For the active membrane of an ESL ... is there an overall better alternative to Mylar? Therefore, any approach to tame the unwanted artefacts emerging from this type of film is welcome. My approach would be first to heoretically test for the effectness of each option in isolation from other influences. So a testbed such as the frame described in this test might be helpful also to test any damping measure. Instead, the complex structure of a perforated stator, partially meshed sandwich might have a masking effect. And then for the gut feeling? Every gastroenterologic specialist can tell you that guts have the potential to betray. And most people might tell from their own experience.
The dust cover films instead, as intended to be tested in this thread, have other properties to match than the active membrane. So obviously besides Mylar many other materials are used for this purpose. I am really eager to test my bunch of rolls by the time I will be ready to do so.
My delay has a simple cause: I have only a limited (=precious) quantity of film, and therefore I want to have the film sheets confectionned in a way that they can serve several times on the test frame, without being destroyed by having to be "torn" off. Furthermore, changing the films has to be quick, easy and reversible. This is why I want to first figure out the best way how to handle these different materials and tasks. An example: In order to test for the material properties of each sheet make in isolation from other influences, I want to hold the film in place and tension it by the four corners only for a first test series. All four borders of the sheet should be freely floating in ambient air, while a vast aera in the center should be reasonably and evenly tensionned. Tensionned from test to test always by more or less at the same, but also variable forces. So no taping to any frame. I want to avoid uncontrolled damping (which would be part of the "therapy"). No standing waves and the like. To figure out how to archieve these goals now turns out to be a more delicate process of trial and error than I imagined before. I e.g. already ripped twice a corner piece away from the Mylar 4.2 sheet while probing tensioning with different approaches to anchor the four stretcher strings into the Mylar sheet corner. But things seem to slowly settle in direction of a useable setup, in the end. All I have to do is to be patient also to myself ...
The dust cover films instead, as intended to be tested in this thread, have other properties to match than the active membrane. So obviously besides Mylar many other materials are used for this purpose. I am really eager to test my bunch of rolls by the time I will be ready to do so.
My delay has a simple cause: I have only a limited (=precious) quantity of film, and therefore I want to have the film sheets confectionned in a way that they can serve several times on the test frame, without being destroyed by having to be "torn" off. Furthermore, changing the films has to be quick, easy and reversible. This is why I want to first figure out the best way how to handle these different materials and tasks. An example: In order to test for the material properties of each sheet make in isolation from other influences, I want to hold the film in place and tension it by the four corners only for a first test series. All four borders of the sheet should be freely floating in ambient air, while a vast aera in the center should be reasonably and evenly tensionned. Tensionned from test to test always by more or less at the same, but also variable forces. So no taping to any frame. I want to avoid uncontrolled damping (which would be part of the "therapy"). No standing waves and the like. To figure out how to archieve these goals now turns out to be a more delicate process of trial and error than I imagined before. I e.g. already ripped twice a corner piece away from the Mylar 4.2 sheet while probing tensioning with different approaches to anchor the four stretcher strings into the Mylar sheet corner. But things seem to slowly settle in direction of a useable setup, in the end. All I have to do is to be patient also to myself ...
While tweaking a method to anchor the fixing/tensioning strings at the four corners of the film, I made a set of quick and dirty measurements with the Mylar 4.2 C film under different tensions.
On my actual iteration of the test frame, the film gets hold in place and eventually also tensionned crosswise diagonally, and the the frame allows for different amounts of tensioning. The tension is generated simply by adding the weight of a 1.5l PET bottle weighting either 500g or 1600g onto each of both of the diagonal tensioning string loops. If tension is applied, then a seemingly near uniformly stretched aera in the center region of the film gets built up.
Red: No Mylar film
Green: Mylar 4.2 fitted, relaxed and nearly without any diagonal tension
Brown (hidden by and as same as Blue): Diagonal tensioning with 500g weight
Blue (hiding Brown): Diagonal tensisoning with 1600g weight ... and inconsistently labeled, should be labeled M42_16 instead of M42_15
These are preliminary and partial results only. E.g. I did not yet assess the basic resonance frequencies of the films at the different tensions. Anyway, in terms of the 1ms artefact the result shows that it seems to play practically no role wheter the film is relaxed or tensioned. Furthermore, the two different tensioning forces show exactly the same artefacts at 1ms (f_res will be different).
What does this mean now? E.g. there is this the nice little story/idea/concept that a dust protective film will quasi compliantly move along with the sound waves, quasi surfing on top of the wave. And therefore a dust protective film might have no sonic impact. At least with this Mylar 4.2 C this is not true. Even the completely relaxed film shows a strong artefact which most probably will correspond to a reflection. And in this very measurement series, the relaxed film's artefact amplitude is even greater that the one when the film got tensionned. Aha.
There are two other aspects on this graph worth pointing to: The green graph's 1ms artefact peaks a bit before the brown/blue artefact. This is because while relaxed, the film was bending with a small convexity toward the mic. The other goodie is the fact that naked red 2ms reflex shows a higher amplitude than the green, brown and blue one. Logically: The reflected energy by the film will no more reach the reflecting plane at the back of the film.
It seems to become a really fun playfield ...
Last edited:
Hi,
what do You mean by ´unwanted artefacts´? As a quite soft plastic the internal damping of Mylar is already good.
Almost all -if not all- artefacts we see in masurement are due to the Mylar sheet beeing fixed and beeing stretched tight.
Which leads to a pronounced base resonance and hash from boundary reflections.
Those are afaik by far dominant.
What else do You suppose to see?
If You reduce for example the tension to almost nil, the base resonance vanishes.
If You reduce the hard clamping at the boundary fixation, You can reduce the hash from reflections.
In theory distributed damping -think of a sandwiched membrane between two layers of foam- should deal with both issues.
This could render the need for stretching and clamping the diaphragm obsolete .... thereby also omitting with the hassle to mount and glue the membrane.
In praxis it´ll be difficult though to find a foamy material with just so high compressability that the forces acting on the membrane and resisting their movement are just enough for damping the hash.
I one tested this approach.
And while the base resonance completely vanished, the foam layers I used almost killed the useful output due their to high stiffness.
I´m still looking for a suitable foam material or an alternative, as in theory there could be some more interisting advantages for a non-stretched non-fixed membrane ESL.
jauu
Calvin
what do You mean by ´unwanted artefacts´? As a quite soft plastic the internal damping of Mylar is already good.
Almost all -if not all- artefacts we see in masurement are due to the Mylar sheet beeing fixed and beeing stretched tight.
Which leads to a pronounced base resonance and hash from boundary reflections.
Those are afaik by far dominant.
What else do You suppose to see?
If You reduce for example the tension to almost nil, the base resonance vanishes.
If You reduce the hard clamping at the boundary fixation, You can reduce the hash from reflections.
In theory distributed damping -think of a sandwiched membrane between two layers of foam- should deal with both issues.
This could render the need for stretching and clamping the diaphragm obsolete .... thereby also omitting with the hassle to mount and glue the membrane.
In praxis it´ll be difficult though to find a foamy material with just so high compressability that the forces acting on the membrane and resisting their movement are just enough for damping the hash.
I one tested this approach.
And while the base resonance completely vanished, the foam layers I used almost killed the useful output due their to high stiffness.
I´m still looking for a suitable foam material or an alternative, as in theory there could be some more interisting advantages for a non-stretched non-fixed membrane ESL.
jauu
Calvin
Am I right in understanding that the double socks (film + fabric) are stretched over the speakers to prevent dust from getting in and that's all, that is, these socks seal the main driver and create compression inside it? If so, what does measuring reflections from the stretched film on "clothes pins" like sheets in the wind have to do with all this?
This is a good point. I plan to roughly and comparatively test the different films for intermal damping (="Eigensound") also.
This goes to the electrostatically driven membrane. But beware: In my preliminary test, also a loose hanging, unstretched film shows a distinct 1ms peak. See graph 2 in my former post.
Any new an improved constructive variant would be most welcome. And in your example, the foam would naturally act als a dust protecion for the membrane also.
Yes. Notice that in a Quad ESL the films are fixed as two separate sheets and are not configured as a sock.
Yes, but only slightly. Due to the (mass) inertia of the films.
The "problem", if we d'like keep on speaking of problems, is the fact that you are slightly too fast.
These already published graphs stem from preliminary tests, as said, intended to set up and refine the testbed for a series of "real" comparative tests for different films later on. And it might be that by comparing then the properties of all the film brands and types, that a rank list from best to worst might emerge from the tests.
I published these preliminary tests in the hope to get some feedback/ideas for constructive details and/or for further variants of testing all these film samples. So for me, this part of the whole story still belongs to the brainstorming/construtive tweaking phase.
