In the paper about B&W pointed just before there is :
In this short text:
I found "coupe vent" means "stop air leakage" as translation for "draught excluder" so weather strip tape. Correct?
Is it the origin of the use of weather strip in DML DIY?
I understand now that I was more focusing on the spring effect of surrounding foam than on the lossy characteristic.
Among the different materials of such "weather strip", is there already some of preferences?
Does somebody as an idea of the deviation of the batch in the original B&W experiment?
Christian
Browsing a bit on those words, I found : Bowers & Wilkins Blog Industry firsts - Fixed Suspension Transducer
In this short text:
After there is the funny beginning of the story where they bought materials at the local hardware store... some draught excluder where it appears after the batch had some defect that made it worked.
I found "coupe vent" means "stop air leakage" as translation for "draught excluder" so weather strip tape. Correct?
Is it the origin of the use of weather strip in DML DIY?
I understand now that I was more focusing on the spring effect of surrounding foam than on the lossy characteristic.
Among the different materials of such "weather strip", is there already some of preferences?
Does somebody as an idea of the deviation of the batch in the original B&W experiment?
Christian
Additional remaks/questions (thinking loud is the right expression I think)
PS
* Seems also it is the right time to let enter in the game an other possibility of adjust the impedance which is the thickness change.
This thought leads me also to consider the slot in at membrane edges as a mean to modify the plate impedance.
Christian
- A full impedance adaption leads to no or at least very low reflections so no modes. Is this situation desirable?
- Each membrane material has its own impedance so the surrounding should be adapted to it. In other words, to take opposite materials, plywood and EPS shouldn't have the same surrounding material or at least geometry.
- For a given material, does a dimension change of the membrane leads to a different surrounding material?
PS
* Seems also it is the right time to let enter in the game an other possibility of adjust the impedance which is the thickness change.
This thought leads me also to consider the slot in at membrane edges as a mean to modify the plate impedance.
Christian
Last edited:
Christian,
I don't recall seeing that in the paper. What part of the paper are you referring to? My undestanding is that the propogation in a beam and plate are quite similar.
Eric
That's my question also. I would like to find out. And interesting consequence would also be that the shape of the panel would be far less important. Perhaps virtually insignificant.
That's what I would think.
Also what I would think.
Do you mean like the Gobel plate?
Eric
From my reading I also note that frequency has a role and may require more than one ‘treatment,
On whether or not edge absorption is required I think the evidence from unsupported free hanging panels and edge damped panels is that both work, but possibly in different mode of operation. One mode is the ‘classic’ DML chaotic action where edge reflection is desirable as it increases the modes available and the edge terminated ‘ripples on a pond’ mode of operation. Gödel are an example of one extreme implementation of the latter.
In experimenting with both I have found either gives a very acceptable performance although I admit I have not measured my free hanging panels as throughly as my edge damped panels so there may be some colouration in the free hanging panels that I have not noticed. The edge damped panels I built for my brother gave a remarkably flat FR in his room, so flat that I opted not to EQ the response.
My current bet is that damping is beneficial
Burnt
On whether or not edge absorption is required I think the evidence from unsupported free hanging panels and edge damped panels is that both work, but possibly in different mode of operation. One mode is the ‘classic’ DML chaotic action where edge reflection is desirable as it increases the modes available and the edge terminated ‘ripples on a pond’ mode of operation. Gödel are an example of one extreme implementation of the latter.
In experimenting with both I have found either gives a very acceptable performance although I admit I have not measured my free hanging panels as throughly as my edge damped panels so there may be some colouration in the free hanging panels that I have not noticed. The edge damped panels I built for my brother gave a remarkably flat FR in his room, so flat that I opted not to EQ the response.
My current bet is that damping is beneficial
Burnt
It’s the question I think.
I do not know if you have seen this paper. It includes a section on 5.3 Simulating an Infinite Plate - Damping page 46. The conclusion is more or less that, if it’s no reflection then it’s not really a DML. Maybe the optimum is with some reflection?
The paper can be found at https://publications.lib.chalmers.se/records/fulltext/61709.pdf
Thomas
Attachments
Eric, in M Hambric's paper https://www.hambricacoustics.com/In19_inf_panel.pdf
I am not sure of the consequences. Below an extract near equation 3
For beam and plate, cb is proportional to sqrt(w). See after eq 4 and eq 6.
