Great feedback from DMLBES and BurntCoil. I am busy making notes all over in preparation to jump into DML’s myself hopefully. In case you have not seen it - here is some nice info from Rich Meinke’s thread on Parts Express regarding EPS/XPS panels:
When we are listening to the panel, we are listening mainly to the panel's "skin". The thinner the better it seems so even a lightweight paper has a pretty large impact on the sound. Its amazing to hear the impact that a 1:1 water diluted PVA (white/wood) glue has on the sound. Thin as hell but it damps the panel well and after hardening it improves the HF response. Shellac is said to over damp the panel.
Panel Prep
Whats important to make the EPX/XPS panels sound natural is to remove the panels "skin". 100 grit with my palm sander does the trick. Round the corners (I user a 6 in. diameter Pyrex bowl to mark it with a pen) using a steak/serrated knife leaving plenty of room off the marked pen line to allow for cleanup sanding with the palm sander. I use a 3/8 in. quarter-round bit on my (big) router table to knock off the edges and form a rough contour. You have to hand sand to finish. First with 100 grit in long straight strokes to even up the less then perfect router work. Then feather the panels from the panels surface/edge down along the contour of the quarter-round surface. Finish with 200 grit.
Order of events. 1 - Round the corners of the panel. 2. If you use a router, round the edges. 3. Sand the panel skin off. 4. Finish by hand sanding the edges.
Panel Treatment
Once panels are ready, I currently put 2 coats of 1:1 water and wood glue on the back surface/exciter side and 1 on the front. Use a blow dryer to speed the process... you can treat both panels fairly quickly with a foam roller. When done, you will get a sandpaper like surface. When you brush your hand over the surface... even lightly...it will be amplified and quite loud. It is this surface combined with the ultra light and ridged XPS panel that gives the panel it natural sound. Untreated panels do not sound Hi-Fi.
I'm accumulating (facts) on my side, not summarizing - it was a straight copy and paste from this informative post by Rich M:
DIY Flat Panel Speaker Love -
Techtalk Speaker Building, Audio, Video Discussion Forum
I have some experience making carbon fiber composites, and I was actually looking forward to building my own composite DML panels eventually. But, since it is expensive AND time consuming, I'd really like to be pretty sure I know what to make before I make it, and I'm not there yet!
As I've mentioned, I do like the accuracy of wood, but it's efficiency is a problem. But so far, everything I've tried so far with better efficiency just doesn't sound as good. My next step may be do try the 40 watt thrusters and see if that gives me a sufficient boost.
XPS has great efficiency, but also that hollow sound you mentioned. I'll try the PVA coating, but if that doesn't fix the hollow sound I'm not really interested. I have tried a foam surround on XPS and that didn't seem to fix it.
I was really optimistic that Gatorboard (basically PS foam with particle board skins) would be great, combining (hopefully) the accuracy of wood with the efficiency of foam. But unfortunately, to me it sounds as hollow as the foam alone.
I'm worried I'll spend a lot of time and money making composite panels and the'll end up sounding no better than XPS....
Eric
I haven’t seen the following products mentioned as panel candidates:
Cellulose Foam Board
https://www.lowes.com/pd/Common-0-1...faced-Cellulose-Foam-Board-Insulation/3369220
Polyisocyanurate Foam Board
https://www.lowes.com/pd/Rmax-R-Mat...ft-x-8-ft-Actual-0-5-in-x-4-ft-x-8-ft/3014172
Polystyrene Foam Panel
https://www.lowes.com/pd/Insulfoam-...ene-Garage-Door-Foam-Board-Insulation/3033278
Any thoughts or experiences?
Cellulose Foam Board
https://www.lowes.com/pd/Common-0-1...faced-Cellulose-Foam-Board-Insulation/3369220
Polyisocyanurate Foam Board
https://www.lowes.com/pd/Rmax-R-Mat...ft-x-8-ft-Actual-0-5-in-x-4-ft-x-8-ft/3014172
Polystyrene Foam Panel
https://www.lowes.com/pd/Insulfoam-...ene-Garage-Door-Foam-Board-Insulation/3033278
Any thoughts or experiences?
