Christian.
In the link in burntcoils post, they use 3D measurements.
This is I believe the only real way to know the full output of a DML panel speaker.
Although I am not sure if they are including the exciter area output?
They show a snapshot of a panel response in fig 2 , but shortly after excitation with an impulse?
Steve.
Pic 1
The small EPS beads inside the panel before applying 50x50 mix PVA.
Pic 2.
With a good splashing of PVA.
Pic 3.
With masking tape holding the panel tightly in place, with wood blocks on top to help.
The kids are coming over this afternoon , so I don't hold out much hope of the panel staying in one piece 😳
Steve.
Ps I seem to have pic 2 first as a full picture somehow ?
The small EPS beads inside the panel before applying 50x50 mix PVA.
Pic 2.
With a good splashing of PVA.
Pic 3.
With masking tape holding the panel tightly in place, with wood blocks on top to help.
The kids are coming over this afternoon , so I don't hold out much hope of the panel staying in one piece 😳
Steve.
Ps I seem to have pic 2 first as a full picture somehow ?
Attachments
Hello Steve
Yes should be very interesting.
The 50mm is excessive for a whole panel but I wonder if an increase of the thickness locally might be good. The reason is that in the hypothesis of a low pass between the voice coil mass and the panel mechanical impedance, to increase the mechanical impedance increase the cut off frequency; a way to increase the mechanical impedance is to increase the thickness (mechanical impedance follow h², square of the thickness). The question is to understand if xe won't go then to other problem where the thickness of the plate is greater than the wave length. Some patents shown a criteria on that.
So if you have the samples of the material, the time, it is better to enlarge the tests range.
So by the way, I have nothing against the fact you communicate again your tricks to get HF from EPS. It is with pleasure I will add a paper from you in the github repository. I can even propose to write it in your name, with your authorization of course, and your inputs.
Christian
Hello Steve
The snapshot in fig2 is a simulation like if the panel was hit with a very fast hammer (impulse).
For the measurement, I think it is written in the Tectonic document that is done in 3D. It is also a technique used for loudspeaker (see M Floyd's work and book) but I guess in an anechoic chamber. It is a kind of standardized method to get a response that is said closed of how the loudspeaker are appreciated in a room avoinding the question of the room. As it is done at distance (2m?) the exciter area is part of the measurement. In our measurement except if we use some time window (window that rejects the signal above a certain time), the room is part of the measurement with the problem to add its own signature.
Christian
Hello Paul,
The range of density seems standard.
If I understand the compressive stress is the limit in compression of a material while the young modulus describe the elasticity in the linear region which is what is involve in plate modes.
I search a little bit and I found some papers showing a linear relation between the density and Young modulus.
See the example below from Selected Engineering Properties and Applications of EPS Geofoam
This kind of papers are about EPS Geofoam. Do you know if it has specific properties compare to what we can find?
By the way, it is consistent with what I have in my notes : about 5MPa for a 15kg/m³ EPS. There is an important dispersion.
We could start from a simple relation like E = rho/3 E in MPa with rho in kg/m³ or more optimistic for the high density E = 2.rho/3 -5.
We'll see where it leads.
More data will be need as the coating change the characteristic. Coating add mass. I measured 0.06kg/m² for the EPS above changing the areal mass from 0.3kg/m² to 0.36kg/m² with 2 layers of coating on both sides. Is it thin, thick... no idea, no reference.
It changes also the bending stiffness which can be translated in E changing from 5Mpa to 8Mpa (big impact!)
I would like just encourage again to have more density or areal mass measurements but also bending stiffness or Young modulus estimations. 2 methods were shared here : 3 points or tap test.
Christian
Steve, could you precise which fig you are referring to for the 3D plots?
Going at the end of the document, fig 11 there are impedance curves. they are strangely smooth. I was expected more "spikes" at the main modes. Like here if the modes are not visible on the electrical side.
It make sense the exciter area is louder. The energy arrives from the exciter here and after is more or less absorbed by the panel.
Christian.
It is very difficult to generalise about what affects what panel type.
I treat an EPS panel totally different from a ply panel.
On harder heavier panels like ply, my exciters shows a large increased HF hump above 10k to 20k.
On lower grades of thick raw EPS there is usually a 5db to 10db drop in response at about 10k ,it is still there up to about 20k but at a lower level.
Sometimes I thin the area in the coil area to alough more sound to come through the panel, this is usually in a concave dome shape as it is easy to do.
But a constant directivity horn shape might be better?
If too deep you tend to start getting a midrange hump , a bit like having a midrange unit in the panel, or a whizzer unit.
It tends to make the panel sound more intimate, and bring the sound round to the front more.
My answer to people saying dml always sounds like it is coming from behind the panel.
