Starting a curved baffle dipole woofer array project

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I'm (finally) getting my hands dirty with a new dipole woofer project that will ultimately mate with a pair of DIY electrostatic panels. I thought it might make more sense to post the woofer part here rather than in the ESL forum.

I wanted to try my hand at making curved baffles for several reasons:
1) It dramatically stiffens the baffle in one direction
2) It gives me a chance to try making kerfed boards for the first time
3) It gives me a chance to try working with fiberglass and epoxy
4) I'd like to break free from the tyranny of rectilinear speakers--in other words I thought the result might look cool.

I'm going to build two vertical arrays, each holding six Dayton Reference Series 8" woofers from Parts Express (RS225S-8).

Each array will end up between 5.5 and 6 feet tall (a bit under two meters tall, for those lucky enough to living metrically). I made the baffles oversized in both height and width so that I could have room to test techniques and adjust dimensions as I go along. So far I've only cut one woofer hole to see how my ideas will work in the real world.

The kerfed board made of a 7 foot long by 16" wide piece of 3/4" MDF:
An externally hosted image should be here but it was not working when we last tested it.


The kerfed board temporarily wrapped into its curved shape:
An externally hosted image should be here but it was not working when we last tested it.


The router support used to make flat cuts in a curved surface:
An externally hosted image should be here but it was not working when we last tested it.


The cutout designed to accept a woofer in its MDF mounting ring:
An externally hosted image should be here but it was not working when we last tested it.


An MDF woofer mounting ring placed into its hole in the baffle:
An externally hosted image should be here but it was not working when we last tested it.


A woofer mounted to its ring and placed in the hole. The woofers will be mounted from the rear (an unfortunate phrase) to hide the mounting holes and provide a cleaner look from the front.
An externally hosted image should be here but it was not working when we last tested it.


Ultimately, I plan to wrap the MDF baffle with fiberglass and epoxy to stiffen it further and smooth the transition from the mounting rings to the curved baffle. Here's a very rough test of what that might look like. That's my unprotected hand handling fiberglass. Nine years of college and still not too bright... Once the fiberglass is applied I'll cut through it to re-expose the woofer holes. I'm also trying to come up with a good way to fill the small gap between the fiberglass and the MDF at the sides of each woofer ring. Epoxy with talc as a filler is one option I'm toying with.
An externally hosted image should be here but it was not working when we last tested it.


That's as far as I've gotten so far. I'm hoping that posting this will force me to maintain construction momentum. I need to find easier ways to do some of the steps I've done so far if I'm going to mount twelve woofers. The Jasper router jig really works well for cutting circles, but you always need a stable center pivot hole about which the router can rotate. That's not always conveniently achieved. Here's a picture of the jig being used to cut the hole in the router support. I managed to make the hole the wrong size and then had to figure out a way to make it larger without having solid material to hold the Jasper jig's pivot pin. At least I made the hole too small instead of too large! The woofer mounting ring, also cut with the Jasper jig, is partially visible at the bottom of the picture.
An externally hosted image should be here but it was not working when we last tested it.


Few
 
One point of clarification: The 1/2" thick MDF woofer mounting rings will not provide a sufficiently stable platform for the woofers, and the baffle will only be 3/4" thick (plus fiberglass). To minimize the excitation of panel resonances and maximize woofer stability I'm planning to include a vertical support of some kind behind the curved baffle and mount the woofers to that support, probably by gripping their magnets. I'll then use a soft gasket to seal the woofers to the MDF mounting rings. One of Linkwitz's Orion modifications included a similar approach. I'd like the vertical support to be stiff along the woofer vibration axis but not so bulky that I lose the attractiveness of the thin front baffle. Has anyone seen an elegant solution to this sort of problem?

