Hello Lekha,
Could you make your question more precise? Which phenomenon do you refer to?
Christian
My question is still actual, strangely enough that coil (and several iterations of such coils) had produced commercial quality flat panel speakers, that is supposed to be using "bending waves" or "distributed mode" waves on the surface.
It has also been found that to create a well-sounding flat panel speaker, one must not only develop the panel membrane, whether composite or otherwise, but most importantly, the driver, whether it is of the "pistonic motion" type or another design. Anyway, @Monterverdi in this forum, but in another thread, is attempting to do both. Kudos to him!
I've provided more than enough information in quite a few earlier posts. The responses to them suggest that no one in this thread would attempt any of it, which, in a way, is not bad at all. 🙂
Hi Christian
I'm so sick of reading patents that I just jumped to conclusions and didn't see any reference to venting...sigh .. I'll just have to read it now I guess
My part is 2.4g, used 0.81m of 1.75mm dia filament and cost 5c. 😊
I guess it's all a compromise between competing elements.. It's nothing compared to the mass of the panel but in the exciter zone it is significant. Does the gain in the reduction in diameter outweigh (pun intended) the added mass.?.. No idea.
Cheers
Eucy
I'm so sick of reading patents that I just jumped to conclusions and didn't see any reference to venting...sigh .. I'll just have to read it now I guess
My part is 2.4g, used 0.81m of 1.75mm dia filament and cost 5c. 😊
I guess it's all a compromise between competing elements.. It's nothing compared to the mass of the panel but in the exciter zone it is significant. Does the gain in the reduction in diameter outweigh (pun intended) the added mass.?.. No idea.
Cheers
Eucy
Tefra seems to have experimented with the flat coil, but they (very obviously) use standard Billionsound produced drivers in their products. In some of the youtube videos you can see them laying on the desk. You can also spot the aluminium/(paper?) honeycomb panels.
The picture you link to shows that they use the "window insulation type" foam on the edge of the panel, not on the back. That provides a very soft/compliant surround, and the light membrane might work pistonic for the low frequencies - hence their quite spectacular low end response?
Hello Eucy,
I'm interested how well your "concentrator" works. My version works well for my paper/nomex panel, but if I try it on other panels (acrylic/plywood/xps foam) it fails miserably, reducing HF output quite extremely.
I can only guess why. Plastic too soft to drive heavier panels at HF? And/or contact surface too small for the softish foam? Does the drumhead like resonance usually provide most of the HF response?
Question, questions.
It remains a mystery but I keep experimenting, stumbling through the dark🙂
Hans
+ @Eucyblues99
HvdZ, how is your paper/nomex panel in word of areal mass (I have in mind heavy but I am not sure...) ?
For now I have done few tests (more should come this week) on an EPS panel. My first concentrator was 1.9g, the second one 0.7g
Below are the measurements of the EPS panel (30x40cm, 10mm thcik)
Blue is without concentrator with the drum effect at about 13kHz
Green is 0.7g model. No more drum effect but no gain in HF
Red is with 1.9g model showing a loss of HF. I can't explain the behavior above 10k...
Christian
Maybe because there are on top other topics which are clear problems for which we have in mind a way to come to solutions. The exciter comes probably at the moment in third row in the topics I would like to work on and I haven't seen or understood in what you provided something to change that. What are the technical arguments (I have some to want to improve the current exciter but a flat coil is not in my current understanding a solution), the analysis, the pratical results or tests. I am ready to consider this point of view but pictures, videos or patents (I don't remember a patent on that flat coil) are not enough.
Christian
My panel is about 4mm thick 600g/sqm. So reasonably light I think?
I just tried the panel without concentrator again, and the most obvious thing is that the screaming 3k peak is there again, so maybe the ventilation holes were beneficial, and the concentrator less/not at all - highs might have been there but masked by the mid peak.
Sorry that I can't show measurements, I'm more trying/listening - which lead to unfounded speculation.
greetings, Hans
That's not the question at all for the engineer (or the others). It's interesting how people overlook the issue. Several videos in between have also highlighted it, for example, this one.
It is the wire that produces sound, in this case, a flattened wound rectangular coil. It could be round, oval, whatever.
In so-called planar panels, the wire is fixed all over the panel in a single strand (sometimes a few strands) and is covered with a multitude of magnets. The panel serves merely as a diffuser/amplifier of the sound. Manufacturers attempt to explain the "phenomenon" by discussing how the thin panel moves from one magnet to another, creating "sound pressure" or using similar terminology.
A thin panel (the thinner, the better) can serve as a carrier for the flattened singing coil, eliminating the need for a spider. And you don't need a panel the size of half a door.
That it! 👍
Christian....How did you get them down to that weight and still function... I'm swoggled! Maybe they're collapsing under hf load??
Eucy..
For me 3D printing is not among the options... So light wood and glue (cyano type)... I don't know if they are reliable enough over the time but they seem not "weak". For tests, they are glued with PVA to the exciter and to the panel. Strong enough for testing, weak enough to be remove without damages to the exciter. The EPS panel was damaged when I removed the concentrator. I guess it will be also the case of the plywood.
The 1.9g
The 0.7g
I see that you have different types of exciters.
What if you were to create a panel with that exciter, instead of that spider, similar to Manger's design but much stiffer? Additionally, the panel could be directly connected to the top of your voice coil. You could make the inner part of the voice coil area much thinner and/or adhere it with thin foil, for the HF. Of course, your panel could be much larger and not circular, but rather rectangular.
You may have noticed that coil diameter quite large for such a small panel.
Awesome! Sorry to re-explain everything, then.
