Bad to confuse "making beautiful music" with "reproducing music in a neutral manner".
But having said that, decades ago using the original Karlson-15, I was amazed at how great the speaker was at reproducing a cello. Maybe my imagination but I also imagine that large metal horns play brass well (heard some great ancient JBLs at the Toronto get-together on the weekend).
But that's quite different than taking an actual cello and putting an exciter on it.
B.
Last edited:
well i dont want to stick a feather in my own arsbut i think mine sounded better. but then again he used very cheap of the shelf exciter so in terms of simplicity he got the better of me for sure
YouTube
another problem with a guitar body is.... well it is a guitar body thats is tuned to a specific task. making guitar noises
The resonance is that of a guitar so guitar would sound as it should. I will put the exciter in front which should result in better vocals and other sounds.
Other DML sound bit thin and inaccurate. Your setup sounds amazing for vocals! Do you have other recordings for other genre of music? DML do very well with vocal reproduction but due to resonance issue sound bit off for other instruments. Do you have a build guide to show how you setup yours? Its not off the shelf driver for sure and what is the white thing sticking out for?
You use an active crossover but a mother sub so in a way its not quite apple to apples. Good work though
I got no other videos or picture im afraid. it has gone somewhere in the bin a long time ago
well that monster sub is an open baffle so it looks like it is more then it is
the white pieces are hpl that clamp the foam board and the reason it is diagonal is it creates the most different resonances that make up a DML. if i would put it vertical i would create less peaks but over a bigger area so the peaks and nulls would be insanely high. i started with just the square foam boards and then adjusted the hpl pieces to get the smoothest response possible for this thing.
The driver is indeed home made, it is a magnet motor of a rubanoide (you can find it here or on google) then i used also a coil from a rubanoide. but there questions arises would not making a complete ruba more efficient ? YES but the dml could win if it would drive a much larger board. you should be able to go lower and still have this high end as well.
on my channel there i tried once with smaller motor and bigger panel what it could do.
YouTube
well that monster sub
is an open baffle so it looks like it is more then it is the white pieces are hpl that clamp the foam board and the reason it is diagonal is it creates the most different resonances that make up a DML. if i would put it vertical i would create less peaks but over a bigger area so the peaks and nulls would be insanely high. i started with just the square foam boards and then adjusted the hpl pieces to get the smoothest response possible for this thing.
The driver is indeed home made, it is a magnet motor of a rubanoide (you can find it here or on google) then i used also a coil from a rubanoide. but there questions arises would not making a complete ruba more efficient ? YES
but the dml could win if it would drive a much larger board. you should be able to go lower and still have this high end as well. on my channel there i tried once with smaller motor and bigger panel what it could do.
YouTube
Panel stiffness vs low frequency response
I have been reading as much as I can find about DML, even renewing my Audio Engineering Society membership, and I especially want to understand what characteristics of the panel mean for the performance. I am stuck on the aspect of panel stiffness vs low frequency response. In the first post of this topic, it states that stiffer is better for lower frequencies. This is counterintuitive to me because if you tighten a guitar string or a drum head the sound goes higher. I also think that one of the research papers stated that less bending stiffness gave lower frequencies. I will see if I can locate that paper.
I have been reading as much as I can find about DML, even renewing my Audio Engineering Society membership, and I especially want to understand what characteristics of the panel mean for the performance. I am stuck on the aspect of panel stiffness vs low frequency response. In the first post of this topic, it states that stiffer is better for lower frequencies. This is counterintuitive to me because if you tighten a guitar string or a drum head the sound goes higher. I also think that one of the research papers stated that less bending stiffness gave lower frequencies. I will see if I can locate that paper.
I have found a very good paper https://pdfs.semanticscholar.org/aa90/bff6113cfcb5ce4979e757d8b749f13ea168.pdf At the bottom of the page numbered 2756 they give equations which show that for low frequencies you want low stiffness, high mass and large area, but for high efficiency you want high stiffness and low mass. So there must be a balance between efficiency and low frequencies, just like with cone loudspeakers. I will think about predicting low frequency response and efficiency of different panel materials by using the values for stiffness and density to scale from a measured panel with known stiffness and density.
just discovered dml's and am reading and collecting materials. got xps foam and 5mm nomex honeycomb. carbon fibre sheet is a bit expensive for a first experiment so i am using i mm whale brand tufnol. it is also called paxolin or srbs phenolic sheet.. waiting for the exciters to arrive from holland. i am uk based. the exciters are the dayton audio 40 watt ones. the aim is maybe a bit different to most on here in that i want to use dml's for a small PA system. the reduced feedback could be very useful in eliminating the need for monitor speakers, which can be a problem in small venues due to lack of space. the tectonics you tube vids look impressive. anyone else played with tufnol instead of CF?