Few
 
Disabled Account
Joined 2007
Few said:


Ultimately, I plan to wrap the MDF baffle with fiberglass and epoxy to stiffen it further and smooth the transition from the mounting rings to the curved baffle. Here's a very rough test of what that might look like. That's my unprotected hand handling fiberglass. Nine years of college and still not too bright... Once the fiberglass is applied I'll cut through it to re-expose the woofer holes. I'm also trying to come up with a good way to fill the small gap between the fiberglass and the MDF at the sides of each woofer ring. Epoxy with talc as a filler is one option I'm toying with.


Hi,
Looks like an interesting build. I've done a few curves myself.
You might want to reconsider the fiberglass wrap as it will be very time consuming to make smooth.

For what its worth, I would stack the rings to go completely through the opening, gluing them in. I'd also mount the woofers from the front, as this will be a lot easier and easier to brace the magnet.
Where the rings meet the baffle, you could use auto body filler (basically talc mixed with epoxy, except in a more convenient package) to fill and smooth the transition. Might be worth filling the kerfs with this too to increase the overall strength.

The front of this amp case is kerfed 5/16" HDF (masonite). sides and top are also 5/16" HDF.
 

Attachments

  • im001621.jpg
    im001621.jpg
    97.2 KB · Views: 1,512
Thanks for the response and the suggestions. You certainly might be right about the fiberglass concerns. On the other hand I'm not looking forward to cutting out twelve woofer rings out of MDF so if I had to do 24 or 36 rings in order to stack them, I think I'd really be discouraged.

As a result of the concerns you voiced I've started reconsidering an idea I had awhile ago. I could carefully sculpt the desired shape around the cutouts and MDF ring I've made already, perhaps using auto body filler to refine the shape. I could then use that shape as a plug, make a fiberglass mold from that plug, and then make twelve baffle fronts from the mold. I'd only have to carefully prepare the surface of the plug once, and if all goes well, I'd be able to make twelve copies whose surface finish wouldn't require lots of sanding before painting. I don't have any experience with this approach but it would be a MUCH simpler project than the one that inspired my thinking---a thread describing the diy fabrication of a carbon fiber motorcycle fuel tank, as described
here

I agree that filling the kerfs would be helpful. I've explored running some rubber sheet down each kerf in order to provide some damping, or I could just use filled epoxy in the way you described. I guess I'll have to experiment a bit more before finalizing a plan.

Thanks again for the thought provoking comments. By the way, I was inspired by your amp case photos in the high gloss thread. I hope to try your approach if I ever manage to get these things built.
Few
 
Well I guess slow progress is better than no progress. Work and life have been getting in the way, but I finally have the woofer rings and the holes to accept them cut into one curved baffle. Here are the various stages in the birth of a ring. I devised a jig that would hold the rings at various stages of their construction to simplify the use of my Jasper jig. Here's the jig and a few rings in various stages of their completion.
An externally hosted image should be here but it was not working when we last tested it.

An externally hosted image should be here but it was not working when we last tested it.


I ended up having to fill the kerfs in the curved MDF baffle before I could cut all the woofer holes because the baffle started to flex too much when more than one hole was cut into it. I filled the kerfs with West System epoxy thickened with talc. When I started to realize how much epoxy I was going to use up in this process ($$!), I quickly dug out my supply of popsicle sticks to help fill the kerfs. The thickness of the sticks turned out to be perfect, and I think I cut the volume of required epoxy down by more than 50%. Here's the concave side of the baffle.
An externally hosted image should be here but it was not working when we last tested it.


I refined the platform I made to support the Jasper jig so that I could cut flat holes into the convex side of the curved baffle. You can see the filled kerfs where the material has been removed to accommodate the woofer and woofer ring.
An externally hosted image should be here but it was not working when we last tested it.


With all six woofer rings resting (but not yet epoxied) in their holes, things start to take shape.
An externally hosted image should be here but it was not working when we last tested it.


