Carbon fiber + nomex honeycomb sandwich DML panel construction project - diyAudio
Go Back   Home > Forums > Loudspeakers > Planars & Exotics
Home Forums Rules Articles diyAudio Store Gallery Wiki Blogs Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Planars & Exotics ESL's, planars, and alternative technologies

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Thread Tools Search this Thread
Old 12th December 2016, 03:41 AM   #1
diyAudio Member
Join Date: May 2010
Default Carbon fiber + nomex honeycomb sandwich DML panel construction project

I made it through most of the hundreds of pages of threads on distributed mode loudspeakers (DMLs) around here, AudioCircle and Parts Express' forum, I decided I wanted to try building a couple of my own panel speakers. I did it yesterday and they sound great. Pictures to come, but I'll start with background first for those who are interested but would rather not have to wade through the gigantic existing threads.

To recap what I found in previous DIY threads:
  1. Basics of the technology:
    1. From a Parts Express customer and DIYer: DML Flat Panel | Parts Express Project Gallery
    2. From Techtonic Audio Labs (owners of the intellectual property from the now defunct NXT): Professional Distributed Mode Loudspeakers |
  2. Material stiffness and size matter, especially with respect to reproducible frequency range and especially in the bass where bigger and stiffer is better. Weight still needs to be kept down though for efficiency and max SPL to be acceptable.
  3. The best panel materials DIYers have found so far are, in no particular order:
    1. 1/8" or 1/4" birch ply
    2. XPS or EPS insulation, around 1" thick, with the paper coating sanded off and new edge treatment for stiffening such as diluted wood glue or shellac.
    3. Heavy-duty foamboard or signboard, the preferred brand being Gatorboard:
  4. Commercially available high quality DMLs use materials not easily available to DIYers.
    1. For live pro audio, Tectonic uses carbon fiber sandwiched around nomex honeycomb:
    2. Podium Sound makes home hi-fi DML panels using what sounds like a one-off phenolic paper honeycomb: 6moons audio reviews: Podium Sound Model 1
I thought, there has to be a way to get the materials together for something like pro construction, although part of the fun with DML seems to be that you can make a speaker out of literally anything, so why not try and see what happens?

In the next post: the search for commercial composites and honeycombs.

Last edited by altie; 12th December 2016 at 04:09 AM.
  Reply With Quote
Old 12th December 2016, 04:13 AM   #2
diyAudio Member
Join Date: May 2010
Sourcing materials

Carbon fiber and nomex are both individually regarded as unobtanium for most of us DIYers. Personally my threshold for a DIY project is that if I'm already investing my time, even if it's probably going to sound really good, I'd like to keep it under $100 in case the whole project is a wash.

You can buy prefab carbon + nomex honeycomb sandwich in form very close to what's in the Tectonic video I posted earlier (probably thicker, but probably not substantially heavier or less stiff): here it is at Dragon Plate for $40/6x6" square: Carbon Fiber Nomex Honeycomb Core 0.5" x 6" x 6"

In larger sheets for a little under $100/sq. ft.:

Even if you buy separate, nomex honeycomb itself is pretty expensive, like $120 for a 40x48" sheet: Nomex Honeycomb Sandwich Core Material in stock | Fibre Glast. Bigger than I thought I'd need, at least for an initial build, but also over total budget.

Likewise carbon fiber: $27/sq. ft. panel:

So I had to get a little more creative...
  Reply With Quote
Old 12th December 2016, 04:23 AM   #3
diyAudio Member
Join Date: May 2010
I started calling around and leaving messages at aerospace CF manufacturers giving a quick explanation about my project and asking if they'd be willing to sell scraps or seconds. I started late on a weekend so didn't get ahold of anybody initially. Got on the line early on a Monday with someone on the east coast who sympathized but said they're only set up to sell to bigger customers (not surprising).

Eventually I landed on Pro Tech Composites. They're near me in Vancouver, WA, and sell a 1 lb. assortment pack of CF for $18.99: 12 Piece Assortment Sample Pack - Protech Composites and a 2 lb. box of seconds and scraps for $34.99.