But for a beam, v/F being the mobility (inverse of the mechanical impedance), if you substitute cb in eq 3, the mobility is proportional to 1/sqrt(w) where for a plate, the mobility is not frequency dependant (eq 7)
Hello Burnt
A free hanging panel is clearly in the situation of no impedance adaptation or damping at the edge.
For a panel with a suspension even all around, it is no so clear for me. Let me explain my thought...
One outcome of our exchange here is to damp or adapt the impedance, a lossy suspension might do the job.
But are all the suspensions lossy?
With the foam I have used for my plywood panel (open cell foam), I have doubt. Seems for me now more a spring than a damper.
A "simple" test of the tape is perhaps simply to pinch it, release it suddenly and see how long it needs to come back to its original shape. A pure spring will do it instantly while a damper needs time; Correct?
I am in the opinion i made a wrong choice of foam.
An electrical impedance measurement should give an idea having a look to the "spike".
I have the idea to go back in the main thread to make a kind of survey of the suspension tapes used. It will be a bit long...
In the mean time, could you remind us Burnt what you used as suspension for your panels?
Christian
PS : I made several tests, not finished panels, with different shapes, surroundings and have been up to now (I mean up to those current posts) disappointed by the poor smoothness of their FR so I agree on the hypothesis an amount of peripheral damping is probably needed. Up to which level is the question. Is there a point where the DML qualities vanished. Among the qualities : the very low directivity, the uncorralated IR in different directions, for a PA application, I would had the low feedback sensitivity. What about also the efficiency?
Last edited:
Hi Christian
Your description of damping is correct, just like the damper in a cars suspension the damping medium should resist the initial impulse and subsequent oscillations at least reducing them in amplitude
On my panels I have used strips of material of two types.
One is a soft dense foam. You can depress a sample with your finger and it returns slowly to its original shape ( slowly is very subjective so I will see if I have a sample and test it for you )
The second material is an elastomeric material used to damp vibration from hard drives. It is very dense, is not foamed and has a soft fleshy feel. It also distorts under one’s finger when pressure is applies and has a ‘long’ recover time.
It would be good to collect some metrics e.g.
Density
Plasticity/ deformation amount under a set load
Recovery time.
Strain ( some materials - e.g. blue tack or silicon jell may not return )
Burnt
Your description of damping is correct, just like the damper in a cars suspension the damping medium should resist the initial impulse and subsequent oscillations at least reducing them in amplitude
On my panels I have used strips of material of two types.
One is a soft dense foam. You can depress a sample with your finger and it returns slowly to its original shape ( slowly is very subjective so I will see if I have a sample and test it for you )
The second material is an elastomeric material used to damp vibration from hard drives. It is very dense, is not foamed and has a soft fleshy feel. It also distorts under one’s finger when pressure is applies and has a ‘long’ recover time.
It would be good to collect some metrics e.g.
Density
Plasticity/ deformation amount under a set load
Recovery time.
Strain ( some materials - e.g. blue tack or silicon jell may not return )
Burnt
Last edited:
Hello LeifB60
The need of break-in is not at the top of the topics. I don't know if there is a need. The phenomena might be masked by the fact a pair of exciters go from one set of panels to a new one...
About rounded panel, do you mean a disk? You should find some tests about that in the main thread but I think this shape has not convinced for now due to its very high symmetry.
If you mean rounded the corner of a rectangular shape, it is more common (I did it for my plywood panel) even if it takes more from the oral tradition than something scientific; the corners being supposed of high displacement If you have seen the last panels from Veleric or Burnt or others, no rounded corner.
Christian
Additional thought...
As the idea of having a true damper at the edges seems to make sense... Let's push it a bit.
A damper as you described Burnt is typically a device based on the viscosity ie a piston with some leakages moving into a cylinder full of oil or of a gas. The damper has no rest position or in other words all the position are possible rest positions.
To have unique firm rest position, some spring effect is needed or some stopper. FR = ressort de rappel; EN = spring return? counter spring?
Do you have followed my thinking?
Lets imagine a plate with an almost true damper along its edges, how to insure a rest position? By adding 4 stoppers!? For the picture see Veleric #8874
Drift of my imagination or reality?
Christian
As the idea of having a true damper at the edges seems to make sense... Let's push it a bit.
A damper as you described Burnt is typically a device based on the viscosity ie a piston with some leakages moving into a cylinder full of oil or of a gas. The damper has no rest position or in other words all the position are possible rest positions.
