Thanks!
That's good to know. I'm curious on how thickness affects response. My gut says thicker = less efficiency, but also probably a smoother response with less resonances. Do you know how low your panels go?
Hello Aagas,
Plexiglass or acrylic allow to make very nice looking things. Do you have more data to share about that (perhaps remind previous posts...). In the synthesis (see post 4144) I think I have listed "only" the aluminium panel from you? additional information like the bending stiffness ant the arial mass is welcome to place the plexi you use in the material graph (see post 4169)
Thank you
Christian
Interesting question...
@Veleric : something tells me it is for you!
In my notes (I think from a thesis about piano soundboard), I have the efficiency linked to R = (E/rho^3)^.5 E = Young modulus of the material (in Pa Pascal), rho its density (kg/m^3). So nothing linked to the thickness... Hmm?
In an other hand, the lower frequency (f0 in previous post) is proportional to (D/µ)^.5 with µ (arial mass kg/m²) and D (bending stiffness Nm) as µ = rho.h (h the thickness) and D = E.h^3/12 it follows (D/µ)^.5 = h.(E/12/rho)^.5... all of that to say the lowest frequency should increase has the thickness (double thickness, double f0); to compensate the total surface should be doubled
Does somebody have experience that proves or invalidates that (always better to have facts against a theory)? Mainly the relation efficiency thickness... Not so intuitive to think there is no relation!
Christian
PS : in addition of interesting topics those exchanges are a way to extend my English vocabulary!
cheapvega | homeswinghome
I had the materials, exciters, and electronics laying about my workshop.I didn't think too much about optimizing the design.
I haven't even analyzed them with REW yet.
When (if) I get around to it I'll post more about them.
The panels measure about 18" by 24".
I have paired them with OB woofers. I think I crossover around 200 HZ (maybe a bit higher).
They sound pretty good.
Not nearly as majestic and immersive as my large aluminum panels, but very satisfying.
Hi Steve, I don't remember seeing any discussion here about putting holes in a "tweeter disc" used in the canvas type of DML. Have you seen this or had any experience with it? The disc is MDF on 5 mm foam board, attached with nylon bolts. http://projectgallery.parts-express.com/speaker-projects/dml-speakers-with-equalization/
Christian.
as I can't do live recordings anymore, I thought I'd try and measure the response from the ceiling.
the problem was that I could not turn on the pink noise while holding the panel up to the ceiling, my big toe would not reach that far 😁
The easiest and quick way was to play a CD with a good low and mid frequency response, the DEQ was set to PEAK HOLD .
I could then press play and stop on the CD remote , then creep silently over to the DEQ and take the photo.
I only played a few minutes of the track as it is not easy holding a panel up close to the ceiling.
I think it shows pretty well what happens to the sound when mounting a panel flat and too close to a wall of ceiling.
But the problem can be resolved by swinging the panel out from one edge.
This panel would have a maximum depth from the ceiling of 6inches.
And could be rolled off at 150hz to a sub.
It would have been better to let you hear the panel recording ,as I was very impressed with the sound output for such a small panel.
Steve.
as I can't do live recordings anymore, I thought I'd try and measure the response from the ceiling.
the problem was that I could not turn on the pink noise while holding the panel up to the ceiling, my big toe would not reach that far 😁
The easiest and quick way was to play a CD with a good low and mid frequency response, the DEQ was set to PEAK HOLD .
I could then press play and stop on the CD remote , then creep silently over to the DEQ and take the photo.
I only played a few minutes of the track as it is not easy holding a panel up close to the ceiling.
I think it shows pretty well what happens to the sound when mounting a panel flat and too close to a wall of ceiling.
But the problem can be resolved by swinging the panel out from one edge.
This panel would have a maximum depth from the ceiling of 6inches.
And could be rolled off at 150hz to a sub.
It would have been better to let you hear the panel recording ,as I was very impressed with the sound output for such a small panel.
Steve.
I just had a listen to my recording of the 6x9 veneer panel being held up to the ceiling.
I noticed something very interesting .
As I moved the panel up to the ceiling, which was moving the panel away from the microphone, the sound became louder ,as if I was moving the panel closer to the microphone ?
I did notice it sounded a bit brighter at the time, but I was standing with my arm outstretched to the ceiling and didn't think much about it as I was so close.
but the microphone 3m away was picking up a general wideband improvement in output .
I may have to look into this a little bit more if I get time.
When I made my recordings of the art panel up against the wall I was holding the panel in one hand and the microphone in the other , the microphone and me were too close to notice the difference, I think ?
I'll have to find them and have another listen?
Steve
I noticed something very interesting .
As I moved the panel up to the ceiling, which was moving the panel away from the microphone, the sound became louder ,as if I was moving the panel closer to the microphone ?
I did notice it sounded a bit brighter at the time, but I was standing with my arm outstretched to the ceiling and didn't think much about it as I was so close.
but the microphone 3m away was picking up a general wideband improvement in output .
I may have to look into this a little bit more if I get time.
When I made my recordings of the art panel up against the wall I was holding the panel in one hand and the microphone in the other , the microphone and me were too close to notice the difference, I think ?
I'll have to find them and have another listen?
Steve
@aagas
@bdjohns
@cheapvega
+@all...
Thanks for the information about acrylic/plexiglas.
I made an update of the "history file" and input the acrylic in the material sheet; including davelang's panel
Considering the good sound result, the density of the acrylic seems not a problem
Does somebody have an evaluation of the efficiency (xxdB/1W/1m) ? A frequency response to evaluate the damping in acrylic (I don't have a method for that but I think we could evaluate the Quality factor of the 1st resonance) ?