I had a nice conversation with a sales rep who was very willing to try to work with me to get the best of whatever was available for larger sheets, which you'll see the results of in the pictures in an upcoming post.

As an aside, they have an option to order with or without adhesive. I wasn't sure which I wanted, so I opted to go simple and do it without. That meant as the project came together later, I was able to (or forced, depending on how you look at it), use double stick tape temporarily, followed by epoxy for a more permanent solution. If you're pretty sure you want to do this, you might think about just ordering with the adhesive pre-supplied. I'm not sure if it comes on the back of the CF or if it's in a sheet that you can apply to the CF yourself. Either way it might simplify things. If you try it, maybe call them up and ask how the strength compares to epoxy. Post back and let us know if you do!
  Reply With Quote
Old 12th December 2016, 04:30 AM   #4
diyAudio Member
Join Date: May 2010
For the middle of the honeycomb, I was about ready to give up on nomex. Cardboard honeycomb seems like it could've been a great alternative, except even that was surprisingly hard to find for cheap and in less than 1" thickness. I also considered a thin sheet of aluminum or birch ply, with or without holes drilled in it a hexagonal pattern (would reduce weight, but I'm also lazy, so...).

I also ordered a cheap, shallow aluminum heat sink on eBay from China or Hong Kong. It's still not here. Oh well. It occurred to me that it might be hard to get the CF to stick to it anyway. For $5 or so, not a huge loss.

Anyway, I was bored one day and decided to go back to looking for nomex honeycomb, and found out it's available cheap... as a radiator protector for race cars! Honeycomb Radiator Protector, 19 x 26 Inch. $22 for 19 x 26 x 1.5".

Also available on Amazon Prime:
  Reply With Quote
Old 12th December 2016, 04:51 AM   #5
diyAudio Member
Join Date: May 2010
The fun stuff: CONSTRUCTION!

The couple pounds of CF seconds from Pro Tech came in a variety of shapes and sizes, with a slight plurality around the size of a letter sheet of paper, maybe a little bigger. Those seemed like a good size to start with. That also turned out to be a good size to match the 19 x 26" radiator protector cut into quarters:

Click the image to open in full size.

IIRC I had about 5 pieces of CF in the lot that were close to that size.

Next step after cutting them down: try a panel!

Click the image to open in full size.

We used 2" or so wide double stick tape to start with, in case we had to modify the honeycomb sandwiches to make them sound decent.

It's amazing how much stiffer both the honeycomb and carbon fiber get when you sandwich them together, even with tape. Individually, in the sizes I got, you can bend them easily with just your fingers. Together and they feel like you could stand with both feet on top of them without crushing them. You will probably not bend them by hand no matter how much strength you apply.

Assembly by hand, careful to get them centered (although you can peel back up easily and without much risk of damaging anything, I think):

Click the image to open in full size.

Attaching an exciter. Shown is the 40W Dayton Ultra: Dayton Audio DAEX32U-4 Ultra 32mm Exciter 40W 4 Ohm. The two bad reviews on PE refer to reliability issues. I haven't experienced that and may or may not be as mad as those respondents if I do. For now though, the audio quality is excellent.

Click the image to open in full size.

I think we initially started with the 3M VHB backing that came on the exciter. This stuff is so strong though that for purposes of trying this out we thought it'd be better to peel it off and replace with weaker generic double stick tape. Even that still requires something like a putty knife to wedge the foot of the exciter off if you want to remove it, especially if it's been left sit a while. DEFINITELY use a tool to pry your exciter loose if you need to. Otherwise you're pulling directly on the voice coil and seeing whether that or the adhesive breaks loose first.

Demo material:

Click the image to open in full size.

Track 5, "The Sinister Minister", has a great mix of highs and lows, dynamics and ambience that ought to make it entertaining, useful and most of all a FUNKY way to break these muthas in:

Last edited by altie; 12th December 2016 at 04:56 AM.
  Reply With Quote
Old 12th December 2016, 04:52 AM   #6
diyAudio Member
Join Date: May 2010
Is inline image posting broken? I only see a spinning "LOADING" icon in Firefox. You can right click and open in a new tab to view the imgur gallery though: Carbon fiber + nomex honeycomb sandwich DML panel construction - Album on Imgur. Those are in whatever order my phone put them in though, not the order that reflects how the project actually went.
  Reply With Quote
Old 12th December 2016, 05:19 AM   #7
diyAudio Member
Join Date: May 2010
Alternative arrangements: bracing and bass response

Click the image to open in full size.

