- High coincidence frequency (fk fo).
In this case, the phase velocity of the bending wave cB must be very low. This is the case for materials with a low E - h/ρ ratio, i.e. for
very flexible and heavy materials. The bending wave near-field at
the excitation point produces an almost omnidirectional sound radiation, as do the
edge effects. However, the efficiency is low.
- Coincident frequency in the transmission range (fu < fk < fo)
Below fk, the transducer operates as a piston with "acoustically short-circuited" bending wave radiation, above fk as a bending wave oscillator with
The disadvantages are the increasing efficiency above fk and the occurring modes.
- Low coincident frequency (fk < fu)
The transducer works more and more as an ideal piston with increasing frequency and simultaneously increasing high phase velocity cB. As long as
the bending wavelength λB is substantially larger than the effective length of the plate or shell, no standing wave can build up by reflection
and the frequency and phase response is balanced. Only when at higher
frequencies, the length of the plate or shell comes into the order of magnitude of the
wavelength, standing waves (modes) are formed.
These modes and the frequency-dependent radiation angle are problematic.
The dimensioning here must be exactly different from the first case: high stiffness (high Young's modulus, high related bending stiffness due to curvature of the
shell) and low specific weight.
An optimization (reduction of resonance influences and thus smoothing of the
According to [German Physiks 1994], optimization (reduction of resonance influences and thus smoothing of the frequency response) can be achieved by selective influencing of the reflections:
8.1. SOUND RADIATION OF FINITE PLATES 95
(a) Shortly after an excitation (b) In steady-state condition
Figure 8.7: Vibration distribution of an NXT transducer or DML panel (from
[Altrichter 1999])
- Frequency-independent termination of the plate or shell edge corresponding to dispersion by frequency-dependent absorbing dampers. This is
not feasible in practice.
- Frequency-dependent mechanical resonance absorbers in or on the plate or shell.
shell. However, this approach increases the diaphragm mass.
above the audible range.
Translated with www.DeepL.com/Translator (free version)
In this case, the phase velocity of the bending wave cB must be very low. This is the case for materials with a low E - h/ρ ratio, i.e. for
very flexible and heavy materials. The bending wave near-field at
the excitation point produces an almost omnidirectional sound radiation, as do the
edge effects. However, the efficiency is low.
- Coincident frequency in the transmission range (fu < fk < fo)
Below fk, the transducer operates as a piston with "acoustically short-circuited" bending wave radiation, above fk as a bending wave oscillator with
The disadvantages are the increasing efficiency above fk and the occurring modes.
- Low coincident frequency (fk < fu)
The transducer works more and more as an ideal piston with increasing frequency and simultaneously increasing high phase velocity cB. As long as
the bending wavelength λB is substantially larger than the effective length of the plate or shell, no standing wave can build up by reflection
and the frequency and phase response is balanced. Only when at higher
frequencies, the length of the plate or shell comes into the order of magnitude of the
wavelength, standing waves (modes) are formed.
These modes and the frequency-dependent radiation angle are problematic.
The dimensioning here must be exactly different from the first case: high stiffness (high Young's modulus, high related bending stiffness due to curvature of the
shell) and low specific weight.
An optimization (reduction of resonance influences and thus smoothing of the
According to [German Physiks 1994], optimization (reduction of resonance influences and thus smoothing of the frequency response) can be achieved by selective influencing of the reflections:
8.1. SOUND RADIATION OF FINITE PLATES 95
(a) Shortly after an excitation (b) In steady-state condition
Figure 8.7: Vibration distribution of an NXT transducer or DML panel (from
[Altrichter 1999])
- Frequency-independent termination of the plate or shell edge corresponding to dispersion by frequency-dependent absorbing dampers. This is
not feasible in practice.
- Frequency-dependent mechanical resonance absorbers in or on the plate or shell.
shell. However, this approach increases the diaphragm mass.
- Mechanically damped excitation of the bending wave.
- Electrical resonance absorbers for the resonance frequencies.
