A Study of DMLs as a Full Range Speaker

Pway.
Basically ,you could say that the heart shaped panels sounded fantastic straight out of the box,boom,boom,:D

You could say My card panels are a composite as I coat them in pva,the 2mm card has two 1mm cards glued together and also coated in pva.

I prefer to have a combination of direct primary sound(the first puls,or ripple from the exciter)and the distributed mode(modal)sound.
The stiffer and harder the panel the more the hf will spread across the panel surface,but if it is too hard it will start to become, unpleasant?
This will swamp the primary first pulse .
I have found that using a more flexible panel, promotes a more even blend of the two.
Using a large very flexible panel will produce a large primary pulse and be quite lossy with little reflections,with very little dml going on.
If you heavily damp a dml panel ,you end up having something more like a bmr.
Cone speakers and bmr are very similar in that they both heavily damp reflections and control resonances.
Which could be the reason why I have not as yet heard a dml that I have liked, yet?
Steve.
 
  • Thank You
Reactions: 1 user
Hi Steve

Pway.
Basically ,you could say that the heart shaped panels sounded fantastic straight out of the box,boom,boom,:D

Did you choose the heart shape deliberately, or have a theory as to why this shape worked well? One thing I can see, thinking about my teardrop shape idea, is that the heart lobes could work like a split 'pseudo-parabolic' reflector, which does not lead the reflections back through the exciter area. I came across heart shape looking for an even distribution of lengths from exciter to edge - (eg polar plot r = theta or r = sqrt(theta), but thought initially that a concavity was probably a bad idea.

Using a large very flexible panel will produce a large primary pulse and be quite lossy with little reflections,with very little dml going on.
If you heavily damp a dml panel ,you end up having something more like a bmr.

How do you characterise your ideal balance of primary vs secondary? Do you have an ideal impulse response in mind?

Paul
 
Pway.
The heart shaped panel was only used because my wife was throwing a heart shaped makeup box away!!
This was very quickly cut up and made into a dml panel.
The idea was to show how quick and easy it is to make a very good sounding panel with very little money.
Obviously as long as you know what you are doing.

Large panels take longer to stop vibrating than small panels,it's the same for cone drivers.
One day I might get round to making a recording of this problem?
There is no reason a very expensive panel is going to sound any better than my cheep panels unless maybe it was designed specifically for that purpose? Which they are not.
Each new panel design has its own set of rules and will react in a different way when vibrating, it's very difficult to predict how it will sound, but you get to know eventually what works and what doesn't.
Mostly?
Steve.
 
Amazing Resonance Experiment! - YouTube

I really like the chaldini patterns. Will a dml make cool paterns like that metal plate?

I could imagine doing the sand pattern then spray painting the backround then removing the sand for a negative.

Or using plaster powder and misting water or printer toner powder and melting with a heatlamp for a positive.

It would be a great converstion piece.

This thread is really long, about what spl should i expect from a foam 18" panel?
 
cheapvega said:
Bernardy, what exciter is that?
It s been 4 Dayton DAEX25FHE-4.

I finished my measurements and was trying to unglue the exciter from the plywood panel.
Got some acetone from the hardware store and kept the glued part in this solution for a few minutes. The superglue was completely unimpressed by this bathing but the ring of the exciter got loose and came off (see pic). The plastic ring should easily be 3d printable which should make exciters fairly easy to reuse even if attached to panels with superglue?

spedge said:
On page 6 there is an explanation of dml impulse responses and why it is ok to have a long tail. http://www.soundright.org.uk/NXTchaos.pdf
Steve.
Interesting read, thank you! First time I get to scratch on the underlying working principles of DMLs. It says:
In fact it actually improves in performance as it is increased in size because the frequency of the fundamental bending resonance is lowered, which not only extends the bass response, but also increases modal density in the mid and high frequencies.
What is the fundamental bending resonance and how can I determine it in a panel?
Model Density means, that with bigger panels, there is a higher number of modes that have overall smaller amplitudes = better balanced frequency response?
 

Attachments

  • 1.jpg
    1.jpg
    237.6 KB · Views: 317
Bernardy.
It might help if you get something like a chisel behind the superglue .
Superglue doesn't like bending much,but the ring is more flexible, so hopefully the two should start to part company?
The easiest way to set off the fundamental bending resonance is probably to just give it a thump with your hand,on large panels such as 6ft , this will set off a boom like a very big drum,this is the one to worry about as every now and then on certain tracks it will get set off by something?

Not sure what he ment by improving the performance in that article, could he have ment you could go lower down before you reached the fundamen resonance?
But I personally would cut off before reaching the resonance frequency ,as you never know when or what will set it off?
Steve.
 
Mpte.
Those vibrating plates always look amazing, but for me they would be more interesting if they offset the vibration point to match say ,the nxt or some other positions ?
This is probably what the nxt software did? At great expense.