It's been well documented through both DIY experimentation and theory that DML panel frequency response is highly sensitive to how the panels are held or braced. Suspending them from a string on the back of the exciter sounds different from pinching an edge with a finger, which also sounds different from holding the long or short edges in your hands, which is different in turn from getting a second set of hands and pinching each corner.

For the theory, IIRC, there was someone from Europe - Sweden, Netherlands...? who published a thesis about DMLs that had some illustrations of modal wave numbers along with which areas of the panel would be vibrating relative to each other. There was some distinctive behavior he documented at lower numbered modes where only the four corners would vibrate, then only one pair of edges, then only the other two edges, before moving on to higher and more complex modes. I didn't completely follow the math but it looked as though maybe these lower, simpler modes involving edge vibration with a stationary middle of the panel happen at lower frequencies.

That certainly seemed like it could have been the case based on our experimentation.

The guy creeping into a couple of the pictures is my dad, whose shop this took place in. He's a violin bow maker, which I have to thank for access to tons of extra materials and adhesives. Both he and my mom are almost as into playing with this kind of vaguely sciency free form experimentation as I am.

Anyway, not pictured: if you've got three sets of hands like us, you may find that you're in a position to do something about excess panel edge excursion when you crank the bass up. We found that if you just set the panel on its edge, or hold it between a pinched finger and thumb, it vibrates way too violently and maybe starts to sound kind of like a rhythmic sequence of farts. Funnier than it has any right to be, but not really hi-fi. Next we tried two of us holding the panel stiffly at the four corners. This resulted in a substantial improvement in sound. We had no means to measure on hand, so we don't know for sure that we didn't just cut enough of the bass out entirely to make it sound better by not distorting, but it sure seemed like we were still hearing something just without the excess flapping around and possible CF self-noise. Next step after that was to get the third set of hands involved, so we're gripping the four corners, plus the middle of each of the long edges. That seemed to result in the best sound.

That inspired the image you see just above in this post (I hope). It's the fully built panel with a piece of 3/8" or so thick weather stripping stuck along each of the long edges. That weather stripping is clamped down with a bit of play between the jaws of two hand screw clamps. We used hand screw clamps both because they provide a nice solid parallel surface, and because they seemed like they were probably heavy enough to provide good backing for the damping without us having to hold onto them.

A quirk of our experimental apparatus: I sized my panels to what I thought would be about right to be powered by one Dayton 40W exciter each, but here you see two exciters attached. That's because our audio source was an old Sony integrated 5.1 receiver found at Goodwill for the princely sum of $12.99 (sticker probably still visible). That receiver didn't have a way we could find to turn off the subwoofer output and send full spectrum sound through the front left and front right channels. Based on some of the papers I've read at Tech Doc Library, I suspect that without some extensive modeling and experimentation, we'd probably get better sound out of a single exciter provided that it's powerful enough to drive the panel as hard as we're asking it to. Even so, this setup sounded really, really good. I'm planning to use something like that weather stripping holding the DML panels gently in picture frames as a more semi-permanent fixture.
  Reply With Quote
Old 12th December 2016, 05:37 AM   #8
diyAudio Member
Join Date: May 2010
Subjective results

Clarity and soundstage are excellent. Just like everyone else in the hundred pager megathread though, I've found that they are different and that after a while I think I like them.

I think I mentioned earlier that we don't have much in the way of measurement tools on hand. They do seem to make for some kind of a sense of distance between the listener and stuff that's often more in the foreground of recordings like vocals. I don't know how much of that is down to the soundstage, vs. maybe a frequency response problem that could be corrected by EQ, or coloration inherent to the construction method or particular to the dimensions and driver placement (slightly off center on both axes, as recommended).

