Hi Eric.
This begs the question, when does a DML panel become a bendingwave panel or vice verse?
Goebel reduces DML activity on his panels to a minimum(whatever percentage that is?) Similar to a Walsh driver.
I let my panels produce as much DML action as possible .
Yourself and others clamp or damp the panels, which I regard as moving towards bendingwave action?
So where is the line drawn , or does it really matter?
Square panels
Not all square panels are bad.
The picture is of a 2ft square proplex panel that JohnnoG sent me, at 1m.
I have done absolutely nothing to this panel except glue on the exciter .
What can I do to improve this response and sound?
I personally would not use a square or round panel, but sometimes they do work, remarkably well.
Is this because proplex is more flexible in one direction than the other?
Panels are fascinating, but can also be very annoying 🤔🤪
Steve.
Attachments
Steve,
I agree 100%. The Proplex is a great example of a square that isn't really square. And, yes, I would say the reason the Proplex panel gives a good response as a square is because the properties are so different in the two directions. I just did some quick modelling and found that a square plate with a modulus 20x higher in one direction than the other has a very similar distribution of natural frequencies as a plate with a 2.5:1 aspect ratio and equal modulus in both directions.
So a square may very well be fine if the properties are different in the two directions, or maybe even if just the boundary conditions vary.
Eric
Steve,
The question of how much damping is optimum is indeed a very interesting question. The papers I've read and my own experience suggests that some damping is good to help smooth the frequency response, perhaps boost the efficiency, and eliminate excessive ringing. But what I really wonder is if there is some level of damping (via ABH means or other) at which the modal behavior is virtually eliminated, and hence (perhaps?) the frequency response becomes much, much flatter. And even (perhaps) that at some level of damping the other aspects of panel speakers that are driven by their modal nature virtually disappear. That is, would panel aspect ratio and exciter position become largely irrelevant at some level of damping? And if so, would there be some other bad consequence? So far, in my own attempts to add damping, I have never been able to introduce enough damping to even come close to eliminating modal behavior, so I don't yet really have a good idea of what would happen if I could, but I'd like to find out!
I do disagree with you somewhat on the meaning of "bendingwave" speaker. I realize that when you say "bendingwave" you mean a damped panel speaker. But to me, any speaker that produces sound primarily via flexural waves is a "bendingwave" speaker, including a largely undamped DML. I think even Goebel himself would agree with that definition. According to him, the addition of (his) damping to a bendingwave speaker is what makes it a Goebel bendingwave speaker (emphasis on Goebel), not what makes it a Bendingwave speaker. At least that's how I read his literature.
Eric
aagas,
That's an awesome looking response, and an attractive, simple design. A little surprised there haven't been more comments. Can you share more details? I presume it's a 3 ply birch, yes? Is it mounted only at (or near rather) the four corners? How? What is the size of the panels? How far away was the mic from the panel? I'm surprised about the great low frequency response in particular, I admit.
Eric
Hello Chdsl
Yes I had a look to web site. Thank you for the link with the translation, it works fine. I agree that a point transfer might be a good idea but not not for the same reason. The site focuses on the relation with the Edison invention but I haven't find explanation to why the DML is cut at 2kHz. I wonder what is the result of the coupling with the tweeter. The DML radiates from both site, the tweeter in front with an increasing directivity.
Nevertheless, the point transfer remains in my test wish list for 2 reasons : it makes sense to think it is in favor of the bending wave (no more separation between the inner or outer area of the voice coil), the point coupler will increase the voice coil to panel distance helping to reduce the exciter noise... but for the moment I don't see (this just says my ignorance?) advantage in the search of a smooth FR.
Christian
Christian.
This is my hdn8 exciter attached to a plasterboard panel.
If I attached this to my rigidly mounted ply panels ,I would basically have the same panel as the jazzmans panel.
Connected with only the screw thread.
Except this unit will go up to about 15k.
So I don't see anything new ?
I'm not sure what panel material he is using, is this special?
Steve.
This is my hdn8 exciter attached to a plasterboard panel.
If I attached this to my rigidly mounted ply panels ,I would basically have the same panel as the jazzmans panel.
Connected with only the screw thread.
Except this unit will go up to about 15k.
So I don't see anything new ?
I'm not sure what panel material he is using, is this special?
Steve.
Attachments
Hello Steve,
Sorry Steve but I don't understand your point. The first target was to check the phase of the rear wave in regard to the front wave for an OB with a standard loudspeaker and a DML.. it was a canvas I had at hands, both having similar overall dimensions The conclusion is both show waves out of phase. The weak point pointed by Burnt is : it is a canvas not a more standard DML. Do you have measurements covering this case?
The second part of the document is about reflections, here, coming on it seems necessary.
Christian
Does a square proplex panel is really square? It is square in the geometry point of view but for waves as the stiffness is not the same in both directions... The current FEM simulations shown here are in my understanding with an homogeneous material.
PS : I should have read Eric's reply first... same content.
Eric.
It seems very clear to me why he calls his panel a bendingwave panel.
He is trying to prevent any reflections from the edge of the panel returning to prevent any DML action being set up.
Hence bendingwave only .
How much he succeeds is another question.
If the ABH does as it says and totally kills any reflections, there will be no DML action at all.
It would be a bendingwave panel.
Steve.
It seems very clear to me why he calls his panel a bendingwave panel.
He is trying to prevent any reflections from the edge of the panel returning to prevent any DML action being set up.
Hence bendingwave only .
How much he succeeds is another question.
If the ABH does as it says and totally kills any reflections, there will be no DML action at all.
It would be a bendingwave panel.
Steve.
Hello Steve,
First time I have a look to this HDN8 exciter. Yes, with its screw thread it is close of the idea of the jazzmans.
The HDN8 as characteristics that would have lead me to reject it as exciter choice : very high mass (38g in the spec, really? incredible!), a high inductance (0.86mH). The good point is the high BL.
Do you have test results with it with different materials? I would be curious which high frequency it reaches with PS, plywood.
About the jazzmans, I haven't found information about the material. do you?
Christian
Thanks Eric.
Some more detail...
Panels:
3mm birch plywood (3 ply).
Raw birch plywood (any thoughts on finishing or not finishing appreciated).
33 3/4" x 15 5/8" arrived at entirely to match with the TV.
I had to dampen (actually wet them) the panels, put them between towels, and weight them down to get the worst of the warp out of the panels.
Even clipped into the frame the panels have a tiny bit of bow to them. I've considered trying more clips.
Sourced from MakerStock.com. Delivered in 2 days. Cut to size (perfect cuts). $37.90 (plus shipping) for the pair.
Frame: The Abstracta frame leaves about 1/8" between the panel and frame on all sides.
Exciters placement: 2/5 from top and side.
Mounting
I currently use 2 Abstracta Shelf Support Clips (see attached image) near each corner. One from the front and one from the back so they secure the panel. I have small felt dots (about 1/4" in diameter) on each of the clips, which then press against the panel to hold it. Not the most elegant solution, but I had some of them sitting around my workbench. I have ordered 8 of the Abstracta Panel Clip 1/4″ (again see attached image) to see if I can work out a simpler mounting. The 1/4 inch clip vs the 3mm (1/8") panel should let me squeeze in some felt or foam padding.
Of note, a significant amount of the weight of the panels rests on additional felt pads placed orthogonally to the front|back clips.
Placement when measured:
Approximately 16" from the untreated brick wall.
5' 8" between the DMLs (with a LG flat screen TV between them).
Room (front to back) 22' 9"
Microphone place 13' 9" from the wall behind the speakers.
I think the size of the room may help the low frequency response. I didn't really do anything deliberately with these other than try to make them as simple as possible.Pretty astonished by how good the birch plywood sounds. They have that DML "...you are there..." musicality to them.
Following the recent posts, has me wondering about whether I could|should:
- clip the panels at just two points on the side to emulate the cantilevered ideas and/or
- radius the corners.
I've also started to give thought to fabricating some simple spine to better support the exciters long term. These Abstracta frames give one lots of options.
A work in progress, but I find myself just listening to stuff. They also do well on movies like Dune or Kingdom of Heaven with interesting soundtracks.
I must admit that they don't sound great with TV shows, but then I haven't found anything that does.
Maybe stick a miniDSP into the audio chain and try Dirac Live?
Any thoughts out there appreciated.
- Andreas
Attachments
He's keeping it a secret.
https://day--pla-net.translate.goog...tl=en&_x_tr_hl=en&_x_tr_pto=sc&_x_tr_sch=http
Was too long to modified the previous post :
Here is the beginning of a track : inventor Toru Maruyama, JP2015136076A, JP2016063283A, JP2018125823A, JP2018121178A. No clue about the membrane for now.
Some translation by google patent
Here is the beginning of a track : inventor Toru Maruyama, JP2015136076A, JP2016063283A, JP2018125823A, JP2018121178A. No clue about the membrane for now.
Some translation by google patent
Thank you for the link to the patent. Will use Google translate to try to fathom it.
I tried an experiment where top half of a long stadium shape (300 x 1200) was simply supported, and the other half free.
This is a non-starter judging from the mode shapes and natural frequencies. Most areas of positive displacement are balanced by about equal areas of negative displacement. You lose the fundamental. And the LF modal frequencies quickly exceed 100 Hz anyway. (16, 71, 98 Hz)
One way to think of this is that any displacement of the supported half will bend the plate in the X direction, stiffening it in the Y direction and removing the possibility of low frequency modes as in the cantilever.
A few pretty mode shapes though...
This is a non-starter judging from the mode shapes and natural frequencies. Most areas of positive displacement are balanced by about equal areas of negative displacement. You lose the fundamental. And the LF modal frequencies quickly exceed 100 Hz anyway. (16, 71, 98 Hz)
One way to think of this is that any displacement of the supported half will bend the plate in the X direction, stiffening it in the Y direction and removing the possibility of low frequency modes as in the cantilever.
A few pretty mode shapes though...
I understand that you interpret "bending wave" as meaning that reflections are inhibited to limit modal (DML) behavior.
But what I'm saying is that is not actually what bending wave really means, at least not historically. For example, below is a snip from one of the Azima patents. It's clear from this quote that he considers DML to be a type, or subset of bending wave speaker, rather than something distinct from a bending wave speaker. He even considers the Heron speaker (image below) to be a bending wave speaker, and it clearly has no means to limit edge reflections, as it is essentially a freely hanging panel. If Azima considers a DML to be a type of bending wave speaker, then I'm not going to disagree.
The snip from the Tectonic website below also indicates that they consider their DMLs to be bending wave speakers.
I contend that there are two types of bending wave speakers. One type that has little or no edge damping (DML) and another type that has substantial edge damping.
Eric
Eric.
I have never stated that "bendingwave" means that reflections are inhibited to limit modal (DML)behavior .
I have stated in the past on this site and nxt RUBBISH, that as far as I know there are three stages of a A DML panel.
1. The pistonic primary drive of the exciter.
2. The bendingwave action is produced by this pistonic action.
3. The DML is produced by boundary conditions reflecting bending waves back into the panel.
If we remove the DML from the panel, what do we call the panel ?
A. Pistonic?
B. Bendingwave?
C. Other
Steve.
I have never stated that "bendingwave" means that reflections are inhibited to limit modal (DML)behavior .
I have stated in the past on this site and nxt RUBBISH, that as far as I know there are three stages of a A DML panel.
1. The pistonic primary drive of the exciter.
2. The bendingwave action is produced by this pistonic action.
3. The DML is produced by boundary conditions reflecting bending waves back into the panel.
If we remove the DML from the panel, what do we call the panel ?
A. Pistonic?
B. Bendingwave?
C. Other
Steve.
The hdn8 is a terrible exciter , and would never recommend it, except for on wall studs, which worked remarkably well.
If I had a TV on a stud wall ,I would definitely use this ,or hopefully a better exciter shaker? To turn the whole wall into a DML speaker, it would sound awesome.
I have previously shown measurements of this unit on various materials including the wall ,also with recordings , somewhere on this forum.
Steve.