit the ground hard enough to crack the upper channel and mess it up a bit. Lost a day gluing it back into place and trying to blend the errors.
@dayneger -- nice concept & execution! End result looks elegant even though it's so tall. Following your steps.
How well did the force cancellation work? I've used it in OB systems ("W"-config dual-woofer), but not in a sealed enclosure.
I am considering one now as a stand for the TV -- around 5'l x 17"h x 14'd. Dayton RSS265HF-8 10" woofers side-facing on either end, connected to each other with a steel or aluminum pipe or rigid wood (maybe made from BB plywood). Potential enclosure volume is up to 5 cubic feet, so way more than enough for 0.7 system Q, probably even 0.5Q. PEQ in a DSP crossover will take care of the FR details.
If force cancellation works well, there shouldn't be any problem having the TV & audio electronics atop, maybe with a resilient pad on the top panel. 🤞🤞
How well did the force cancellation work? I've used it in OB systems ("W"-config dual-woofer), but not in a sealed enclosure.
I am considering one now as a stand for the TV -- around 5'l x 17"h x 14'd. Dayton RSS265HF-8 10" woofers side-facing on either end, connected to each other with a steel or aluminum pipe or rigid wood (maybe made from BB plywood). Potential enclosure volume is up to 5 cubic feet, so way more than enough for 0.7 system Q, probably even 0.5Q. PEQ in a DSP crossover will take care of the FR details.
If force cancellation works well, there shouldn't be any problem having the TV & audio electronics atop, maybe with a resilient pad on the top panel. 🤞🤞
Finally ready for assembly! I'd painted the ends white, along with what I could reach of the inner chamber. Tucked some antires stuff along the sides of the inner chamber just in case it might buzz.
Then got the amp into place! I was super grateful that I'd done the test earlier since it went just fine.
To my pleasant surprise, I was just barely able to get the leads on the amp without extending them. @tonyEE, the leads on the servo amps are supposed to remain short in order for it to work well. You can actually make them a couple feet longer if needed, but too far and the time delay will mess up the servo system's response.
First assembly and test using the high-tech jig I'd organized:
And it made noise! That was encouraging!
Then I routed the light strip. In my original version it went around the bottom driver twice, up the channel in the back spine, and several laps around the top driver. Hard to see here, but there's a strip of wood that covers up the cables in the channel and provides the mounting surface for the light strip.
Here's the back side:
As it turned out, my original vision of this glorious light shining up and down didn't play out quite as anticipated. The uplight looked wonderful, and the strip along the spine as well, but the downlight made a very distracting hotspot on the floor that totally drew the eye to the cables and stand. I later removed that section and it's much better!
Then got the amp into place! I was super grateful that I'd done the test earlier since it went just fine.
To my pleasant surprise, I was just barely able to get the leads on the amp without extending them. @tonyEE, the leads on the servo amps are supposed to remain short in order for it to work well. You can actually make them a couple feet longer if needed, but too far and the time delay will mess up the servo system's response.
First assembly and test using the high-tech jig I'd organized:
And it made noise! That was encouraging!
Then I routed the light strip. In my original version it went around the bottom driver twice, up the channel in the back spine, and several laps around the top driver. Hard to see here, but there's a strip of wood that covers up the cables in the channel and provides the mounting surface for the light strip.
Here's the back side:
As it turned out, my original vision of this glorious light shining up and down didn't play out quite as anticipated. The uplight looked wonderful, and the strip along the spine as well, but the downlight made a very distracting hotspot on the floor that totally drew the eye to the cables and stand. I later removed that section and it's much better!
My son helped me get the phase dialed in, we got it crossed over pretty well (don't yet have measurement gear, will do so in a bit) by ear, and we were really happy that it was done in time for his birthday dance party.
Five days later I got a call from my wife saying that she'd heard a really loud "boom" from upstairs. Oh crud.
Somehow it had stood there for several days, but that lower joint was overloaded and not executed well enough!
Totally cracked the laminate as well.
Fortunately, the really expensive $$ stuff survived the impact. And while the surface was messed up, the rest of the sub seemed ok. Saved by that fiberglass?
Five days later I got a call from my wife saying that she'd heard a really loud "boom" from upstairs. Oh crud.
Somehow it had stood there for several days, but that lower joint was overloaded and not executed well enough!
Totally cracked the laminate as well.Fortunately, the really expensive $$ stuff survived the impact. And while the surface was messed up, the rest of the sub seemed ok. Saved by that fiberglass?
So naturally, this occurred just a few days before said birthday dance party, so I immediately set to work redoing a bunch of stuff I'd just finished doing.
In a serious heat wave with no AC. Working with contact cement at over 90 degrees F isn't amusing!
I ground out the cracked zones and filled with a smear of bondo (that hardened in about 45 seconds at that heat).
I decided to rebuild the existing joint, using screw pressure and leaving metal in place, then support the tube where it overlapped the legs to take up the moment (torque) on that joint.
Courtesy of this semi-disaster I also got to try out my improved method of getting the laminate started. Lo and behold, it worked great! Much much easier to control the initial cement contact.
Trimmed everything again, painted exposed bits, reassembled, redid the light strip, and poof! We were back in business.
In a serious heat wave with no AC. Working with contact cement at over 90 degrees F isn't amusing!
I ground out the cracked zones and filled with a smear of bondo (that hardened in about 45 seconds at that heat).
I decided to rebuild the existing joint, using screw pressure and leaving metal in place, then support the tube where it overlapped the legs to take up the moment (torque) on that joint.
Courtesy of this semi-disaster I also got to try out my improved method of getting the laminate started. Lo and behold, it worked great! Much much easier to control the initial cement contact.
Trimmed everything again, painted exposed bits, reassembled, redid the light strip, and poof! We were back in business.
So far I've been really enjoying the clear extension of this subwoofer. It easily plays below 20 Hz, seems to be surprisingly smooth in the room response, feels tight and snappy with music, give me confidence that if it's in the music I'll hear it... fun stuff!
I've also really enjoyed the visual effect in the evening, where the tube seems to kind of float in space. My wife finds the aesthetic quite acceptable and appreciates the dimmable amount of indirect light it provides, and it's actually really nice being able to just use it as a lamp regardless of whether there's music playing!
Here are the initial photos again, for reference:
I've also really enjoyed the visual effect in the evening, where the tube seems to kind of float in space. My wife finds the aesthetic quite acceptable and appreciates the dimmable amount of indirect light it provides, and it's actually really nice being able to just use it as a lamp regardless of whether there's music playing!
Here are the initial photos again, for reference: