[sedge]

morning
I have used a few different  coatings but pva is flexible to use ,sounds good and is used to glue the exciter on,its also cheep, basically it ticked all the boxes,the fact that I had a gallon in the shed had nothing to do with it 
you can apply as much or as little pva to as much or as little of  the panel as you like , to adjust the sound to how you want it ,removing the skin or coating it,it gives us more options to play with and you can choose the sound you want and I think that is very important .
the 1 to 1 mix of pva and water was just thrown together in a jam jar ,it worked well so I kept using it , if you feel you want to use it thicker or thinner go ahead and give it a try, I would say though that gluing the exciter on with pva is the best option for good HF beyond 20K,rather than the tape ,which tends to damp HF.I have used thick waterproof pva to glue exciters on to fill uneven gaps in some panels, as there is much less shrinking and fills the gaps ,I find that using as little as possible works well.
In another tech sheet I was reading a few days ago they mentioned using Loctite 4212 to glue onto rohacell 51 panels ,a foam type 3.5 mm thick ?
steve.

[sedge]

on the tech sheet EP1349407 B1 the nxt driver uses 2 small weights to keep control of the area in the middle of the coil, as I said before I used to (years ago)use blu-tack which cured the peeks and dips , but to me adding weight to the exciter centre just slowed down the sound, I'm talking about a 2mm ball aprox pressed on to the panel with a wetted finger, obviously after some months It would ping off, on harder panels such as ply I used to use strips crossed in the centre to guide the sound waves away.
steve

[j gale]

I have wondered for awhile if the sound of EPS or the extruded stuff could be changed with a coating but, didn't''get around to trying it. NOw I will. Thanks Sedge for the info. I also wonder if a PVA coated panel would benefit from a coat of shellac on top or if that would make it sound harsh. One question I have for you Sedge--When you talk about poly and EPS in the same sentence, do you mean that the poly is the extruded stuff? If so do you prefer it to EPS? What you said about the PVA becoming the skin makes sense to me. It makes me happy that doing these experiments will cost very little and I'm encouraged by the fact that you are happy with the sound from treated panels. I noticed that in previous posts you made the panel thinner in the vicinity of the exciter. Do you recommend that? If so, why?

[zygadr]

I second j gale's questions here sedge.
We need to know if you are using the infamous VH grade of "eps" or something else?.... There is a big difference here.

By hollowing out the section of the panel where the exciter is mounted may possibly affect the whole principle of a DML and bring us towards the realm of point source twin cone loudspeaker territory????.......just thinking ot loud here?

[zygadr]

Sedge, what thickness is your panel?

[jeffac]

Warning – very long post (and my apologies but I missed this rash of posts while drafting this tome)

Terrific stuff steve,

I also figured that the thin pva film would provide a ‘surface-skin’ to aid wave propagation across the panel back and also negate some of the perceived hollowness to the sound of VH-EPS. My panels were cut from a block, and thus have no ’skin’, which from what you report in your investigations might provide a clue as to why I’ve not had too much concern about that “unnatural” EPS sound. I’ve wondered about also coating the panel front surface…….. so might be time to experiment…

I took some photos during building my Mark 1 VH-EPS panels (see below). These have moved on to Mark 2 and 3, but some of these and my meanderings below might provide ideas for panel fiddling and panel support, which can prove to be a struggle to find a workable/flexible/visually acceptable solution. Maybe not quite as important as panel material selection, size, exciter positioning and fixing, etc, but still a significant challenge.

So…

1.   As I didn’t like the idea of squared panel edges, I rounder them over using coarse and then fine sandpaper glued to small diameter pipe pieces cut in half. How edge rounding affects edge diffraction and rear/front panel interaction I’m not exactly sure, but I figured it wouldn’t hurt based on what I’ve gleaned from open baffle/dipole studies… although DML being something very different will behave quite differently

2.   Rounded panel corners because this is known to avoid corner-related problems

3.   Initially used an exciter mounting support spine, but IMO this spine creates more trouble than it’s worth. Why? It’s very tricky to get the exciter-support damping right with whatever mounting material is used, and get it wrong and you’ll stress the exciter voice coil when trying to push the panel pistonically with low bass.  With the el-cheapo but good exciters I have, I’ve had no problems with voice coil sag when attached only to the panel by their foot. If the DML panel is vertical, this I think helps the exciter attachment cause. On some initial dinky tilted-back test cardboard panels that I flogged because they sounded so good, I did manage to detach voice-coil spiders from the exciter edge. But that was with panels tilted back, since then, I’ve had no exciters falling off using 3M VHB tape, and no drooping or spiders detaching. Additionally, using VH-EPS, whatever spine (I used 43 x 19 mm pine in Mark 1) is used for exciter support will reflect enough sound to be easily heard on the front side of the panel, and I didn’t like what I heard……. So for more sensitive panel materials like VH-EPS, I’m not convinced that an exciter mounting support brace really helps the cause