The standard panel sheet is now divided in 2 for readibility, an sort by post number.
Christian
@bdjohns
@cheapvega
+@all...
Thanks for the information about acrylic/plexiglas.
I made an update of the "history file" and input the acrylic in the material sheet; including davelang's panel
Considering the good sound result, the density of the acrylic seems not a problem
Does somebody have an evaluation of the efficiency (xxdB/1W/1m) ? A frequency response to evaluate the damping in acrylic (I don't have a method for that but I think we could evaluate the Quality factor of the 1st resonance) ?
The standard panel sheet is now divided in 2 for readibility, an sort by post number.
Christian
I see...
What about exporting a pink noise from REW, transform it in stereo and burnt a CD?
Attached a PN I use if it can help (2 channels in phase). Both files are compressed in zip. Is it a method you could use to send again mp4?
Christian
Yes it is the same ,sorry I missed your link.
I must admit that I forgot that I do have test discs with pink noise 😁
but it was just easy to use the music CD I already had in the CD player, as it had good coverage of the frequencies I wanted to test.
the CD had a very strong output from 40hz to at least 4k , this showed the panel roll off below 500hz when The panel was placed flat against the ceiling.
The pink noise looks pretty similar to this anyway(below 4k that is).
Steve.
I must admit that I forgot that I do have test discs with pink noise 😁
but it was just easy to use the music CD I already had in the CD player, as it had good coverage of the frequencies I wanted to test.
the CD had a very strong output from 40hz to at least 4k , this showed the panel roll off below 500hz when The panel was placed flat against the ceiling.
The pink noise looks pretty similar to this anyway(below 4k that is).
Steve.
i would not say that the density is no problem.
i would have loved to have see through panels ,they look fantastic.
but I was not impressed with the sound.
it has all the problems of a heavy panel plus a very bright harsh treble .
steve.
Christian,
I see it the same as you do. That is, at least based on D/mu^3, the thickness has no effect (theoretically) on efficiency for a homogeneous material. Simply because D and mu^3 are both proportional to thickness^3, so it cancels out. And yes, to match the same F0, the thicker material would require a larger panel.
I'd be interested to know the particular thesis about the piano soundboard with the E/pho^3, if you can recall it. I have seen that and the equivalent D/mu^3 parameter referred to many times in the patent literature, but not in a technical paper with proper references or derivations.
Eric
I had grand ideas for this. I was looking to build a SLOB bass driver with a DML on top. The thing was I was building smaller panels for tabletop use and in order to get good bass response the drivers needed to be large and/or in multiples and that would have made the speaker too big and cumbersome for my needs. Plus it would significantly increase cost, and the bass drivers would require a lot more power than the handling limits of the exciters, so you'd have to either do mixed wattage bi-amping or something. Also I had concerns about vibrations from the bass driver interfering with frequencies on the DML. It just ended up looking to be more hassle than it was worth and I decided if I was going to go down that route I would just build some giant OBs 😅
I know the answer will vary depending on the substrate and size of the panel, but hoping for some generalized input from the hive mind regarding the efficiency of DMLs. I ask because I am starting to get into tube amp building, and the single ended class A designs that I'm looking at building first are probably 3-5 WPC RMS at <1% THD. For those numbers you're usually looking at a speaker with a sensitivity in the high 90's. Would be fun to run some small to medium XPL or Balsa panels with a ~5 watt exciter but I'm not sure if the sensitivity is up for the task. It'd be a cheap experiment regardless but does anyone have experience with such a low-wattage scenario, or insight into approximate sensitivity potential vs FR potential? I'm used to using 25-40 watt exciters and having way more power available than they require, not the other way around. Thanks!
Christian,
This is fun detective work! But there are still some discrepancies for me. From the plot in the spec sheet I referenced earlier, I am convinced that the Fc of the panel is between 5 and 6 kHz. You calculated 14.5 kHz above, but I think your denominator for Fc should be 2*pi, not just pi. So in that case your estimates of 0.4 and 2.6 would give Fc of 7.3 kHz (a bit closer!) But assuming that the 0.4 number is correct for mu, D would have to be between about 5 or 6 Nm to get Fc down below 6 kHz.
I understand your reasoning regarding the D/Mu from the apparent F0. But personally, I'm more inclined to infer the relationship of D and Mu from the Fc equation than from "2.5*F0" rule. From my own measurements, I have found that rule is often not reliable. I suspect maybe the exciter may also limit the low end of the panel sometimes, not just the panel size.
But also I'm having a hard time figuring out how he panel could be over 3 mm thick. Any likely carbon fiber honeycomb panel I can model with a D of 5 or 6 is at most about 1.5 mm thick. Was the thickness mentioned in one of those videos? Because I don't see it on any of the spec sheets.
It may be hard really know what the Tectonic plate specs really are. It's hard to know how reliable the information they gave really is. And maybe they used a different panel at different times so not all our information is actually even the same panel! But it's fun (and I think worthwhile) to try and figure it out anyway.
Eric
NG,
Did you mean "XPS" instead of "XPL"?
I'm not a big fan of the sound of extruded or expanded PS (XPS or EPS) for DMLs, but I suspect they might be your best bet for this application. It might be possible too with some fancy carbon fiber honeycomb core panel, but also probably not anything you could buy "off the shelf".
Eric