Sometime ago somewhere in this diyAudio I wrote about cone speakers as somewhat DML, that there are vibrations going from the bobbin top to the surround at a given plane, and those waves give out additional sound, or that is the only sound we get from the cones. I had lot of resistance there. Everyone was sold on the idea of the 'pistonic' movement pushing the air... At that time, I didn't know about rubanoides, curved flat panels, etc. I took a long holiday from this site. When I came back here, one led to another, and got to know of rubanoides and Paul W Haddock's patents, actually his thinking. And, that's 3 decades too late! This new interest of speaker building, especially the distributed modes started in the covid times.
Paul Haddock had (and maybe still has) very good ideas, the type of ideas that tunes with me--back to the DMs in/on a paper cone. He knew about every flat curved panel connected to the voice coil is a separate speaker. He mentioned that in many ways, but not directly.
Regarding your sketch, the standard voice coil has a small perimeter, whereas the flat coil can be stretched, say for example, to an A4 size, and have a 29,7cm long curved speaker, actually 4 of them.
By the way, have look at this claim.
Last edited:
Oh, that state in normal for us.
I have zero knowledge on planar speakers. But, I am going to dig in. Will read all Haddock's patents, claims, applications etc and whatever related. Go back and watch Joppe building planars. Planars are DMLs...and doesn't push air to make music.
I'm glad I prevailed with the idea that a speaker cone is a DM surface, and that it gives out more sound perpendicular to its surface. I always had the commercially available "bookshelf" speakers always up-firing. and with nothing blocking the speaker plane, except the back wall.
Last edited:
In your sketch, the vibrations start at a tiny edge, while the end edge is much larger than the beginning. The vibrations won't get larger, just because the end edge is much more than beginning, but would die on the way. That's why its better to have both edges the same as wider as can be, and one edge fixed tightly, while the other edge with flat voice coil incorporated allowed a restricted movement. This allows the whole bent panel work as DML. Sure, at the fixed end vibrations die, but that's not much of a problem. If 60-80% area gives out sound from the voice coil, that's quite good.
I suppose you've noticed the width of a cone is not much in speakers. In a smaller ones, that might be twice the width of the dust cap, but in larger ones, they could be equal, or nearly equal.
When you take out the surround, you are left with ~2" wide angled paper cone from this 6.5" to do the work, in one plane if there's one.
But Haddock's speaker has same width speaker panels in 4 directions.
If you have a spare speaker driver, or you can buy one from the nearest flea market, just cut the basket and the paper cone out, you'd have your exciter. Only make sure the spider is intact, so the bobbin won't jump out. Touch the bobbin head to the tambourine.
You can also buy an old portable stereo, from which you can salvage two speakers, and more than that.
Hello
Seems plate and membrane don't have the same physics. Does a membrane keep the diffusive property of a DML? Not found the answer in a some minutes browsing.
Drum heads also a possibility. Different sizes available. Different skins. Pretty inexpensive. Some discussion earlier in this thread. As I remember the consensus questioned viability of a circular DML.
homeswinghome,
I believe this is the crook of the issue of panel vs. membranes. With panels, one has the resonances that show up and contaminate the original sound.
With thin membranes it eliminates the resonances and stored energy of a thicker panel. In my opinion gets a truer representation of the sound. The DML experience has never really impressed me as a great sound like some many others here have. I put myself as an outlier in this group and enjoy the diversity that this daily audio community has
I believe this is the crook of the issue of panel vs. membranes. With panels, one has the resonances that show up and contaminate the original sound.
With thin membranes it eliminates the resonances and stored energy of a thicker panel. In my opinion gets a truer representation of the sound. The DML experience has never really impressed me as a great sound like some many others here have. I put myself as an outlier in this group and enjoy the diversity that this daily audio community has
In post #7,548, where I mentioned one voice coil with 4 speakers, it should've actually written one speaker with 4 bent DM panels.
The voice coil is the speaker. The 4 panels bent in different directions are the 4 amplifiers of sound coil. Depending on the angle/curve these panels are bent, the sound (quality) changes. I have yet built this--awaiting the magnets to arrive, but I think, even if the membrane stays straight, it'd still give good sound. Only a standing alone membrane would just flutter. I wonder, if we can wound the flat voice coil around a glass panel. Or, paste it between two glass panels.
The voice coil is the speaker. The 4 panels bent in different directions are the 4 amplifiers of sound coil. Depending on the angle/curve these panels are bent, the sound (quality) changes. I have yet built this--awaiting the magnets to arrive, but I think, even if the membrane stays straight, it'd still give good sound. Only a standing alone membrane would just flutter. I wonder, if we can wound the flat voice coil around a glass panel. Or, paste it between two glass panels.