4.   Because of sound reflections transmitting back through the panel, to try to mitigate at least some of this, I used contact adhesive to glue strips of outdoor/car-type carpet to the support frame edges (again cheap 43 x 19 mm pine). I also wanted this carpet covering to provide infinite flexibility in panel attachment to the frame

5.   The frame support I made approx. 5 cm smaller than the panel width/height, so as not to be seen and to leave 2.5 cm of each panel edge extending beyond the frame seeing that I’d gone to the trouble of rounding them to remove their edges, I wanted these edges to float free

6.   For panel mounting to the frame, I used the “hook” side of Velcro stuck to the panel. Obviously Velcro is very light in small pieces and thus will affect panel vibration minimally. This allows lots of flexibility in where YOU think panel-frame attachment will work best for panel edge rigidity or alternatively flexibility in your situation, and this to so some extent is voodoo or trial-and-error and dependent on all matter of things related to panel dimensions, frame position, exciter placement etc etc…… 3M VHB tape was used to attached the Velcro to the panel as the glue used on the cheap stuff I tried first didn’t attach it securely enough long-term, and some Velcro pieces began to detach from the panel with extended use.

7.   The panel bottom I have sitting on 2 x small car carpet pieces doubled-up to hold it up, which leaves the side and top Velcro attachment pieces to work in the horizontal plane only to hold the panel vertical and to resist against/damp vibrations. It’s worth Googling images of how Velcro works – for me it seems an almost ideal panel attachment material, being very loosely connected by the thousands of little hooks to allow the minute high frequency panel deflections to pass by it with little alteration/attenuation but providing some damping for low frequency pistonic movement, and of course holding the whole kit and caboodle together. Interestingly, even if the Velcro is attaching lightly it seems to be able to self-tighten with time due to more “hooks” attaching through repeated vibration. 

8.   For spaces between the “hook” Velcro pieces, pieces of “loop” Velcro can be attached to the panel as a bumper against the car carpeted frame to dampen pistonic movement to what you think is acceptable. Better to use multiple small pieces 1-2 cm long positioned at 10-15 cm spacing’s to avoid any “mild contact noise” that can be heard with ear very close to the panel on transient mid-bass if longer, more distantly-spaced pieces are used

9.   For larger panels one or more cross-braces can also assist in avoiding pistonic chaos with low bass. Short pieces (again to avoid being heard) of car carpet can be placed between the carpet glued to the brace and the panel to help avoid this chaos of the panel wobbling in-out for too long beyond the bass note. Not absolutely essential to glue car carpet pieces to the panel for this as they tend to hold themselves in place. Carpet fuzzy side facing the panel to minimize noise.

10.   To increase high frequency energy, I’ve played with piezo tweeter cones with their flat outer flange glued to the panel with pva and connected in parallel to the exciters. I’ve just connected them naked, but there are easy means of rolling off low frequencies and reducing SPL that can be considered. I used the cheapest of cheap useful >3.5 kHz type piezo cones as the larger and more expensive >1.5 kHz types reached down more into vocal ranges that at the time I was playing about with them, seemed to add excessively to unnaturalness. They do tend to make the VH-EPS panels hyper-detailed, but not ‘too hot’ for me as I'm getting on and can’t hear much above 14 kHz and as I like HF energy and have H-frame woofers to help balance this HF energy with LF energy. And I can easily disconnect one or more if my tastes change.

Hopefully the few old photos below will help decipher the musings above, and maybe provide a few additional panel building ideas.

I’m still learning and thinking about ways to make better panels and there is terrific information posted in this tread that is guiding this adventure, so thanks to everyone for contributing to try and DIY DML panels cheaply that push the fidelity envelope.

cheers.. jeffac

This is where it all started, half a cardboard box


Some origami and hey presto, we have DML panel


Attach some $2.50 exciters, and these little pieces of cardboard sounded awesome and got me hooked on DML


OK something more serious, VH-EPS sheets with test round-over of panel edges on one of the freebee protector sheets provided for the 8 good sheets, better job done on the real panels


Frame with car carpet attached


Car carpet close-up, notice wood removed for a spine that was omitted in the Mark 2 panel


Velcro hook strip


Velcro hook strip back with 3M VHB tape attached


Mark 1 VH-EPS panel with exciters/piezo cones/Velcro attached


Piezo cone close-up showing VH-EPS panel surface ‘roughness’ from the cut


Exciter on VH-EPS panel surface, note panel sanded smooth with fine-grain sandpaper prior to attachment using 3M VHB double-sided tape. Initially used 5 um thick tape, as that is what I had, but now have 2 um VHB tape and this works far better in close coupling of the exciter to the panel surface, never had an exciter detach even with the panels being pounded regularly