Here's the side view of a ring sitting in its hole in the baffle. If I go ahead with the fiberglassing as planned (see earlier posts), I'll use the gap between the ring and the baffle to provide access to the space between the fiberglass and baffle. I'll need to fill that gap with sand or expanding foam, or...other suggestions? Now that I have a clearer idea how much epoxy costs and how much you can go through in short order, I know I'll not be filling that space with epoxy.:)
An externally hosted image should be here but it was not working when we last tested it.


I have to say that now I've put together a few photos it doesn't look like as much progress as it felt like when I turned off the light in my shop this afternoon. The good news is that things are moving much faster now that I've devised some jigs to crank things out reproducibly. I'm hoping the other channel will take a small fraction of the amount of time this one did. Before I deal with the second channel, though, I still have to delve into the world of fiberglass and epoxy. I've given a lot of thought to the suggestions that I just forget about the fiberglass, but at the moment I'm still leaning toward using it. It will stiffen the baffle, and part of the reason for taking on this project was to gain a little experience with some new (to me) materials and techniques.

Few
 
I'm going to wait until I have some measurements in hand before making a firm decision, but expect it to be between 200 and 500 Hz. I was originally leaning more toward the high frequency end of that range, but I don't want the woofer-woofer or woofer line-to-ESL distance to cause major lobing issues so I may end up closer to 250 Hz.

Few
 
I asked because I was trying to figure out if you could use one solid piece to mount the woofers on then mount the one piece smaller stiffer baffle to the curved baffle without mucking up the sonic benefit of the curved baffle.

I see you are using Russ's panels - NICE!
 
Thanks for the interest, Magnatar, but I'm afraid you lost me with the "Russ's panels" reference. Which panels do you mean? These are home-brew. I did look into buying some pre-kerfed panels but had trouble locating a local source; companies never got back to me after I emailed asking for information. Thanks for the aitwood.com link, Jorge. I did know about them but the larger sizes looked expensive to me when I was planning things out. Since I already had MDF hanging around I decided just to do it from scratch. Now that I know what fun it is to fill the kerfs with epoxy, and how expensive the epoxy is, a commercially available precurved panel of the right dimensions might be a good option.

Magnatar, when you mention the "smaller stiffer panel" are you referring to the panel grabbing the woofers by their magnets, or the electrostatic panel these will be mated with? When I first read your post I thought you were referring to the ESL panel so I wrote up the following response. As I reread things I realized I may have misunderstood. I've already got this typed up so I'll just leave it in here in case it's of interest to anyone.

Regarding attaching the ESL panel to the woofer panel:
I thought a bit about coupling the ESL panel directly to these curved baffles, but I'm worried that the front/back pressure difference generated by the woofers might force the ESL diaphragm into the stators. Leaving a gap between the ESL and the woofer array provides at least some chance to short circuit that pressure difference. Also, I want to minimize the chance that woofer panel vibrations will modulate the ESL's output. Of course the acoustic coupling risks causing the same problem, but others have gotten this configuration to work well so I'm going to give 'er a try. Here's one of the better known commercial examples:
ML Statement e2

If you were writing about a panel that the woofers' magnets would be mounted to and you have some suggestions about the geometry, can you tell me a bit more about what you have in mind?

Few
 
I thought I read you were using "Just Real Music" electrostatic panels - Russ is the man behind that.

I was referring to using a stiff one piece panel (maybe baltic birch) to mount the woofers to then use the curved panel (attached so it is damped) behind it - The panel would be just big enough to mount the drivers to ------- like the circles you cut out but one piece for all of them
 
Magnatar: I now get the Russ connection. Thanks for the clarification. Rather than buying commercial ESL panels I'm going to build stators with stretched magnet wire supported on garolite frames (garolite is McMaster-Carr's version of a paper-phenolic composite). I want to be able to control the directivity by electrically breaking the panels up into vertical strips---but I guess that's a discussion for the "ESLs, planars, alternative technologies" forum.