Carbon fiber composites is not the same as Honeycomb (paper/Kevlar) composites so they are not going to sound the same. Like I said before with certain materials the down fall is the cost. It wouldn't be to bad if one had already perfected there design with the more costly materials but to do experiments and fail will surely cost you a lot. One of the reasons for the huge hike in price of commercialized speakers is the experimentation (research and development) process which could cost in the hundreds of thousands to even a million and some of that cost will be passed on to the consumer. Personally I spent over 2k in experiments but my final product design only cost me around $200.00 pair (or less) to build. If I knew back then what I know now I would have saved me a whole lot of money. lol
Compromise to wood is efficiency. Higher powered exciters can help like the Thruster or the Steered high flux exciters like the DAEX25SHF-4 or the DAEX30HESF-4.
A single technique (like using pva glue) will have a subtle change in sound. The combination of many multiple techniques will significantly change the sound. That is why I stress All basic foundation techniques like rounding the corners, sanding panel, pva glue, using a frame and spine, using some sort of damping material to attach panel diaphragm to frame, using 3-4inch frames and recess material into frame to act as a wave guide. These are standard foundation techniques to build upon, without a solid foundation it will be very difficult to build anything on top of it.
Experiment with Gator board by peeling of the ply wood in the front side of panel but leaving the ply wood on where the exciter resides and see if it sounds better. The back side were the exciter resides can be harder/stiffer then the front. For those who didn't catch on even Rich M. uses 2 coats of PVA glue for the back where the exciter resides and 1 coat on the front because you want the back side stiffer/harder. If the front is to damped the vibrations cant pierce through it. Like I said before you can combine different materials together, one just needs to know how to do it right.
Time and money mostly come from experiments. Each material will have different designs due to different characteristics of each material. I've tried a small piece of carbon fiber and I didn't care for its sound I preferred the EPS. If you have the time and money then go for it but it not I wouldn't venture in that direction as experiments can get really costly especially on expensive material.
Another fine summary from DMLBES. I'm pleased to say that I'm following pretty well all the foundational guidelines in the post above. My panels will be up and running later today, although the spine part is waiting on something arriving in the post.
One thought. Plywood has been mentioned, specifically birch and spruce. An alternative might be poplar ply. It's lighter than either spruce or especially birch and 3mm poplar is very affordable (just over £3 for a 600x400mm sheet plus postage). Perhaps this is worth a go too, with or without a thinned PVA treatment.
One thought. Plywood has been mentioned, specifically birch and spruce. An alternative might be poplar ply. It's lighter than either spruce or especially birch and 3mm poplar is very affordable (just over £3 for a 600x400mm sheet plus postage). Perhaps this is worth a go too, with or without a thinned PVA treatment.
True I'm sure. I didn't mean to imply that I necessarily thought carbon fiber was the right material. I meant only that I had experience with composite construction and would enjoy designing and building composite panels of whatever construction should perform the best. In terms of construction techniques, building with Kevlar is virtually identical to carbon.
I think polar plywood is a good choice indeed, for the reason you mentioned (light weight, or more specifically, lower density). I've been looking for it myself, but have not found exactly what I want here in the US. I bought some 1/4 inch "poplar" plywood, but it's only thin poplar veneers over a unspecified core. It's still pretty light but also pretty warped. It sounds about the same as my "sureply" underlayment plywood, but the sureply is nice and flat. If I can find a better poplar plywood myself I will definitely try it.
Eric
200mm x 300mm x 5mm carbon fiber honeycomb panel for $38 shipping included.
This might be a good mid/tweeter for DML?
AliExpress
This might be a good mid/tweeter for DML?
AliExpress
The poplar ply here tends to be 100% poplar. However, it can be prone to warping, especially as it's so thin. I may have my panels up and running later today (EPS, PVA etc.).
