Depending on where the driver is and how it is supported, you may not get a classic 1,2 mode.
It would be interesting to see the results of a chladni test
I see Xcite now have a mounting ring for there exciters, seems odd that the picture has it screwed to the panel it's attached too, I'd think that would constrain the movement and sound output?
They also have an 8ohm version which I guess could be good for other options of multi exciter panels.
They also have an 8ohm version which I guess could be good for other options of multi exciter panels.
It's attached to a bracing panel behind the driver, and not to the moving membrane.
I’m not finding those images on the website. Can you share a link or direct me to the menu item under which to find them?
The key point is to avoid the loose screws! The vibrations at the exciters are terrible. Strange design (but no other choice?) with one screw axis in the membrane plane in addition with metal insert.. In the bottom mounting (see the spec linked in the above Sound Import page) the 4 screws are no more accessible but the fifth one yes. In the top mounting, it is the opposite. Some advice about theadlocker might be welcome. Why not having 4 small brackets with hole like in DAEX serie directlly from moulding? In my last design which is a sound reinforcement panel (reduce bandwidth PA?), with a DAEX30FHE, I glued a stud at the exciter central thread and then a nut. After a 2 hours test, one nut was lost, the other one loose. Next test with self-locking nuts (nilstop?). Good point, the 3M adhesive is also available as spare part at Sound Import for Xciter exciters.
The photo shows the screws inserted from the front of the driver, from the panel side. But there's no reason they cannot be inserted from the brace side.
This is how I've supported my drivers. No screws...
Those are simply 55mm holes drilled with a standard hole saw. It's a perfect press-fit for the magnets. But I would prefer a 56mm hole for a slightly looser fit. I would then adhere the driver to the front panel first before running bead of silicon around the back of the magnet where it sits in the hole.
As it is, with a tight fit, the drivers are aligned, or possibly mis-aligned, with each other and with the panel. I have to very carefully mount the brace first, with the drivers away from the panel, and then when everything is in place, very carefully encourage the drivers to mate with the panel in such a way that there's no misalignment. Not easy.
This is how I've supported my drivers. No screws...
Those are simply 55mm holes drilled with a standard hole saw. It's a perfect press-fit for the magnets. But I would prefer a 56mm hole for a slightly looser fit. I would then adhere the driver to the front panel first before running bead of silicon around the back of the magnet where it sits in the hole.
As it is, with a tight fit, the drivers are aligned, or possibly mis-aligned, with each other and with the panel. I have to very carefully mount the brace first, with the drivers away from the panel, and then when everything is in place, very carefully encourage the drivers to mate with the panel in such a way that there's no misalignment. Not easy.
Christian, the vibrations are extreme only around the resonant frequency of the assembly? (around 30Hz). I find that at anything above, say, 100Hz, there's almost no magnet vibration at all.
Good remark! The first part of my test was done "full range" but as the displacement was too important in the bass, I then used a filter. Maybe the nuts came loose in the first part of the test. It makes sense. Currently my panels run mostly in full range (in room application, moderate level). The HW to get a filtering possibility is on progress.
Hi Andre,
Is this one you used with a single exciter or 4 of them? What does the foam around the edges help with?
On my quick dirty canvas test panel I did have much improved bass response but at louder volumes there seemed to be distortion coming from around the edges, I'm not sure if it was the canvas to wood edge or the EPS canvas edge.
When you say it has nice punchy bass I assume it can play at a loud volume nicely? I am lacking in that from my current workshop panels. These are HD Eps 300/500mm in a frame held in place with epdm foam and pressure in a few places. The panels measured to around 110hz before dropping off in close measurement but at 3m in the room they had less output below 200hz than I am wanting.
Today I reconnected some 15 inch PA speakers I have in the workshop and adjusted my tuning and crossovers. I didn't realise I was missing that mid bass punch till I had it again, I have plenty of sub but was missing that 70-180Hz range from my panels and subs alone.
I was thinking I may need to try more exciters or make some better canvas panels to have a bit more punch but my quick experiments have me worried that canvas may not be the answer if I want to have the SPL as well as the bass?
Hi Joska,
Yes, this was the small single-driver panel. There was too much ringing on the panel, so I shoved the foam strips in between the canvas and the frame to damp it a bit.
I put a spacer, a thin dowel, all all around between the canvas and the wooden frame, both to pull the canvas a bit tighter, and to stop the canvas flapping against the frame. The tightness of the canvas will determine the fundamental resonance of the canvas/insert assembly.
No, not too loudly, but just very nicely. But I don't know what "loud" is in a domestic environment... I play live gigs with my PA speakers less than a metre behind me, loudly enough for the drunks to hear me on the other side of the garden. I hardly ever play music at home, so I can't really tell what 'loud' is.
But I think I'd have to use multiple panels like that, with one driver per panel, to get battle-level SPL. Multiple drivers per panel in that small size don't deliver the same bass sound.
You need to play with the stiffness of the material that you glue onto the canvas. Low stiffness with a relatively heavy material will get you good bass (you will have to tune it), and a high stiffness, light material will get you better tops.
Thanks Andre,
I appreciate your reply and it's given me some good ideas to try out when I get some more time.