- Raising the first resonance to a frequency above fo by material and shape selection of the plate or shell. This would be the optimal solution if
above the audible range.
Translated with www.DeepL.com/Translator (free version)
If you think the patents are basically the same , can you please just say so.
I am amazed at the amount of information that you have found on the Web old and new, I've looked for years and found very little.
Maybe you could re-find the patent on panel shapes that I lost and have tried to find again, but, without luck.
Steve.
I am amazed at the amount of information that you have found on the Web old and new, I've looked for years and found very little.
Maybe you could re-find the patent on panel shapes that I lost and have tried to find again, but, without luck.
Steve.
peril,
That is a translation of what exactly?
I get that fk is coincidence frequency, but I'm not sure what fo, fu and fD are. Any idea?
Eric
Steve,
As I read it, the claim is that anywhere along the line EB is good, but the X spot, where EB/EX=1.62 is "best".
Funny thing is, they claim this is substantially different from the NXT patents, but if you do the geometry, it turns out that their best (X) spot is at (0.38, 0.41) which is pretty darn close to the NXT position of (5/13, 3/7), i.e, (0.38, 0.43).
That said, I'm pretty convinced that the idea that there even exists a good simple general rule for exciter positioning is incorrect. The best evidence is in this article:
https://www.researchgate.net/public...iated_Sound_Power_of_a_Flat_Panel_Loudspeaker
It's all based on modelling, not actual measurements, but they find that the "best" position depends strongly on the size of the panel, even for a fixed aspect ratio (1.25 in the case of the paper). I think it's reasonable to suspect that the ideal position also depends just as much or more on the aspect ratio of the panel, anisotropy of the panel, and boundary conditions of the panel.
If you don't want to make any effort, the 2/5 is rarely a bad place, I suspect, but this or any other simple rule is likely to be the best place, I think.
Eric
Eric.
As with the proplex panels, a square panel works very well if it can flex differently in different directions.
Even with the exciter in the centre,and nothing else done to it, it still measures better than most panels I have tested, apart from the oddities from the flutes.
I intend to look at all (eventually) my panels , large and small, to see what happens to the out of phase response (out of phase with each panel ,that is), it is very interesting.
Steve.
As with the proplex panels, a square panel works very well if it can flex differently in different directions.
Even with the exciter in the centre,and nothing else done to it, it still measures better than most panels I have tested, apart from the oddities from the flutes.
I intend to look at all (eventually) my panels , large and small, to see what happens to the out of phase response (out of phase with each panel ,that is), it is very interesting.
Steve.
It is the matter of reading in the given language. When you read few documents comparing them, you get the feeling, even if the sentences are somewhat different. It is the style of writing, and the way the information is given out. When you read a document/a book, it is like the person is talking to you.
You have to place a task, and start searching on that task, not branching out. Found what I looked for, read the patents, posted them here for all to read. You find one patent, in this case, and that somehow points to the next one. It is not to branch out and go on the side roads, keep at it, until you find all.
Now, that I've found them, I've lost interest, especially because of that Karavashkin "thesis."
If anyone wants, compare Herger patent and Petrushevski+5 patent on the exciter.
As I said here, #8,264, I am going to read henry Azima's works for now.
Last edited:
Hi Eric,
I agree with you on the need to experiment with exciter positions rather than rely entirely on simulation. Differing panel proportions and material anisotropy can and usually do deliver quite different locations that sound and measure best. I rely on the manual test method these days although you can rely on panel symmetry to simplify this approach to some extent. Test one half or one quarter of a panel and generally the same position is reflected about the midpoint axis which saves time.
Burnt
- The diaphragm of a Distributed-Mode Loudspeaker (DML) vibrates in a complex pattern over its entire surface. Close to the diaphragm the sound field created by this complex pattern of vibration is complex also, but a short distance away it takes on the far-field characteristics of the DML radiation. This is close to the directivity of a true point source - i.e., approaching omnidirectionality - even when the diaphragm is quite large relative to the radiated wavelength.
- The radiated sound is dispersed evenly in all directions. Diffuse radiation of high order becomes omnidirectional in the far field.