Mark 1 panel back showing some of the support frame


And some finished DML panels

[sedge]

hi all
jgale
when I say poly I generally mean foam and eps , they sound much the same (to me) , and the treatment is the same, the difference is that the foam can be thinner for smaller panels and the eps can be thicker for larger panels, although I have not tried 1inch foam, a smallish 5mm panel will go down to say 1or 2 hundred HZ mounted freely, and a larger 25mm eps will go down to 20HZ or so freely mounted (depending on room problems).
I have had problems with el cheapo exciters falling apart which could be from poor quality control ,hence their price, but running them down into the sub-sonics is always going to be a problem, I believe they came with capacitors fore a reason,also some of the lead out wires are a little short or loose and can bang on the spider.
I still have some 2inch bandor drive units that I had to get repaired because I was driving them to low ,Doreen herself (lovely lady)moaned about people who think they can run them lower 
I roll my large panes off at about 40 to 80 HZ depending on the panel used.
As for vh eps I only have ever had 2 40cm sample panels which are very rigid for the size which is a problem for me.the larger panels will have more flex in them,but that's something else to talk about another time.
must go.
steve

[sedge]

rob
If you remember back on this forum somewhere, and in some tech papers, the discussion on how many ways in which these panels radiate sound ,one of the ways it radiates is as a direct radiator in the area of the exciter, this is easy to hear if you stand up close to the panel ,I was looking at ways to minimise the thickness of the panel and to increase the small detailed sounds of the panel and keep the rigidity, at first I made a small 1inch inverted dome type tweeter shape, and then just  carried on even thinner till I reached a 2inch inverted dome(this is on a 25mm panel).this I did because of some comments made on this site or your other site about the sound not being very intimate sounding. its just another option If you feel the need but not mandatory. but works quite well. maybe If I had the knowhow I would make the shape of a constant directivity horn ?
you can push these panels heavily into nxt dml type panels or more into a point source type panel the choice is yours , as far as I know these are all unavoidable principals of a dml panel
on another post here I should have said the exciter area acts as a direct radiator,  not as I said, a piston, hope that clears that up.
steve

   

[sedge]

jeffec
that's a very dusty job sanding the edges ,I did the same years back with a 25mm eps panel ,still don't know if it was worth it ,pull it out now and again to have a listen
looks more professional than mine, keep up the good work.
just remembered It was that panel I striped both sides(skins) off ,I have coated the back of the panel, but only a 4 or so square inches of the exciter area on the front, sounds ok as far as I can remember ,never got round to enlarging the front coated area
steve

[zygadr]

My God!!!
I can not believe what a monster of a post I have created!!??????
You guys have impressed me immensely with what you have done with your experimentation and end results!
I am truly honored that after so much effort, worry,and at times, massive disappointment and heartbreak, you have continued to not give up and carry on with this simple but very effective technology that brings a new world to owning an inexpensive panel loudspeaker.
I am truly touched!
Rob.

[sedge]

rob
In the patent I mentioned  EP 1340407 b1 the first fig shows the masses they apply to the inner area of the coil on a bmr driver, but in figs 15 to 18 they show methods of thinning and fattening  the panel but only on the driver side ? they use a large coil, you would not get a lot of thinning on a 25mm coil. Its just a slight difference having the thinning on the front and the larger cone shape I think helps throw the sound forward ?
wish I had seen this back in 2008  so many bits of info all over the place ,this could probably all have been sorted years ago If we had it at hand in one place
steve

[zygadr]

Ok.......some news on a possible contender for a panel material.
Today I had a guy over to provide a quote on a new garage door for insurance purposes.He inadvertently mentioned a range of "HONEYCOMB" polycarbonate panels used in the sections of garage doors.
I examined his 4inch square samples of many different colours and my mind began to tick over!

Even though the samples were very small, they were also very light and indeed had a honeycomb interior cross section.
I will state right now that this material was not the softish Corflute stuff. It was rock hard plastic and if (!?)available in large sizes, it may be a very interesting test material for panels.

I will ring the door company tomorrow to find out if I can obtain a large sample and what the maximum size available will be. Talk about everything happening at the last minute sedge and all!!!!!!

[sedge]

zygadr
as if garage doors were not enough ,have you seen piezo drivers made by sonitron can handle 60v peek to peek and supply their own amps. only mentioned it as it was how I saw your other thread had started ,not too expensive now ,I saw this a year or two ago and am sure it cost a lot more.
steve

[zygadr]

Sedge, I wonder if those piezo drivers are what they are cranked up to be?
When I saw these PC  honeycomb samples, it reminded me of a guy on the net that had some positive success with thin solid Acrylic apart from the usual sensitivity issues    ........so who knows, it may be a waste of time, or maybe not?
My guess is that it will be either unavailable in larger sizes or way too heavy........ and we don't want that crap again!!!!!!