I appreciate your reply and it's given me some good ideas to try out when I get some more time.
Just came across something that might interest you.
Apparently one can dilute silicone adhesive with acetone sufficiently to enable one to spray it a panel.
Even a pump sprayer will work.
The acetone evaporates leaving a very thin layer of silicone.
Only do this with plenty of ventilation. Outside best.
Could prove interesting.
A very good thought!
I suppose the idea is not to move the panel, as in pistonic movement of a "normal" speaker, but to have enough stiffness of the panel, so only the bending waves carry on the surface. It is not easy to get rid of the pistonic motion, as I understand some here had achieved that one way or another.
I don't believe you need to buy an expensive neodymium magnet, to get the coil to react on the panel, the exciter being such a tiny thing compared to the panel itself. The exciter has restricted the pistonic movement of the coil by a small plate (or something like that), with that plate transferring vibrations to the panel. The interesting matter is that the exciter transfers vibrations to that reverse side of the panel, but we are looking for the front side to give us the sound. We have considered as though the panel has no thickness, so the bending waves happen also on the front, after the vibrations had crossed over to the front.
By using the exciter(s) at the back, we have disregarded the main bending waves happening at the reverse side of the panel. But, placing the exciters on the front would look somewhat ugly, which is a problem. Now, coming back to the coil and the magnet separately, one might consider placing the coil on the front side of the panel, and have the magnet(s) at the back, a few mms away from the panel, so that eventual pistonic motion won't knock the panel on the magnet structure.
There also a question of how to flatten the coil, should it be a round one, or a rectangular one, should the magnet(s) cover the whole of the coil, or part of the coil. Also, should it be a Tesla coil, or normal solenoid, but flattened. If it is a flattened solenoid coil, how should it be flattened, left to right, or into the middle, or outwards -- on how the alternating current (sound signal) react with the flattened coil, in the three ways a rectangular coil can be flattened? Or, should the solenoid coil be flattened as it is, so we might get an ellipse, or an oval?
Lot of questions to consider, but having just the coil on the panel will keep the panel lighter, without the weight of the exciter (or magnets). We can fix the panel on a structure with some sort of a suspension (rubber, silicon). Maybe, we can make the panel much smaller. Maybe, we can cover everything later with a speaker fabric. Would be interesting to see everyone's thoughts on this.
Its quite interesting what you say here. I saw the picture of your panel with the Nidaplast insert. In a way, the stretched artist canvas can be considered as the suspension, instead of rubber or silicon, and the smallest possible gap between the frame and the insert would make the speaker system even stiffer, hoping that canvas won't lose its strength later and expand. The canvas glued to insert would give a roughened surface for the bending waves, which is what most people try to achieve by sandpapering the polystyrene surface.
Very interesting thoughts. Not only for the "normal" DML panel with an exciter, but also for what I was thinking above, with a flattened coil on the panel, and separate magnets(s) as the motor, an assembly.
Anyone here tried to 3d print a panel similar to Nidaplast? I saw Bambu Labs is introducing some Nylon Carbon Fiber filament that may meet the stiffness requirements? Maybe 3d print the honeycomb structure and epoxy between two sheets of "store bought" carbon fiber panels?
I'm in the process of trying my fist set of DMLs... going with the Lowes EPS 2'x2' panels lightly sanded then a coat of Hyde glue then Shellac. Excited to give them a try...
P.S... I found a new use of my Sous Vide machine... warming Hyde glue to the exact recommended temperature.
I'm in the process of trying my fist set of DMLs... going with the Lowes EPS 2'x2' panels lightly sanded then a coat of Hyde glue then Shellac. Excited to give them a try...
P.S... I found a new use of my Sous Vide machine... warming Hyde glue to the exact recommended temperature.
I have done some experiments with printing, and concluded that thermoplastics are not stiff enough. They will give bad HF reproduction and carbon fibre enhanced filaments are not that much stiffer and doesn't have long fibers but more like CF dust. Nylon it relatively flexible, which is one reason it is stronger than PLA which is more brittle, but makes it even worse for DML plates.
The flexibility of thermoplastics also makes it not that great as honeycomb material, and if you are doing carbon fibre skins I would not think it is worth saving on the cost for nomex honeycomb instead.
The flexibility of thermoplastics also makes it not that great as honeycomb material, and if you are doing carbon fibre skins I would not think it is worth saving on the cost for nomex honeycomb instead.
Roger that. I suspect the materials science part of 3d printing will catch up with things like nomex soon enough.
If I hear potential in my panels I'm building now perhaps I'll try what you guys are up to now with the artist canvas? @Andre Bellwood do you like these (the canvas based DML's) better than the ones you did where you hot-glued a poly disk on a panel or were those two different use cases (PA vs Home)? I've been intrigued by that ever since I read it on the DML for PA thread...
If I hear potential in my panels I'm building now perhaps I'll try what you guys are up to now with the artist canvas? @Andre Bellwood do you like these (the canvas based DML's) better than the ones you did where you hot-glued a poly disk on a panel or were those two different use cases (PA vs Home)? I've been intrigued by that ever since I read it on the DML for PA thread...