The electrical impedance of the driver is related to the velocity of the coil, so if the velocity of the panel is high, the velocity of the coil is high (even if the magnet isn't moving at that frequency). Panel velocity is more or less inversely proportional to frequency, so the impedance peaks decrease in amplitude as frequency increases.
Acoustic loading and losses are (again, more or less) proportional to volume velocity, which is proportional to the average velocity on the panel surface. The lowest-order mode has a high average velocity because there are no positive/negative peaks that cancel each other out. High-order modes have a lot of cancellation, so the average velocity over the panel surface is much lower in magnitude, and acoustic losses are much lower. This is the same idea as dipole cancellation, where the font and back motion of an unbaffled speaker cancel out. As @homeswinghome mentioned, most researchers ignore acoustic loading, but I probably shouldn't - it seems like it has a dramatic effect on the Q of the lowest mode, especially when the material is light!
There is only one type of exciter on my pictures (and in general in my tests, few exceptions) which is the DAEX25FHE. It is shown in the post with 2 kind of concentrator.
Sorry, I haven't caught the idea. The picture of you post is the Manger membrane with its voice coil and the peripheral absorber (bending wave driver).
Christian, I appreciate your craftsmanship! I wonder if coil former stiffness/deformation is in play (especially in your 0.7 gram version), as discussed in Zenker's paper here.
Bruce
In the Manger membrane, the coil former is glued to the membrane; in other words, it is part of the membrane (in direct contact), unlike in exciters, but similar to "normal" cone speakers. The Manger membrane is quite flabby. For distributed mode operation, the membrane needs to be relatively stiff while also being quite thin.
Christian,
Quite a departure for me, right? That path is complicated, and not even clear for me to be honest. But basically I eventually came to understand (from Hambric) that the fundamental is pretty effective over a broad range above it's own resonance and maybe doesn't need the "help" of the next higher order modes. So, I wondered, are the next bunch of higher order modes really all that helpful? Or do they just add more peaks, lobes and ringing? I read also Dave's papers and liked the concept of really controlling the modal behavior. And while I don't have the skill to create such a complicated driver network, I could do some fun and interesting things with multiple exciters and judicious exciter placement.
I have not given up on the concept of high aspect ratio panels for spreading the most efficient modes evenly across the frequency spectrum, but I think I have found another way of getting pretty decent response across the first two and a half octaves.
Absolutely yes. I was planning to make those measurements and share them with you. Of course, it is the same panel for which we already looked at the coincidence frequency, so I don't expect those results to change, but it will be interesting to see the results below Fc, with the new exciter placement.
Eric
Dave,
So are losses proportional to average velocity or average absolute velocity? My thinking thinking is that acoustic loading losses should be proportional to the average absolute value of the velocity, rather than the average velocity. Taking just the average velocity would suggest that modes like the 1,2 mode of a simply supported panel would have no losses, since the positive and negative velocities perfectly cancel each other.
I think I understand why the the positive and negative peaks would cancel in terms of on-axis acoustic power output in the far field. But it's not clear to me that the damping effect should be similar in that regard.
As an avid cyclist, I know that pushing against the air slows me down regardless of which direction I go. So I don't see how having parts of the panel going opposite directions offsets the drag on each side. Seems to me it should double it. What am I missing? Are we talking about different effects?
Eric
True, lots of information. But I for one still don't know for sure what your question is.
But my best guess is that the answer is this:
If a wire carries electric current, and is in a magnetic field, then a force is generated on that wire. As a result, that wire accelerates in the direction of the force applied to it by that current and field. If that wire is attached to a thing, like a panel or cone or tape or tabletop or any other thing, that thing also accelerates along with the wire. When that thing (be it a cone, or panel) moves, due to that applied force, it pushes air. If the electrical current in the wire is time varying, with a characteristic frequency (or frequencies), the resulting force and motion of that thing creates pulsations in the adjacent air at the same frequency (or frequencies). Those pulsations in the adjacent air travel to your ear, and are perceived by your brain as sound.
Eric
WOW - There's a lot going on there Christian!
I think these things through intuitively, and while I can easily accept coil mass being a major influencer of performance on standard, light weight pistonic cones, I find it harder to see that a 0.7g addition to a panel weighing 30 grams (?) is going to have such a deleterious effect.
If you consider the standard ring mount fixing, that ring becomes an 'anchor' which can only move with the coil.
That very small ring zone is in fact wholly pistonic in action. Vibrations within and without the ring are a function of the stiffness of the panel. Transverse waves will travel across the ring region but the wave action must inflect (new use of the word) across the ring. Hence, within the ring zone, the panel with a large diameter voice coil will vibrate differently to a smaller one. We know that as the frequency rises, the vibrations centre around the exciter and if the panel material can transmit high frequencies without compression effects then the a smaller ring zone should be more effective at high frequencies than a larger one.
BUT - That pre-supposes that we still HAVE a ring zone to vibrate in response to, but with some degree of freedom from the coil.
In your designs, you pin the whole central exciter zone to the coil and hence remove the inner ring vibration from the panel. Although my plastic piece also has a central 'pin', it is not connected directly to the membrane in the same manner. (A bit hard to describe here at this time)
I wonder what the effect would be if you simply had a smaller ring atop your 'concentrator' and ditched the penetrating pin?
Something like this:
Obviously (or at least, intuitively 🙂), there is a limit to how small the top ring can be and still allow inner ring vibrations to be measurable, but maybe 10mm inner dia may work.
Also - I think it may be instructive to make small samples of your panel, say 50x50 or so, attach one to an exciter with no concentrator and do a free response test, then repeat with the same size but using the concentrator. The concentrator version will have larger cantilever sections than the non case but it may give some insight.
Anyway - As I said - intuitive.
Cheers 👍
Eucy
Last edited: