This is the sort of material I am using with the epoxy panel.
I think I only paid about £20 for 19 sheets,some time ago.
I had to shop around quite a bit before I found a reasonable price.
From memory ,I think the epoxy sounds better than the pva ? On this material.
But it was a long time ago.
The only xps panels I have for comparisons are coated in vinyl.
But I will have a listen.
Steve.XPS Foam Underlay (9.79m2 Pack) Wood and Laminate Flooring Underlay - 5mm Thick - Insulation/Floor Levelling Boards by FloorPro(R) : Amazon.co.uk: DIY & Tools
I think I only paid about £20 for 19 sheets,some time ago.
I had to shop around quite a bit before I found a reasonable price.
From memory ,I think the epoxy sounds better than the pva ? On this material.
But it was a long time ago.
The only xps panels I have for comparisons are coated in vinyl.
But I will have a listen.
Steve.XPS Foam Underlay (9.79m2 Pack) Wood and Laminate Flooring Underlay - 5mm Thick - Insulation/Floor Levelling Boards by FloorPro(R) : Amazon.co.uk: DIY & Tools
Ok, Steve. Now I've checked that recording of yours and it sounded very nice.
Was it just the veneer panel that was used, that one with the dome?
Are there comparison recordings with and without the dome? What the dome contributes with?
The closer I might get around here to your veneer is thin plywood, which is veneered on both side, of course. Would that apply?
Has anyone tried Formica, alone or with some damping?
Remember my application will never use any eq of any kind, and will have to be complemented with a sub, because of distance from wall limitations on two of the places.
I will also try the DML boards on my main listening room, where I can separate a little bit more from the wall, around 50cm. Not possible on the other rooms.
Carlmart.
I believe the veneer panel was on the left and the card panel was on the right ,in that recording.
The dome helps the sound in many ways ,but the main thing is to prevent the cavity resonance at 2.5k.
Once again I would recommend the art canvas panel,it is easy to build and cheap.
You don't have to use it full rang but it is handy having a full range panel.
I think you would be surprised how good they sound .
Unless you want to spend years trying to find your own ideal panel ?
Steve
I believe the veneer panel was on the left and the card panel was on the right ,in that recording.
The dome helps the sound in many ways ,but the main thing is to prevent the cavity resonance at 2.5k.
Once again I would recommend the art canvas panel,it is easy to build and cheap.
You don't have to use it full rang but it is handy having a full range panel.
I think you would be surprised how good they sound .
Unless you want to spend years trying to find your own ideal panel ?
Steve
Carlmart.
On page 220 post 2201 you will find a recording of my art panel ,just one art panel on its own playing just one channel only ,no sub.
I think you will agree that it sounds pretty good and has a full bodied sound.
Imagine how good it would sound with two panels in stereo.
There are some poor quality frequency plots further down which show a max deviation from 100hz to 20k of +- 5db ,this panel did do better but the cascamite coating I applied cracked up and split.
I would hope to get a response nearer -+ 3db with no large peaks in the 400hz region.
Steve.
On page 220 post 2201 you will find a recording of my art panel ,just one art panel on its own playing just one channel only ,no sub.
I think you will agree that it sounds pretty good and has a full bodied sound.
Imagine how good it would sound with two panels in stereo.
There are some poor quality frequency plots further down which show a max deviation from 100hz to 20k of +- 5db ,this panel did do better but the cascamite coating I applied cracked up and split.
I would hope to get a response nearer -+ 3db with no large peaks in the 400hz region.
Steve.
Also on page 347 post 3468 you will find a recording of the art panel and veneer panel both with domes.
This is a simple recording,using natural reverberation ,which comes through in my recording.
There are a lot of natural sounds and acoustics .
After 3minutes it reverts to mono as it is a free app !
Steve.
This is a simple recording,using natural reverberation ,which comes through in my recording.
There are a lot of natural sounds and acoustics .
After 3minutes it reverts to mono as it is a free app !
Steve.
Edge modes, canvas and long speakers
I've been reading the archives back when discussion of canvas panels started, around 100-110. Around the same time, Steve shared his learnings from the long panels he made for his brother (and name-dropped Wittgenstein
).
Well Im not really a follower of Witters when he says 'whereof we cannot speak, therof we must remain silent' Where would this forum be if we adhered to that dictum ?
This may well have been discussed in these terms since those earlier pages, but if I read all the intervening pages I'd forget what I wanted to say anyway. So here goes...
Anyhow, to get to the point I would like to make concerning edge modes. The theoretical far-field SPL from an infinitely large bending-wave speaker at low frequency (ie below critical frequency) is zero. This is because the speed of waves on the panel is lower than that in air, and sound cancellation occurs between adjacent high and low pressure areas across the face of the panel. Apart from the lowest pistonic (1,1) mode, the far field LF SPL we get from DML speakers is due mainly to edge and corner modes, which essentially are the un-cancelled fragments which are left over at the edges, with no adjacent fragment of opposite phase close enough to cancel it before it can escape as a sound wave in air to the far field.
Anyhow, this fact goes a long way to explaining the LF performance of long panels and canvas. Long panels have a higher ratio of edge length to area, so a greater portion of the energy of the propagating wave will reach an edge, and consequently a chance to reach the far field, before it decays away.