To have unique firm rest position, some spring effect is needed or some stopper. FR = ressort de rappel; EN = spring return? counter spring?
Do you have followed my thinking?
Lets imagine a plate with an almost true damper along its edges, how to insure a rest position? By adding 4 stoppers!? For the picture see Veleric #8874
Drift of my imagination or reality?
Christian
@BurntCoil
Burnt
The second seems to be Sorbothane no? What is the first one?
For the metrics, we should use the metrics from the sorbothane pages pointed by Lenta. If I remember there is a mechanical impedance or mobility. To be checked.
Acoustic guitar use a lot of different brace technique on very thin wood panels. Granted, some of the function of that is to hold the string tension, but it also effects the way the top vibrates. One Trevor Gore, the guy who "wrote the book" on acoustic guitar empirical analysis, said that one type of guitar you want reflection at the boundary, another type you want the wave to keep going around the sides and into the back. It just makes me wonder if those guys know a little bit about this impedance match at a plate boundary, in terms of what to do in the context of construction.
Thank you.
I just started with exciters. Have watched some videos on youtube and it piqued my curiosity. There were several types of different panels of different materials to choose from. Have you figured out which material is best to use?
Started with cardboard.
I was surprised by the result. The small exciter can handle a large area. Maybe not as good as a driver in a box.
I have read all the posts in this thread and you seem to want to solve edge reflections or maybe something else that is bothering.
It's great that you want to share what you come up with.
I'm just a happy amateur who wants to try this thing with exciters.
@LeifB60
From my side, I started with DML something like 3 years ago and the result was a good surprised. We are in the same category of "happy amateur". We try to push the topic with our different backgrounds, knowledge to get the best of the panels.
Have you had a look to the main thread A Study of DMLs as a Full Range Speaker ? Very long, heavy, difficult to read but a great area of exchange. This one here is more specialized.
To help you in finding topic, you can use the history file I posted. See under my signature on the left of the post and follow the link to github. In the history file there are links of posts of realizations or topic on interest (of course of my interest!).
Hope it will help you to start but do not hesitate too much... Plywood (poplar plywood) is a good entry material in my opinion easy to find, cheap, giving good performances.
From my side, I started with DML something like 3 years ago and the result was a good surprised. We are in the same category of "happy amateur". We try to push the topic with our different backgrounds, knowledge to get the best of the panels.
Have you had a look to the main thread A Study of DMLs as a Full Range Speaker ? Very long, heavy, difficult to read but a great area of exchange. This one here is more specialized.
To help you in finding topic, you can use the history file I posted. See under my signature on the left of the post and follow the link to github. In the history file there are links of posts of realizations or topic on interest (of course of my interest!).
Hope it will help you to start but do not hesitate too much... Plywood (poplar plywood) is a good entry material in my opinion easy to find, cheap, giving good performances.
Hi Christian,
The first material is a foamed rubber. Here is a link to the Amazon page https://www.amazon.co.uk/Self-Adhes...id=1675254893&sprefix=foam+tape,aps,87&sr=8-6
The second is indeed a Sorbothane
I agree using the metrics from the Sorbothane page is good in principle but I would like to find out how they measure samples so we can standardise measurement of non-Sorbothane samples as well, e.g. the Foowong foamed rubber I am also using. Deflection percentage is a variable that we should be able to measure with a fixed load and probe diameter. We can calibrate any measurement method using the published figures from the Sorbothane Inc website
Its a good thought Christian,
Yes our damping material has to provide support as well as the damping action. Support in the car suspension is provided by the coil spring to suspend the car mass, it is the shock absorber that does al the work absorbing the input from the road surface.
But you have prompted the thought that we might be able to separate out the functions. We could provide the suspension element either by suspending the panel from the top of the frame or by providing compliant mounts on the bottom edge, a couple of small rubber contact points on the base would be enough to suspend the mass from the frame. We would then be able to explore a wider range of damping media including sealing material like silicon? Just a thought.
Burnt
Yes our damping material has to provide support as well as the damping action. Support in the car suspension is provided by the coil spring to suspend the car mass, it is the shock absorber that does al the work absorbing the input from the road surface.
But you have prompted the thought that we might be able to separate out the functions. We could provide the suspension element either by suspending the panel from the top of the frame or by providing compliant mounts on the bottom edge, a couple of small rubber contact points on the base would be enough to suspend the mass from the frame. We would then be able to explore a wider range of damping media including sealing material like silicon? Just a thought.
Burnt