I have no idea if it will work on this 35mm soft panel, the 50mm was too thick for my liking.
But you don't know until you try?
Steve.
I don't mind people quoting me as long as it is not turned into meaning something totally different.
When parts express started promoting using PVA on XPS , because of my posts on NXT RUBBISH using EPS, I was not happy.
He acknowledged my idea but It was not something I would have recommended.
I was never satisfied with XPS and PVA did very little to help.
If you want to improve the sound and performance of XPS I would now recommend epoxy.
All I can say is pva sounds very good with EPS, if the instructions I give are followed to the letter.
A different exciter could also change the sound and performance, so once again generalising is very difficult.
Steve.
It is very difficult to generalise about what affects what panel type.
I treat an EPS panel totally different from a ply panel.
On harder heavier panels like ply, my exciters shows a large increased HF hump above 10k to 20k.
On lower grades of thick raw EPS there is usually a 5db to 10db drop in response at about 10k ,it is still there up to about 20k but at a lower level.
Sometimes I thin the area in the coil area to alough more sound to come through the panel, this is usually in a concave dome shape as it is easy to do.
But a constant directivity horn shape might be better?
If too deep you tend to start getting a midrange hump , a bit like having a midrange unit in the panel, or a whizzer unit.
It tends to make the panel sound more intimate, and bring the sound round to the front more.
My answer to people saying dml always sounds like it is coming from behind the panel.
I have no idea if it will work on this 35mm soft panel, the 50mm was too thick for my liking.
But you don't know until you try?
Steve.
I don't mind people quoting me as long as it is not turned into meaning something totally different.
When parts express started promoting using PVA on XPS , because of my posts on NXT RUBBISH using EPS, I was not happy.
He acknowledged my idea but It was not something I would have recommended.
I was never satisfied with XPS and PVA did very little to help.
If you want to improve the sound and performance of XPS I would now recommend epoxy.
All I can say is pva sounds very good with EPS, if the instructions I give are followed to the letter.
A different exciter could also change the sound and performance, so once again generalising is very difficult.
Steve.
Could you please give your subjective feelings, what your ears hear, rather than what the mike hears, and the computer program interprets? You've made a lot of DMLs with lot of materials and gave us lot of info. It'd be interesting to read a summary. Thanks!
My understanding is fig 2 is a kind of snapshot of a simulation of an impulse response at a certain time. Fig 3 are also probably simulations, each for a specific frequency, showing the SPL in different directions; no influence of the time here. In fig 3, the distance is probably set at "far enough" to take all the element of surface (including the central one)
First picture shows the area around the exciter I coated in thin mix of PVA .
The second is the cocktail stick and tube I use to slide the exciter into position , I also use this to make holes in my panels.
Third picture is with the exciter in position.
The fourth is with a little weight to push the coil into the EPS a little, for a good bond.
The panel join feels pretty strong now , but I shall try and restrain myself from listening until tomorrow ,if I can 🤗
Steve.
The second is the cocktail stick and tube I use to slide the exciter into position , I also use this to make holes in my panels.
Third picture is with the exciter in position.
The fourth is with a little weight to push the coil into the EPS a little, for a good bond.
The panel join feels pretty strong now , but I shall try and restrain myself from listening until tomorrow ,if I can 🤗
Steve.
Attachments
Some of the current discussion on this thread re: cantilevers and my wife flying to Toronto this afternoon, leaves me some time to experiment this weekend.
Please advise if you think this experiment sounds nuts.
I've spent a lot of time with my new set of Abstracta frame + 3mm baltic birch DMLs that I have on either side of my TV.
I've now started wondering about adding some kind of center channel to help clarify voices particularly in films.
I've got an LG TV (see attached image). The side view shows how its screen (some kind of plastic/acrylic) cantilevers up from the its base.
The freely cantilevering area measures 57" wide x 19" high (3 to 1).
I've got an extra Dayton Audio DAEX32Q-4 exciter.
Can anyone recommend a temporary (easily removable) way to attach the exciter?
And again, please advise if you think this experiment sounds nuts.
Please advise if you think this experiment sounds nuts.
I've spent a lot of time with my new set of Abstracta frame + 3mm baltic birch DMLs that I have on either side of my TV.
I've now started wondering about adding some kind of center channel to help clarify voices particularly in films.
I've got an LG TV (see attached image). The side view shows how its screen (some kind of plastic/acrylic) cantilevers up from the its base.
The freely cantilevering area measures 57" wide x 19" high (3 to 1).
I've got an extra Dayton Audio DAEX32Q-4 exciter.
Can anyone recommend a temporary (easily removable) way to attach the exciter?
And again, please advise if you think this experiment sounds nuts.