Similarly with regard to membranes I believe their enhanced LF response is due to the fact that a greater portion of the energy reaches the edge rather than being wasted in the centre. Bending waves propagate more rapidly and expend more energy moving and compressing the air near the panel - both the mass and the elasticity of air playing a role. But the transverse waves on a membrane are much slower, similar to the moving waves on a taught rope when you shake it up and down. So they interact less with the inertia and elastic nature of the air - they just push it along as they move. Because they lose less energy in transit, more reaches the edge where it can do some good in the far field.
I've been reading the archives back when discussion of canvas panels started, around 100-110. Around the same time, Steve shared his learnings from the long panels he made for his brother (and name-dropped Wittgenstein
).Well Im not really a follower of Witters when he says 'whereof we cannot speak, therof we must remain silent' Where would this forum be if we adhered to that dictum ?
This may well have been discussed in these terms since those earlier pages, but if I read all the intervening pages I'd forget what I wanted to say anyway.
So here goes...Anyhow, to get to the point I would like to make concerning edge modes. The theoretical far-field SPL from an infinitely large bending-wave speaker at low frequency (ie below critical frequency) is zero. This is because the speed of waves on the panel is lower than that in air, and sound cancellation occurs between adjacent high and low pressure areas across the face of the panel. Apart from the lowest pistonic (1,1) mode, the far field LF SPL we get from DML speakers is due mainly to edge and corner modes, which essentially are the un-cancelled fragments which are left over at the edges, with no adjacent fragment of opposite phase close enough to cancel it before it can escape as a sound wave in air to the far field.
Anyhow, this fact goes a long way to explaining the LF performance of long panels and canvas. Long panels have a higher ratio of edge length to area, so a greater portion of the energy of the propagating wave will reach an edge, and consequently a chance to reach the far field, before it decays away.
Similarly with regard to membranes I believe their enhanced LF response is due to the fact that a greater portion of the energy reaches the edge rather than being wasted in the centre. Bending waves propagate more rapidly and expend more energy moving and compressing the air near the panel - both the mass and the elasticity of air playing a role. But the transverse waves on a membrane are much slower, similar to the moving waves on a taught rope when you shake it up and down. So they interact less with the inertia and elastic nature of the air - they just push it along as they move. Because they lose less energy in transit, more reaches the edge where it can do some good in the far field.
On page 156 I found some plots I made of my large sanded panels ,if I'd fixed the 2.5k and 9k spikes and used the 20db scale that everyone seems to use,this would probably look like a straight line from 150hz to 20k.
I came across this as I was looking at the first mention of the art panel by offgrid on page 101.
I got carried away and started reading it all again, I'm up to page 158 and he hasn't yet given full details of his build?
I hope I find it soon.
But I am enjoying the read ,even if it is going around in circles.
Steve.
I came across this as I was looking at the first mention of the art panel by offgrid on page 101.
I got carried away and started reading it all again, I'm up to page 158 and he hasn't yet given full details of his build?
I hope I find it soon.
But I am enjoying the read ,even if it is going around in circles.
Steve.
Finally found the original post of the art panel on page 88 with pictures.
Steve.
Steve,
Offgrid said in a later post that the sticks that he had affixed to the back of the canvas were overkill, and not necessary. He ended up only using the 4x6" wooden placard in the center for the exciter. Later work confirmed that the caulking that he did around the edge of the placard and where the canvas meets the frame were also unnecessary.
Offgrid said in a later post that the sticks that he had affixed to the back of the canvas were overkill, and not necessary. He ended up only using the 4x6" wooden placard in the center for the exciter. Later work confirmed that the caulking that he did around the edge of the placard and where the canvas meets the frame were also unnecessary.
i was getting a little board as my back is playing me up so i cant go out in the garden digging
so i have done another recording this time of my two epoxy panels .
not that anyone will notice , the left channel has now got a paper tissue type of dome, a strong tissue that is used for cleaning glasses.
tea bags just popped into my mind, mmmm i wonder if my lovely wife would like a cup of tea
tea bags are pretty indestructible ,will not fall apart when wet and breath when dry.
anyway this is the recording , the dome panel has totally lost its 9k and 2.5k peaks, which is nice
steve.
so i have done another recording this time of my two epoxy panels .
not that anyone will notice , the left channel has now got a paper tissue type of dome, a strong tissue that is used for cleaning glasses.
tea bags just popped into my mind, mmmm i wonder if my lovely wife would like a cup of tea
tea bags are pretty indestructible ,will not fall apart when wet and breath when dry.
anyway this is the recording , the dome panel has totally lost its 9k and 2.5k peaks, which is nice
steve.
Last edited:
Me too. I went to check them out and the ones I saw seemed a bit floppy. But that doesn't seem to matter (and can be fixed). I have more exciters on the way so I will play with them.
I made this dome and had it up and running in about half an hour, with the use of a hair dryer.
Or would have ,if I had not ripped the first dome
I was in a hurry, so it's not my best effort.
This is the first epoxy panel .
The plot is from about 3 inches in front of the dome , please do not regard this as the room response.
As can be seen , there is no 9k or 2.5k peaks
Eventually I intend to cut the panel in half and see how this affects the response.
There is a very slow roll off above about 2k or so .
Steve.
Or would have ,if I had not ripped the first dome
I was in a hurry, so it's not my best effort.
This is the first epoxy panel .
The plot is from about 3 inches in front of the dome , please do not regard this as the room response.
As can be seen , there is no 9k or 2.5k peaks
Eventually I intend to cut the panel in half and see how this affects the response.
There is a very slow roll off above about 2k or so .
Steve.