[zygadr]

A word of warning ladies and gentleman, boys and girls!!!!!!
I may have been too quick to judge VH EPS....... I used two "new"exciters on my remaining test panels and have suddenly noticed a massive sound quality improvement in week 2 of use!!!
Break in possibly?............I would say so............STUPID, STUPID!!!!!! 

[jeffac]

sedge,

Rounding over the edges wasn't too painful or messy with my thin (10 mm) VH-EPS panels, it's a very brittle material and there wasn't that much corner to wear to round compared to the much thicker panels you've been experimenting with. Plus I just tend to chip away at stuff when I get time. I'll get to examining addition of a PVA skin front side when I can, but TBH I'm pretty contented so don't what to mess too much with what I have now.. until I retire at least when time won't be so much the conqueror.

zygadr,

VH-EPS is very good!! but burn-in?.. maybe new exciter spiders/voice coils loosening up...  I've never ceased to be impressed with the sound of my large VH-EPS panels even without having exciters positioned ideally. It's got to be very close to an ideal panel material, particularly if very subtle changes like adding a minutely thin PVA "skin" can shift its tonal balance towards the PVA skin itself. Does make you wonder how much other thin sink coatings might alter its sound. Relatively cheap too if you have a foam manufacturer nearby like I do, 8 x door-sized sheets cut to size for about AUS$65 was a great deal at the time. I built some reasonably substantial timber frames/supports for my original tongue-oil permeated 4 mm single-flute CC-cardboard panels that I was playing with years ago. I haven't got these panels sorted sound-wise exactly as a want yet, but I'll try and post pics of these sometime as they might provide a few more panel support ideas to anyone interested.

Keep the faith... DML panels really are a revolutionary rather than evolutionary speaker technology that deserves a spotlight like this thread to iron out the wrinkles and spread the word.

cheers.. jeffac

[oldschoolVlad]

... I was wrong...    ... Rob, I'm really sorry about my wrong advise on DAEX32EP-4 thruster. In fact it is 2 times more forceful than your DAEX58. I realise that doing some calculations of my own exciter. Because the force on the coil imposed by the magnet is BL multiplied by the current through the coil. Both exciters have aprox. equal BL, but he thruster is a 4Ohm one, hence the current through the coil is 2 times stronger. So, that thruster might drive plywood +3dB louder than DAEX58. 

[sedge]

Jeffec
When I sanded the vh panel it made the surface feel soft and a little fluffy,does the cutting you have on your surface do the same?,I was just pondering on whether the sanding might damp the sound  a little bit more.
Suggest that you do as I did and only coat 2 to 3inches of the exciter area to start with and see if you like the sound ,depending on whether the sanding is softer ,you can easily sand it off.
Zygadr
As you say ,the piezo units might be no good ,but it is a bit tempting 32 pounds for amp and piezo unit mono ,Xmas is coming the piezo roll off at 200 but that could be the 80mm or so size .
Your test panels, are they coated ?

Steve

[jeffac]

sedge,

Not sure the rounded edges are softer, they're certainly not fuzzy macroscopically as very fine sandpaper was used for final smoothing... they feel velvety smooth to touch. Fingernail scratch test indicates the rounded edges sound softer than the panel surface, but with cut VH-EPS, this could be because the surface is quite rough as can be seen in one of my pics. So I don't know really... maybe rounding damps sound at the panel edges compared to square-edged panels. I've got a gut feeling that it should help eliminate vibration diffraction nasties caused by hard corners, but have zero evidence to support this. I'll have to set up spare panel sheets to listen how square-edged ones sound.

Have litres of PVA but will play with drops.

cheers.. jeffac

[zygadr]

Hi all. When I meant burn in, I was referring to the exciters............sorry about the confusion.
Interesting to hear about the thruster Dayton exciter Vlad. These will be ordered today hopefully.
I cannot help but imagine what the thruster exciter will perform like on a large EPS panel?????

My current VH EPS panels are not coated as yet. I am happy with the tonality at this point in time. With a good quality recording, the mids and highs remind me of an electrostatic quality similar to my electrostatic phones.
As they are from a similar family of plastics, (sort of) it's not surprising I suppose, even though there are two totally different operating principles used.

I must point out that using multiple exciters in the past (on VH EPS) has NEVER produced such a sensational top end such as that when using a single exciter which I am using at the moment!!
Is the use of multiple exciters a serious compromise towards sound quality?!
Rob.