@tagis having built electrostatics for twenty years I would caution against assuming thin films are superior. My electrostatics used 10 micron PET and yet none of them had the sound field projection or the ability to recreate the leading edge of drum strikes or a piano keys percussion that DML’s do so easily. Timbre reproduction is also improved with DMLs. The release of stored energy certainly has an effect, I used to own a pair of Celestion SL600’s which sounded cleaner than the SL6S’s I had before due to the reduction in cabinet mass, but there are so many factors involved even though DML’s appear simple their mode of operation is not. When I built the Tall Blondes two supposedly identical Bitch ply panels had completely different sounds. One needed damping pads added to the panel in specific positions to control resonance peaks, the other didn’t need any at all. It took time to correct the bad resonances but once done the sound quality was excellent. DML’s can be frustrating but I find I still prefer them to my electrostatics despite having spent much more time and money on the stats.
Burnt
Burnt
Christian,
Good question, I am wondering the same thing. It's complicated to by the fact that not all membranes (planars) work the same way. I have very limited knowledge about planars but my understanding so far is that for most of them the motion of the memberane is at least intended to be "pistonic" in nature, by which I mean the entire membrane moving in phase. (if someone know this to be wrong, please correct me). The Quad ESL on the other hand, appears to be a special case where the intention is apparently to create a "ripples on the pond" type waves in the membrane. According to the Wiki article:
But I'm not sure what the implications of that operation mode is for dispersion characteristics.
Eric
Damn, I've been using the wrong ply all this time! But in retrospect it's obvious you would use bitch for the Tall Blondes. (sorry, couldn't help myself)
Eric
Burnt,
Glad we have electrostatics experience here! Is the usual design (except the Quad) intended to produce a "pistonic" (i.e. everywhere in-phase) motion of the membrane? Or something else entirely?
Eric
I could not get to sleep as I was thinking of a cone , with either cuts made to the cone (to the right of the picture) which would make the cone look like my drawing from the side.
Then I thought instead of making cuts ,what if you just made slits in the cone (on the left of the picture) making sure the sides do not touch.
Waves could move along the cone from the coil to the roll surround,to be absorbed, (bendingwaves,) not across the whole cone surface.
Would this cone still be able to keep enough of its pistonic pressure to still produce low frequencies?
Would light foam strips in-between the slits help damp the panel segments and increase pistonic pressure?
It would be very easy to convert a pistonic cone speaker, it's a pity I just threw out a couple of bass drivers which I could have cut up🙄
The cone in the picture was from a unit I used to make an exciter.
Depending on the size of the cuts or slits you could have up to something like 10 rubanoide foils driven from 1 coil ?
I have no idea if anyone else has ever tried or even thought of this ,if so I apologise.
If not, bang goes another patent 😱😭😫
Steve.
Then I thought instead of making cuts ,what if you just made slits in the cone (on the left of the picture) making sure the sides do not touch.
Waves could move along the cone from the coil to the roll surround,to be absorbed, (bendingwaves,) not across the whole cone surface.
Would this cone still be able to keep enough of its pistonic pressure to still produce low frequencies?
Would light foam strips in-between the slits help damp the panel segments and increase pistonic pressure?
It would be very easy to convert a pistonic cone speaker, it's a pity I just threw out a couple of bass drivers which I could have cut up🙄
The cone in the picture was from a unit I used to make an exciter.
Depending on the size of the cuts or slits you could have up to something like 10 rubanoide foils driven from 1 coil ?
I have no idea if anyone else has ever tried or even thought of this ,if so I apologise.
If not, bang goes another patent 😱😭😫
Steve.
Attachments
Hi Eric,
The Quads are intended to be pistonic, you can’t avoid that as the entire surface is driven evenly assuming a constant charge design like the quads. However, the later quads introduced the concept of delay lines so that the time a signal reached a section of the diaphragm could be delayed. The electrodes were arranged in concentric circles in an attempt to simulate a point source. The centre section received the first arrival of the musical signal then the first ring received a delayed signal then the third concentric section received a slightly later signal. The design intent was to simulate a pulsating sphere wave front as seen by a plane set 1 metre in front of the theoretical point source. All this was done using analogue methods rather than digital. Peter Walker was a bit of a genius.
Burnt
Last edited:
I pertained with the idea that the cone surface is DM surface (I got a lot of resistance to that idea). Whether the pistonic motion makes physical waves on the cone surface, or whether the vibrations are transferred from the bobbin surface through the bobbin top to the bottom of the cone surface has to be determined first.
If it is just pistonic motion, the slit pieces would flap, and the vibrations coming from the voice coil trough the bobbin would just give distorted sounds.
In a given second (or millisecond) how many sounds you hear listening to music coming from a speaker? 10s, 100s or 1000s? When a piston moves, there's only one motion at a given time in one direction. There won't be 10, 20 different movements in the same direction.
Voice coil is the speaker, not the bobbin, not the cone that's connected to it. Sound travels many times faster in solid material than in air. The 'pistonic motion' we see is very much slower than the sound that had travelled through the bobbin surface and through the paper. Speed of sound in plastic is ~1070m/s, and in paper ~3000m/s. Our eyes cannot register that speed at close quarters.
Look at the exciters we have. The voice coil is restricted of any motion. Paste the exciter to a table and you have music.