If by foam ,you mean xps ,I would expect very loud dbs, but even more so with eps.
This is even with a 10watt exciter !
If you play the du hast recording using headphones.
But play it so that it is getting very uncomfortably loud.
This was the volume I was listening to them and I was 10ft from the panels.
Any louder and I'd have had to put my fingers in my ears.
This was also only using my small card panels!!
As as you can hear,they went loud without any stress.
Steve
 
  • Like
Reactions: 1 user
prime test



ciao come state?
eccomi di nuovo ho dovuto aspettare che gli altoparlanti consigliati arrivassero dall cina, ho provato solo un pò su qualsiasi cosa anche sulla mia scrivania ma fa solo rumore ma suona gracchiante come una cornacchia perchè lo appoggio a mano mano e vorrei evitare di usare il suo desivo, quindi chiedo come posso provare diversi materiali e avere un suono accettabile senza usare un adesivo? mi serve un metodo per non far vibrare senza controllo i materiali e inoltre consigliatemi su quali materiali posso usare per le prime prove
Hello, how are you?
here i am again i had to wait for the recommended speakers to arrive from china i just tried a little bit on anything even on my desk but it just makes noise but it sounds croaking like a crow because i support it by hand and i would avoid using its desivo , so I ask how can I try different materials and have an acceptable sound without using an adhesive? I need a method to not make the materials vibrate without control and also advise me on which materials I can use for the first tests
 
Will a defined chaldini pattern form on horizontal 1/2" xps panel with sand like it would with a metal plate?

I was thinking of spray painting and the areas with the sand will not get paint but the backround will. Alternatively taking a picture printing it and cuting holes to make a stencil to paint with.

Maybe using a plaster powder then misting it with water and leaving it on the panel and not painting.
 
What is the fundamental bending resonance and how can I determine it in a panel?

You can estimate it with calculations. If you can open the paper at this link, look at equations 2 and 23.

(PDF) Development of panel loudspeaker system: Design, evaluation and enhancement

Another similar equation is presented here:

Eigenfrequency Analysis

Scroll down to the section on plates. For the fundamental bending resonance use m=n=1.

The graphics showing the various mode shapes are also quite helpful if you have not seen depictions like these before and are wondering what "mode shapes" (i.e. "eigenmodes") are.

Model Density means, that with bigger panels, there is a higher number of modes that have overall smaller amplitudes = better balanced frequency response?

Yes, that is basically correct, or at least, that's the theory.
But I think it's more accurate to say that bigger panels can potentially give flatter response toward the low frequency end (but may not). I suspect that for stiff, light panels that have a sufficiently low coincidence frequency, bigger panels should give flatter response, all else equal. But, for less stiff and heavier panels, bigger is less likely to help. Bigger won't likely hurt, but also won't likely help.

Eric
 
thank you
this is my first experiment with this system, at the moment I can't say I'm happy because the exciter even at half the power it should support croaks (I don't know how to translate the hum of the empty cone without loading it on the panel) in practice buzzes a lot and then the coil heats up. As for the vibration and consequent hum of the panel, surely the loom method could work. I bought these exciters because I am interested in high powers it is therefore obvious that maybe I have to use a very rigid frame.Any advice to use powerful exciters and which material safe for an acceptable result is very welcome and above all to always avoid using the tape in this experimental phase speaker sticker.Thank you
 
pixel1,

A few possible aluminum skin on aluminum honeycomb panel suppliers (in the U.S.).

They often use similar panels for:

  • highway signs and
  • art restoration

You may find sources in those worlds.

To my mind and experience, one can get aluminum panels like this with an internal/internal edge framing the same thickness as the honeycomb and covered by the aluminum skins. I've seen such framing made from wood or aluminum.

I've mentioned this before, even 3/8" (9.525 mm) thick panels have extraordinary rigidity to weight ratio.
I think only titanium would exceed aluminum in this.

They don't come cheap, but they should handle high power and volume easily (at least mine see to).
 
spedge; said:
It might help if you get something like a chisel behind the superglue. Superglue doesn't like bending much,but the ring is more flexible, so hopefully the two should start to part company?
Thank you for the tip. I ll give it a try!

spedge; said:
Not sure what he meant by improving the performance in that article, could he have meant you could go lower down before you reached the fundament resonance?
Are you specifically referring to the following sentence on page 7?
This works in our favor, making a well-designed NXT panel sufficiently modal for use as a loudspeaker by a frequency only twice that of the panel's fundamental bending resonance (Figure 6).
I am struggling with this (too). Can someone unpack this for me?
2* f0 = sufficient modal density; Doesn't make any sense to me...

Veleric; said:
You can estimate it with calculations. The graphics showing the various mode shapes are also quite helpful if you have not seen depictions like these before and are wondering what "mode shapes" (i.e. "eigenmodes") are.
Thank you for the links and the explanation!

  • So, fundamental bending resonance = natural frequency = f0?
  • What do the indices m & n stand for?
  • What integer do I use for m & n if I want to calculate the 2nd, 3rd, 4th etc. eigenmode?