I think I also now have a better idea what you were envisioning for woofer support. Here's the sort of configuration I'm working toward at the moment. I haven't bothered trying to get SketchUp to depict the fiberglass stretched over the curved baffle, but hopefully the basic shape is clear. The sketches don't show a finalized way of connecting the woofer magnets to the rear panel or connections between the curved baffle and the rear panel because I haven't worked out those details yet.

Few
 
Hi,

the curved design looks well, but do yourself a favour and add some ´distance´ for the sound waves, i.e make it not just a baffle but a U-Frame. With just 6 of Your drivers You have to fight for every dB of output! Believe me, with a good panel crossed over between 150Hz and 250Hz (not higher!) You´ll find yourself more than often cranking up the volume to take a sound shower :D
Your drivers Qt asks for considerable equing of the freq-response. With just a baffle You´d loose too much -at least for my taste. With a U-Frame, or something similar, you need less equing, but the useable bandwidth is limited to ~250Hz. Since most DIY-Panels have a d/s of >1mm the crossover frequency can be as low as 150Hz. I wouldn´t opt for higher than 250Hz crossover freq at all. Basically everything that´s so charming about ESLs will be lost. If you have a panel that is capable of reproducing down to 150Hz...then use it so!

jauu
Calvin
 
Hi Calvin,
Thanks for the suggestions. I may end up needing to add the "wings" you're referring to, and will try to build things to make that not too hard to do.

For now, though, my plan is to use two 12" woofers in sealed boxes for the bottom 1.5 - 2 octaves. Given the fact that volume displacement goes as 1/(frequency)^2, if I use the array of 8" woofers only down to 60 Hz instead of the woofers' Fs (27 Hz), I reduce the excursion required of them by a factor of 5. That gain in headroom will be much larger than anything I can achieve by increasing the effective baffle size or number of woofers. I will have to sacrifice whatever dipole magic might be achieved below 60 Hz, but I reduce the risk of box-like resonances that a U-frame or its variants would have, and don't have to build woofer arrays that are larger than I want.

The subwoofers will also give me multiple sound sources in the deep bass region which I should then be able to locate to minimize the standing wave effects in the room. My thinking is that having multiple sources that can be placed to provide the most even bass response will be more important than maintaining dipole operation down to the last octave. I may find this thinking is all wrong, but I guess that's the fun of these little exercises!

Few
 
A bit more progress. This sequence shows the application of some knit cotton fabric over the woofer rings and baffle after the woofer rings were epoxied in place. The fabric is similar to what might be used to make a heavy T-shirt. The black color turned out to be an excellent choice as I discovered when I laid fiberglass over the cotton---but I'm getting ahead of myself.

MDF woofer rings epoxied into the holes in the kerfed MDF baffle (click to enlarge):

An externally hosted image should be here but it was not working when we last tested it.



Black cotton knit fabric laid over the front of the baffle and taped to the back of the baffle to keep it stretched tautly over the baffle. This fabric was used to create the nice smooth curves that were part of my motivation for embarking on this experiment (sorry the black fabric makes it hard to see the contours):

An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.



West Systems epoxy was brushed into the fabric to attach it to the baffle and stiffen it. The stiffened fabric defines the shape for the fiberglass layers that will be epoxied on later. I thought about going straight to the fiberglass and skipping the cotton fabric but I'm glad I didn't. The fabric is MUCH easier to manipulate for this sort of application. This heavy cotton absorbs a hell of a lot of epoxy! When it hardens, though, it's makes a surprisingly stiff shell:

An externally hosted image should be here but it was not working when we last tested it.


Next step, laying fiberglass over the cotton and epoxy composite.

