In fact, that's what they truly are—a party gadget. Aiyima never refers to them as being intended for professional use.
I suppose you've heard of a guy named Oleg Rullit from Munich, who has been making his own handmade speakers? I'll post some images of his speakers in your thread if it might help you create an interesting speaker. I'm quite interested in redesigning a speaker membrane—whether concave, convex, or flat. I'll share some here for inspiration. This is not exactly a sandwich, but rather integrated, balsa and paper, his secret paper.
Suede leather suspension, with felt around the cast metal frame, in front and back (not shown here). If you've watched the Vietnamese videos, you might notice some concepts, such as the round flat wooden panel in a speaker box, wooden spiders at the back, and so on. It's a mixture of various ideas in flat, irregularly cut panels within a box.
Speaker parts before assembly—not a sandwich, but it could serve as inspiration.
Last edited:
@lekha
loudspeaker from scratch - difficult task
but standard paper cones are not bad already and with a little tweaking even standard drivers excel sonically
this 15cm cheapo driver has good linearity and phase response naturally
loudspeaker from scratch - difficult task
but standard paper cones are not bad already and with a little tweaking even standard drivers excel sonically
this 15cm cheapo driver has good linearity and phase response naturally
By the way, this looks like mulberry paper. It seems that Rullit has used that paper, as I read elsewhere. It appears to be resistant to moisture. To create a sound diffuser panel, this paper can be glued together, either one sheet over another until the desired thickness is achieved. Someone else experimented with tearing the paper into pieces and then gluing them back together. Another method involved combing the paper (resulting in a panel) until the fibres stand up.
True, but I am considering it in terms of panels for "DMLs."
Rullit seems to be out of internet circulation at the moment. Some of his former countrymen are not particularly pleased with him, but those who have heard his speakers praise them—at least that's what they say. I've heard that his speakers are quite expensive. In fact, the lack of perimeter suspension is quite intriguing. And that top held the "spider" also made of slit balsa strip.
Even though its shape is conical, it doesn't seem to move much, at least not at the perimeter; it behaves more like a "pushed-in" DML panel.
Last edited:
Hello Sandasnickaren.
All panel materials have a sound of their own.
Wood sounds like wood , eps sounds like eps, acrylic sounds like acrylic, and so on.
This is what I call self noise.
I try to minimise these sounds.
Coating some materials in the Ali foil might help reduce these sounds ?
Distortions and such like is another matter.
Steve.
All panel materials have a sound of their own.
Wood sounds like wood , eps sounds like eps, acrylic sounds like acrylic, and so on.
This is what I call self noise.
I try to minimise these sounds.
Coating some materials in the Ali foil might help reduce these sounds ?
Distortions and such like is another matter.
Steve.
Until a material is caressed, knocked on, scratched, etc., it won't produce any sound by itself. However, since we need that specific material to work with the exciter/transducer, we must find a way to compensate for that. I believe this is why, most of the time, the speaker diffuser material is so thin and the diffuser is so small—larger panels tend to present more problems. So, attempting to use panels the size of half a door is likely asking for trouble.
An experiment done some time in 2021 by someone for inspiration.
Last edited:
No, it would introduce extra noise, depending on the type of foil used. If it is fully glued to create a sandwich, that would produce a different sound altogether. Self-noise can sometimes sound better than the additional sound introduced by the transducer, which has its own internal noise.
Hello,
Here are some measurements from my 3mm poplar plywood panels (45x120cm, full foam suspension, light frame and spine)
First picture below is about 85dB/1m which is unusually high as volume for me, below is about 75dB/1m which more usual.
The level is evaluated from a 1/2octave smoothing which seems increase it a bit (by 2 or 3dB?) compare to what seems to be the mean value on the distortion plot.
I show here only the THD, it is mainly a H2 harmonic (see 4th picture with H2 and H3). I still wonder why it is mainly H2 (H2 means some asymmetry between positive and negative displacement, like a tube or FET transfer function)
The third picture is also about 75dB but here the panel was laying against my previous 3way column (normal position) so I searched a bit to understand the reason of the peak at 400Hz to understand it might be a problem of resonance in the frame (which is for sure too light). So for the 2 first measurements, I was holding the frame to avoid this resonance and also the contact behind with the column loudspeaker.
In addition I tested an USB insulator on the Umik1 mic. The harmonics of the main we can see on the noise floor were a bit reduced (by 5dB?).
All of that to remind that almost anything around the panel can be a source of resonance. In this panel, the frame is for sure much too light.
The distortion measurement with REW is a good test to detect any problem (like the 400Hz peak)
The way I use to locate the problem is to play a pink noise (1/3rd of octave, maybe less) centered at the frequency of the problem.
It is like that, that some weeks ago, I understood I was unable to hang a simple EPS test panel with 2 simple (kitchen) strings... their resonances (noise?) was much too high.
An other anecdote is while doing directivity tests, a panel (30x40cm polystyrene glass with a light frame) was on small table with a wood pin at its basis to define the axis (there is a picture not to far in the posts before). The exciter being of this axis by about 3cm was able to make the panel turning a bit during the test.
Hope it helps.
Christian
About 85dB/1m :
About 75dB/1m
About 75dB/1m, panel in its initial conditions... with some not expected behaviors.
May I ask what the standard distance is at which you listen to your DMLs?
I'm asking this because the standard distance from which I listen to any speakers is between 2.5m to 2.8m. The 1-meter distance is more typical for computer speakers; my monitor is situated about 90cm away from my eyes, so any computer speakers would be positioned behind it, a bit further than 1 meter away. Those computer speakers are meant to produce sound softly and not disturb. In contrast, the living room speakers are intended for leisurely listening.
Even though you are conducting quite good tests at a 1-meter distance, that information isn't particularly useful for a 2.5-meter leisure listening distance, I'm afraid.
i once owned an elctric car and had exciters inside as the plastic surface of the car is thin and it went down low in bass!
this car is a combination of car and bike for two persons, called Twike
this car is a combination of car and bike for two persons, called Twike
going to have to be a contact adhesive water based will well and probably the only choice on foam though with a absorbent substrate (such as wood) you could use a solvent or water based adhesive but setting time will be much longer.
Hello Lekha,
Yes you can.
Those measurements were done at 1m from the panel which is a 45x120cm 3mm plywood. The listening distant for them in our living room is 2.80m, maybe a bit more.
The choice of the measuring distance depends on what we want to focus on.
The most important is to be in what it is called the far field so that all the possible sources from the loudspeaker are combined in a way similar to what happens at the listening distance. In sources, are included the possible different ways (multi way speaker) but also the edges of the cabinet which are supposed to act as secondary sources. A rule of thumb I have in mind is that distance is 2.5 time the largest dimension of the speaker.
Unfortunately, most of our measurements are done in a room that adds its reflections.
So there is a compromise to find between a representative FR (far field) and the reflections entering too early in the impulse response.
For the directivity test shown before for example, the distance is 1m in order to get most of the information coming from the speaker, not from the room that reflect sound from any direction.
For the distortion test here, I think the distance is not that important, the goal being to evaluate the level of the distortion (30, 40..50dB below the signal) and to detect possible problems (ie peak at 400Hz)
In room audio measurement is not easy. The most important is to keep in mind the limitations. What do we want to see? What do we collect?
Christian
I believe we are merely deceiving ourselves with that 1-meter distance testing. What one hears at a 2.5-meter distance is not the same as at 1 meter, especially if we are not aiming to create computer speakers.
It should, after all, as we live in rooms filled with furniture. The whole idea of creating a speaker for oneself is to design one that works harmoniously with the room in which it will be placed.
Should I have written : "I think in this case the distance is not that important, the goal being to evaluate the level of the distortion" ?
Nor do I propose to conclude from the FR more than an average level of SPL.
The last sentence was in this idea : "In room audio measurement is not easy. The most important is to keep in mind the limitations. What do we want to see? What do we collect?"
The goal of a measurement is not necessarily to have a full view.
It is known that even if the standard distance is 1m, the real measurement distance (except with very sophisticated tools like from Klippel?) is often 2m (see the far field)
My experience is in some cases, it is better to not mix the "room sound" too much with the direct sound.
This depend also how the room is reverberant or not... 2.5m might be above the critical distance (distance where direct and reverberant sound are equal)
Anyway feel free to measure at 2.5m. ll you are more confident with that... some results will be probably just less easy to read... not to say impossible (ie coincidence frequency detection).
Christian
The reality is that we can't change the room, so whatever we create must work with our own space, not someone else's. Considering that the sound from flat panels is often nicer than that from cones, for example, we can actually enclose it in a box, which would allow us to position it as we desire. I've been listening to his speakers a few times today, and I will likely do so a few more times in the future until I decide which box (and drivers) to dismantle, whether to try wood cones or wood panels. I've been searching for ideas on wood cones, and that's how I came across his videos. However, placing a flat round panel on the cone speaker is much easier to attempt than making wood cones.
lekha,
Am I reading that correctly? That you are going to build something of your own?!
Eric
Has anyone tried something like this? It is two layers of "sign grade" aluminum sheet bonded to a polyethylene core. Total thickness is 3mm, in the ballpark for the more-rigid materials. A 24" x 24" panel weighs 4 lbs, probably on the high side compared to 3mm plywood.
The core material might provide some damping but that remains to be seen.
Given the mass it might make a decent "subwoofer" panel. But, since I haven't completed even ONE DML what do I know 😉. Even after going through all 661 pages.....which took awhile!
I have some experience with stuff like this but in a completely different context. I tried using some in a high vacuum environment (a scanning electron microscope), as a sample holder for large samples like 8 inch silicon wafers -- but the plastic core outgassed so much that the SEM wasn't able to pump the sample chamber down to the required operating pressure. It must have a certain amount of permeability, which might improve its damping factor. But......"what do I know?"....
Mark
The core material might provide some damping but that remains to be seen.
Given the mass it might make a decent "subwoofer" panel. But, since I haven't completed even ONE DML what do I know 😉. Even after going through all 661 pages.....which took awhile!
I have some experience with stuff like this but in a completely different context. I tried using some in a high vacuum environment (a scanning electron microscope), as a sample holder for large samples like 8 inch silicon wafers -- but the plastic core outgassed so much that the SEM wasn't able to pump the sample chamber down to the required operating pressure. It must have a certain amount of permeability, which might improve its damping factor. But......"what do I know?"....
Mark
BTW, the same supplier I referenced above has a similar product, but it is just 2mm thick. It weighs a little more than half the 3mm stuff.
The foam is probably very soft, that might be a problem (too much damping and lack of HF). The paper poster boards have this same problem. But maybe this will work because of the aluminum skins? I'interested to hear if you try it.
Hans