The balancing point is
High stiffness
Low mass
Large area
With a powerful transducer
And with weights on out of control resonant modes, you can tune the highs and lows at the same time
I would say that was a double guitar!
Fingers plucking strings which resonate. This is the input. This in turn goes through the guitar transfer function which is basically a Helmholtz resonator.
What he is doing is taking that and putting it through another guitar transfer function so you end up having a double guitar! If you don't want a double guitar this is not ideal.
Anyway I digress.
The DLM speaker acts much like the soundboard of a guitar. I have been researching this. It turns out luthiers have been using nomex hexagonal composites for years now.
Making Guitar Double Tops with Nomex Part 1 - Luthier Tips du Jour Episode 46 - YouTube
Luthier Tips du Jour - Double Tops - Part I of II - YouTube
(there are more videos if you search)
A few points:
- They say in part 1 that the hex core makes wood grain irrelevant. This is interesting as Young's Modulus (E) is 20-30% the value across grain as it is with the grain in wood.
Acoustic radiation coefficient = sqrt(E/p^3) where p = density.
So basically loudness is determined from high stiffness but more so low weight. The very best wood for soundboards has traditionally been spruce which has just about the highest radiation coefficient of any wood (except balsa which is too fragile for guitars). It makes sense to use composite hex core as soundboards in guitars.
- In part 2 they show exactly how to make it. They use 2mm nomex so the OP used nomex that was too thick. I ordered some from UK - it is cheap except for delivery costs!
Also they glued differently. They used thixotropic glue and applied it to the nomex not the panel. This would greatly reduce weight. Of course they also used vacuum but I understand this is not practical for most of us.
And they used tonewood. I wonder if you can just use veneer instead of carbon?
Anyway I have ordered 2mm nomex, some 0.5mm carbon fibre sheet and also a variety of veneers to try out. Spruce veneer is unavailable here but basswood has virtually the same radiation coefficient and is cheap - available from craft stores. Grain is also less pronounced. I suspect the nomex sheet will dominate the effect over the type of panel used.
You can also buy assembled CF honeycomb sheet now. Price is not too bad. I ordered this from Aliexpress:
500mm x 500mm x 5mm 3K twill glossy carbon fiber and Aluminium honeycomb panel, sandwich carbon fiber plate|carbon fiber plate|fiber platecarbon plate - AliExpress
Price is ouch but if it works it is not bad. It is enough material to make up to 6 panels depending on size. If it works well even just for 1 panel price is OK.
I think people have been concentrating on big panels in the other thread. I like the ideas in this one ... small panels ... hex composites ... maybe just a few tweaks can get them working better (less glue and thinner, maybe use much cheaper wood veneer too). Even large panels need woofer augmentation generally so why not just use smaller ones?
I like the idea of wooden speakers. It freaks me out!
Last edited:
PS. I made up this spreadsheet. I left out a few that were obvious no-brainers like balsa (too fragile and most of the heavy woods). I also used only veneers I could buy locally. As I said I expect the core material to have dominant effect. I also wonder about how to finish the wood. I like hardwax oils but do not know if they will work without testing.
wood Youngs modulus(GPa) density(kg/m^3) R
Bamboo 18 500 0.000379473
Spruce 11.03 425 0.000379057
Basswood 10.07 415 0.000375355
Aus Cedar 9.22 485 0.000284284
Coachwood 15 600 0.000263523
Dillenia 12 575 0.00025124
QLD Maple 10.83 560 0.000248332
Blackwood 14.82 640 0.000237768
BH Sassafras 12.6 620 0.000229931
Silver Maple 7.86 530 0.000229772
Tas Myrtle 12.62 625 0.000227358
African Mahogany 10.6 640 0.000201087
Ash 12.31 680 0.000197863
Silky Oak 7.93 590 0.000196498
Maple 11.8 690 0.000189525
Zebrawood 16.37 805 0.000177146
wood Youngs modulus(GPa) density(kg/m^3) R
Bamboo 18 500 0.000379473
Spruce 11.03 425 0.000379057
Basswood 10.07 415 0.000375355
Aus Cedar 9.22 485 0.000284284
Coachwood 15 600 0.000263523
Dillenia 12 575 0.00025124
QLD Maple 10.83 560 0.000248332
Blackwood 14.82 640 0.000237768
BH Sassafras 12.6 620 0.000229931
Silver Maple 7.86 530 0.000229772
Tas Myrtle 12.62 625 0.000227358
African Mahogany 10.6 640 0.000201087
Ash 12.31 680 0.000197863
Silky Oak 7.93 590 0.000196498
Maple 11.8 690 0.000189525
Zebrawood 16.37 805 0.000177146
This is a great paper and basically tells us how we should be designing our panels rather than just guessing.