Few
 
Disabled Account
Joined 2007
Looking Good! Ah, the smell of curing epoxy...brings back memories of my childhood, as my father built a fiberglass boat in our basement. What's the harm in a little brain damage? :dead:

The last photo could be a scene out of an audio based Dexter episode, what with all of the plastic sheeting, and the victim carefully laid out. :)
 
After the epoxy in the cotton fabric cured I sanded the very rough surface. This was necessary because I waited more than 24 hours between applying the epoxy and getting ready to apply the next epoxy coating. I might have been able to avoid the sanding if I had acted more quickly, but the cotton and epoxy shell came out pretty rough so I would have wanted to sand it anyway. Here's what it looks like after sanding with a random orbit sander. The process went quicker and more painlessly than I thought it might. It was gratifying finally to see the shape I had envisioned actually materializing. I probably could have sculpted the curves around each woofer ring using Bondo or something, but this method creates the curves almost automatically and does so more uniformly than I could have done by hand.

An externally hosted image should be here but it was not working when we last tested it.



I used a rotary cutter (a cross between an X-acto knife and a pizza cutter) to cut fiberglass cloth to size, wet down about a third of the baffle with epoxy and then laid one end of the fiberglass over the epoxy. I added some more epoxy on top of the fiberglass to thoroughly wet it and make it adhere to the baffle. This is where the black cotton fabric paid off. It made it very obvious where I had gotten the fiberglass stuck to the underlying layer because the black color showed through clearly. The not yet stuck areas looked white. I repeated the process for the middle and last third of the baffle by folding the fiberglass out of the way, applying the epoxy, and then folding the fiberglass back over and onto the wet epoxy. I used super cheap "chip brushes" from Home Depot to apply the epoxy and they worked great. In this photo the white fiberglass is clearly visible in the center of the woofer holes because I didn't apply any epoxy there. The black color elsewhere indicates the fiberglass is thoroughly wetted with epoxy and the fiberglass is adhering to the baffle.

An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.



When I first bent the kerfed mdf I was impressed with how much the panel was stiffened by the curvature. When I started punching woofer holes into it it became much flimsier and I thought I might be in trouble. One layer of fiberglass really stiffened things up so I'm back to feeling optimistic. Nonetheless, I decided to play it safe and apply a second layer of epoxy to the front of the baffle. I'm waiting for that to cure as I write this. Tomorrow morning I'll apply more epoxy to the fabric to fill the weave texture and get closer to a smooth paintable surface.

I'll probably apply at least one layer of fiberglass to the back of the baffle under the assumption that the sandwich construction will help reduce panel flexing problems. I'll have to wait and see whether it makes sense to double up the fiberglass on the back as well. The shape of the back of the baffle is simpler so I'm hoping the application of the fiberglass will be straightforward. Then again, sometimes the "easy" steps cause all the trouble...

I'm getting far enough along with this one baffle to have confidence that this crazy scheme might actually work and that it's worth starting to build the second baffle. I hope I've learned enough from the experience with the first baffle to move more confidently and quickly through the process. I guess we'll find out.

Few
 
John: I hadn't caught any of the Dexter series so I had to do a bit of Googling to follow the reference. I have to admit I see your point. Does this make me the serial killer or one of his victims? I kept thinking of Darth Vader every time I saw the black monolith shaping up but now I see things may have even a darker side. The fact that I've been avoiding breathing epoxy fumes by holding my breath for an hour at a time just adds to the effect.

Brandon: Thanks for the encouragement. I was wondering whether I was providing too much construction detail on a project that might not be of very broad interest.

Few
 
Disabled Account
Joined 2007
Few said:
The fact that I've been avoiding breathing epoxy fumes by holding my breath for an hour at a time just adds to the effect.

I was wondering whether I was providing too much construction detail on a project that might not be of very broad interest.

Few


Hi,
Yeah, the fumes, I remember them well. I have to hand it to you, you are doing a first rate job and it looks like it will be well worth the effort.
I did a bit of fiberglass myself a few weeks ago, when I put a hot tub back together. It was cut in half in order to remove it from a house. I "bandaged it together and now it's as good as new. I just need to find the time to paint it (with car paint).

You say you are using epoxy, not the regular polyester resin for fiberglass?

Lots of construction detail is great. Too many threads here are long on talk and short on results. It's nice to see some industry happening. :up:
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.