Well, the panels are finished and I'm currently testing before getting down to the serious business of doing measurements and adjustments on the miniDSP after the rugby tomorrow.
I guess the first observation is that the technology actually works at all! Secondly, the panel (EPS, PVA, Dayton Audio DAEX25FHE-4) is much more efficient than the coned drivers (about 90db) I was using before. XO is at 160hz LR24 with a 15 inch Monacor PA bass driver crossing over to the panel. Other elements are pretty much as per DMLBES's recommendations.
EQ is obviously needed, but treble seems pretty good.
If the system makes a decent fist of Radiohead's Reckoner from In Rainbows, then we're in with a fighting chance, since this sounds like an incomprehensible mess on poorer systems.
I guess the first observation is that the technology actually works at all! Secondly, the panel (EPS, PVA, Dayton Audio DAEX25FHE-4) is much more efficient than the coned drivers (about 90db) I was using before. XO is at 160hz LR24 with a 15 inch Monacor PA bass driver crossing over to the panel. Other elements are pretty much as per DMLBES's recommendations.
EQ is obviously needed, but treble seems pretty good.
If the system makes a decent fist of Radiohead's Reckoner from In Rainbows, then we're in with a fighting chance, since this sounds like an incomprehensible mess on poorer systems.
I have now had a chance to do some listening and measuring, the latter with mixed results.
Overall, the approach has some merit. Initial impressions were of a rather 'papery, scratchy' sound. This has dissipated to a degree, but far from completely. There's a definite lack of warmth overall. Treble falls rapidly after 7khz. However, there's a large hump centred around 4950hz, starting from about 3300hz and coming back down by around 7000hz and continuing to fall.
I have not heard the distant or hollow sound referred to by some. I'm used to, and prefer a soundstage spread behind the speakers, not in front of them.
The key issue then is tonal balance. I will draft in a SEAS 27TDFC for duties above 7khz. In addition, I have some 100% 3mm poplar plywood on order which should arrive mid-week. It will be easy enough to try plywood rather than EPS and I am anticipating an improvement in the overall tonal balance. One can go so far with the miniDSP, but I suspect that it's the overall sonic character of the EPS that's currently not quite doing it for me.
Overall, the approach has some merit. Initial impressions were of a rather 'papery, scratchy' sound. This has dissipated to a degree, but far from completely. There's a definite lack of warmth overall. Treble falls rapidly after 7khz. However, there's a large hump centred around 4950hz, starting from about 3300hz and coming back down by around 7000hz and continuing to fall.
I have not heard the distant or hollow sound referred to by some. I'm used to, and prefer a soundstage spread behind the speakers, not in front of them.
The key issue then is tonal balance. I will draft in a SEAS 27TDFC for duties above 7khz. In addition, I have some 100% 3mm poplar plywood on order which should arrive mid-week. It will be easy enough to try plywood rather than EPS and I am anticipating an improvement in the overall tonal balance. One can go so far with the miniDSP, but I suspect that it's the overall sonic character of the EPS that's currently not quite doing it for me.
I'm eager to hear what you think of the poplar. I think you will like the sound quality better, but it's still far from a perfect material. As I'm sure you are aware, you can expect a big decrease in efficiency. You may find you need two (or more) of the DAEX25FHE-4 exciters per panel to get decent output. And don't be surprised if you still need a lot of EQ. My DML "laboratory" amplifier has no equalizer, only bass and treble controls. And while the plywood panels are my favorite so far, they still require turning both bass and treble way up to get close to a flat overall response.
What size are your EPS panels? What size are you planning for the poplar panels? Sorry if I missed it, but how are you supporting your panels, and where did you position the exciter(s)?
Eric
Hi Eric
I'm using a miniDSP for EQing, so have quite a few parametric equaliser levers to pull. The panels are 400mm x 564mm (1:1.41 ratio as per NXT technical document). They are set mid-way back in a 70mm deep slightly oversized frame and secured to the frame with as little closed cell foam tape as I can get away with (including avoiding the corners). The frame site atop a 15 inch Monacor bass driver in a U frame.
The exciter is positioned in the Monacor position (4/9 X, 3/7 Y) and secured to the frame via an MDF spine, with some butyl rubber to help with decoupling (not perfect due to screws in frame, but better than nothing).
I can comfortably cross over over at 320hz with anything from a 6db Bessel to an 48db LR one. This of course frees up the panel for mid/tweeter duties.
I can get away with reduced efficiency, especially if the midrange performance is a little flatter. There is obviously no guarantee of this! A second exciter may or may not be needed, so we shall see.
One thing I cannot decide on is whether to apply a thin coat of diluted PVA to the sanded panel, or indeed some French polish.
Frames
From your posts and many other posts in other threads it looks like using a frame and spine can contribute much to better sounding panels (a basic foundation or requirement according to you). I have seen a few (not many) pictures of frames, but what I am intrigued about is the statement you made about the frame to act as a wave guide. Sorry, I actually know nothing about wave guides and their shapes. Could you possibly expand a bit on this and maybe share a picture or two? Because the frames I have seen to date are just very straight forward rectangular pieces of wood. Do you imply that these frames could be profiled with round inner edges like some picture frames maybe? And will that have an effect on the sound?
DMLBES – thank you, I find your contributions to this thread very valuable and I am busy accumulating notes. For newbies like me it is rather simple to get started by hanging the panels with strins from a beam or the ceiling, or even just letting them stand on a chair like R Meinke’s panels. I’m trying to get my head around building frames from the start.
From your posts and many other posts in other threads it looks like using a frame and spine can contribute much to better sounding panels (a basic foundation or requirement according to you). I have seen a few (not many) pictures of frames, but what I am intrigued about is the statement you made about the frame to act as a wave guide. Sorry, I actually know nothing about wave guides and their shapes. Could you possibly expand a bit on this and maybe share a picture or two? Because the frames I have seen to date are just very straight forward rectangular pieces of wood. Do you imply that these frames could be profiled with round inner edges like some picture frames maybe? And will that have an effect on the sound?
Last edited:
Its just a rectangle wooden frame. The wood used for the frames should be 4inches width. Most use 1inch EPS/XPS so it will be in the middle of the 4inch frame leaving 1-1/2 inch gap in the front and 1-1/2 gap in the back which basically sort of acts like a wave guide.
Picture of frame in link > 2x4 Furniture Pdf | Easy-To-Follow How To build a DIY Woodworking Projects. :Wood
second picture of a pair of frames.
This technique helps focus the sound making it sound less diffuse and more accurate sounding.
Another one of my techniques of adhering panel to frame is to use Alexandria 3/4inch molding with 3/4 Frost king foam on one side of the molding. The frost king foam side is adhered to the panel material while the other side of the molding is adhered to the frame using wood glue and clamps. This needs to be duplicated on the back side as well to SANDWICH the panel together to firmly hold it in place. This technique will reduce unwanted resonance.
https://www.homedepot.com/p/Alexand...med-Shoe-Base-Moulding-WM126-90096C/205958745
Rubber Foam Weatherseal | Frost King(R) Weatherization Products
Besides reducing resonance this technique will help the bass jump off the panel due to acting like a wave guide within the frame because its curved shape helps to direct bass frequencies.
This is a picture of a wave guide used in conventional horn speakers > Community 1" Exponential Horn 90x40 Non-Standard 2-Bolt for UC1 Compression Driver
See how its shape gradually curves and tapers outward and it helps guide the frequencies outwards.
I was reluctant to give out this technique but most don't listen to what I say anyways as they rather hang there panels by strings. LOL
I for one cannot see how hanging panels with strings is going to work as well as some kind of clamping or securing with foam tape. Surely one needs to allow the panel to do its thing without being free to waft around.
Anyway, it's time to cut the poplar plywood. It's half the density or thereabouts of birch ply, but with a similar modulus of elasticity (which I take to be bendability).