Another possibility is that the bi- or omni-directional nature of the speakers is leading to weird reverbs from the back side sound radiation. The panels are currently playing on my desk where they sit in front of my Monitor Audio Silver 3i's (if that tells you anything about what I think of their sound quality). They seem to improve from that position if I stuff a throw pillow in between them and what's behind them. Sorry Monitor. It doesn't totally eliminate the effect... but then again I'm not sure it's unpleasant.

I do have a measurement mic buried in my desk somewhere. That will have to wait for a more permanent fixture though. When I get them framed with weather stripping damping, hung on the wall, and hooked up to a Behringer digital EQ, I'll see what I can do. That may be a month or more out.
  Reply With Quote
Old 12th December 2016, 05:56 AM   #9
diyAudio Member
Join Date: May 2010
Replacing double stick tape with epoxy for more permanent construction

The double stick tape version of the panel sounded decent, but knowing that, I wanted to go with something more permanent that also had a chance of sounding better: epoxy. We went with 5-minute dry, I think polyester based but I'm not sure. The QuickCure stuff we show here was just what we happened to have around. It doesn't seem to be the cheapest but I'm not at all sure it matters.

Click the image to open in full size.

Our gluing jig:

Click the image to open in full size.

On a flat surface, grab two sections of 2 x 4 or something of about that stability and height. Place a flat board on top wide enough to fit your DML panel. The 2 x 4s create space beneath the workspace so that when you've got the sandwich together, you can get clamps on without having to pick the whole sandwich + clamp apparatus up and fool around with it - the clamps fit underneath.

Put down a sheet of wax paper to catch any stray epoxy.

Mix the epoxy in a small container. We found that about 1 fl. oz. total of epoxy + binder was just about the right amount to cover one sheet of CF.

To the side of the jig, put down two more sheets of wax paper, one for each of the CF faces.

Click the image to open in full size.

Spreading the epoxy:

Click the image to open in full size.

Be sure to use gloves. Use a popsicle stick or something similar to lay down your mixed epoxy + binder, and spread with a fiberglass spreader as shown (I gather they're available at hardware stores. We just happened to have a few around. Thanks again dad!).

On our first go around, we just ended up doing a single face at a time because we only mixed up enough epoxy for that one. The single face, plus wet epoxy, plus honeycomb seems to be plenty strong enough to safely clamp in a jig:

Click the image to open in full size.

You can see we put down a second sheet of wax paper on top of the facing. Next comes as flat of a block of wood as we can find, and six clamps to go along the edges, seen here on edge:

Click the image to open in full size.

Leave sit to dry for 5ish minutes. You'll probably have some left over epoxy on the fiberglass spreaders. You can flex the fiberglass spreaders around and see how the epoxy responds to get a sense of how dry the stuff in the jig might be before you decide to disassemble it.

If you mix up 2 fl. oz. of epoxy at a time for panels of roughly the size we did, there's probably no reason to do a single face at a time. On the second panel, we did both faces at the same time from a single batch of epoxy, halving the drying time, and it seemed to work out great.
  Reply With Quote
Old 12th December 2016, 06:57 AM   #10
DAMIC is offline DAMIC  Seychelles
diyAudio Member
Join Date: Jan 2006
Location: Beograd
Unable to open any of pic's, posts 5-9.
  Reply With Quote


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off

Similar Threads
Thread Thread Starter Forum Replies Last Post
Honeycomb composite panel enclosure, dual opposed configuration rhapsodee Subwoofers 6 1st September 2010 04:26 PM
A Carbon Fiber store!! DaveThreshold Analogue Source 9 11th May 2009 05:52 AM
carbon fiber earing Instruments and Amps 8 13th March 2007 06:04 AM
How do I use carbon fiber + MDF wood + glass fiber to make speaker? sonyacer Multi-Way 16 10th May 2004 04:57 PM
CArbon fiber enclosure jouch Multi-Way 1 14th March 2002 11:49 AM

New To Site? Need Help?

All times are GMT. The time now is 10:07 AM.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright 1999-2017 diyAudio

Content Relevant URLs by vBSEO 3.3.2