Simulation becomes quite hard, in the far field. No one really sits at 1 meter in front of one speaker in a normal usage, more or less 2-3m away, and not exactly in the centre.
Last edited:
Hello Steve,
I built countless panels with an incredible number of different materials and composites, but I always came across the same problems with each new attempt: the stereo image and the soundstage wasn't precise, and, above all, each type of material "colored" the sound in a particular way. The type of exciter, its position on the panel, the way it was connected to its support, the size of these panels… All parameters had a consequence on the resulting sound !
As I said in a previous post, I have done a lot of recording and mastering, so I had the voices and the sound of the "live” instruments in my ears and my head, by the way I had a "real" reference to judge the quality of reproduction, not by using commercial recordings for which we are neither at the origin of the recording, nor in the studio for its mastering, and thus without any reference as to the original sound…
Not a single panel reached the level of "high fidelity” of the several studio (i. e. professional) monitors I had the chance to work with. So, I tried to understand the reasons for these differences, and I discover that the physical characteristics and in particular the radiation pattern of these transducers isn’t adapted to deliver the awaited soundstage.
When it comes to music reproduction, there are two large populations that will never agree on anything, one simply seeks to reproduce the recorded sound as faithfully as possible (remember the QUAD adage: "For the closest approach to the original sound!"...), this is the raison d'être of HiFi, the other seeks to please itself by constantly "improving" its system without worrying about whether or not the sound reproduced is the one that was recorded… There is no "good" team, each one like one or the other approach, and there isn't any truth ! Reproducing sounds is feasible by a very large number of means, but not all of them lead to High Fidelity.
But, as suggested by some workers in that field, DSPs could perhaps be of some help to improve the reproduction of DMLs, since it remains any hope, I still keep an eye on this present thread...
Last edited:
Ondesx.
It is strange how I have the totally opposite opinion of the sound from a DML panel.
I find the soundstage to be far superior to ordinary speakers and I include electrostatic in that.
You move your head just a little and you loose everything with ordinary speakers.
Realism is the main reason I like them.
Colouration can be a problem and every effort should be made to reduce this.
Can you describe or show pictures of the panels you have made, to give an idea of how things could have gone so wrong?
Steve.
It is strange how I have the totally opposite opinion of the sound from a DML panel.
I find the soundstage to be far superior to ordinary speakers and I include electrostatic in that.
You move your head just a little and you loose everything with ordinary speakers.
Realism is the main reason I like them.
Colouration can be a problem and every effort should be made to reduce this.
Can you describe or show pictures of the panels you have made, to give an idea of how things could have gone so wrong?
Steve.
No, I'm sorry Steve, it's been in the garbage for a long time now, especially since I moved from a big house to an apartment 3 years ago!
The issue here is that everyone relies on their own judgment to evaluate what they hear... Unfortunately, the human ear is not very... reliable! It is even sometimes favorably impressed by certain distortions... The only thing you can say is that "you prefer the sound of the panels and their image". So you belong to the second population of audiophiles, those who want to enjoy listening to reproduced sounds but aren't concerned about fidelity of this reproduction.
The only thing you can do to really "compare" panels and monitors would be to record voices and instruments with subjects you know well in different acoustics and then, with a very few processing of these recordings, listen to them on the different transducers: you will very quickly understand what I am trying to explain about the soundstage and the "coloring" of panels. But even after that comparison, you may still prefer one to the other !
Anyway, I'll say it again, building transducers, here DMLs, and liking the sound they deliver is perfectly understandable. But, if these transducers are not widely used by professionals, who "make" the music that everyone listens to, there are reasons, and these reasons are the ones I explained.
Chdsl.
All materials have colourations even normal speakers suffer from this problem.
Everyone seems to have different methods to deal with it.
Steve.
All materials have colourations even normal speakers suffer from this problem.
Everyone seems to have different methods to deal with it.
Steve.
That's not what I asked. I'll rephrase it. Can you explain what you mean by colouration? Especially in the DM panel?