I hope my tests in the end will help to better understand the physical properties and the best practice uses for different films. That's all. For dust protecion as an example. We all know all the relevant properties of titanium, aluminium, magnesium, beryllium for use in our tweeters. How much then do we really know about the physical and acoustic properties of different films variants?
As said, in the test series I also plan to test for the internal damping of the films. You also may say test for the "Eigensound" of the different films. For this purpose, I will mechanically excite a loosely fixed, non tensionned film and check it's acoustic response. I expect that a highly internally damped film then will produce less acoustic response than a nearly non-damped film. Imagine the noise of a mechanically stimulated towel vs. a bit of mylar sheet vs. a foil of houshold aluminium wrapping sheet and guess which one will be the loudest.
So may I present the actual iteration of my test frame, newly set up and completed for this damping/Eigensound test:
On these picture you can see the suspension mechanism fixing the 4.2um Mylar C film in place by the upper corner fixing points only. The lower corners are not connected to the tensioning string system, and logically there are also no weights tensioning the strings. Therefore and apart the uppermost few centimeters of the mylar, the film hangs loose, completely untensionned.
You also can see an oblique metal tube centrally fixed at the back of the frame. And at the frame's upper corner, you see a neodym magnet with two steel spheres attracted to it. Now imagine taking one of these steel spheres and introduce it into the upper end of the oblique tube. The steel sphere will roll/fall through the tube with a forward momentum, then leave it at it's lower end, fly in forward free fall down and at the same time against the film and then still falling loose it's forward momentum while mechanically scratching/displacing the centerline of the film on it's way. By the time the film ends, the steel sphere will be in free fall again and finally hit the floor.
While the steel spere is falling/rolling through the tube, this will make some noise. This noise is gating the recording by the microphone. The recording is performed by ARTA which allows noise gated recording on external stimuli. The recorded response then can be exported as a 32-bit wav.
Red is the response of the whole recording section. More selectively,
Please note: No interpretation by now! Later on, in the comparative testing phase I expect the different films to produce different responses. The test setup hopefully should ensure near-constant conditions from test to test, with quasi equal exciting energies.
So may I present the actual iteration of my test frame, newly set up and completed for this damping/Eigensound test:
On these picture you can see the suspension mechanism fixing the 4.2um Mylar C film in place by the upper corner fixing points only. The lower corners are not connected to the tensioning string system, and logically there are also no weights tensioning the strings. Therefore and apart the uppermost few centimeters of the mylar, the film hangs loose, completely untensionned.
You also can see an oblique metal tube centrally fixed at the back of the frame. And at the frame's upper corner, you see a neodym magnet with two steel spheres attracted to it. Now imagine taking one of these steel spheres and introduce it into the upper end of the oblique tube. The steel sphere will roll/fall through the tube with a forward momentum, then leave it at it's lower end, fly in forward free fall down and at the same time against the film and then still falling loose it's forward momentum while mechanically scratching/displacing the centerline of the film on it's way. By the time the film ends, the steel sphere will be in free fall again and finally hit the floor.
While the steel spere is falling/rolling through the tube, this will make some noise. This noise is gating the recording by the microphone. The recording is performed by ARTA which allows noise gated recording on external stimuli. The recorded response then can be exported as a 32-bit wav.
Red is the response of the whole recording section. More selectively,
- Green is the noise of the steel sphere in the metal tube, then a short free fall with ambient noise only, until the spere hits the film
- Brown is the sequence when the steel sphere scratches/jams against the film
- Blue is free fall again. Any recorded sound is ambient noise during this phase
- Grey is the hitting of the steel sphere onto the carpet and then ambient noise again.
Please note: No interpretation by now! Later on, in the comparative testing phase I expect the different films to produce different responses. The test setup hopefully should ensure near-constant conditions from test to test, with quasi equal exciting energies.
Last edited:
It seems to me that the question should be put as follows: how many drivers in percentage terms fail due to the influence of dust, for this we take the statistics of the failure of the Martin Logan (without socks) and
The next stage of progress will be the influence of the protective sock as a compression element in the driver design, and here even without measurements it can be calculated that the greater the mass and the less compliance it has, the greater the compression will be in low frequencies.
I will not stop asserting that electrostatics are a very, very weak motor in low frequencies, by artificially creating compression elements on its path we reduce the output of low frequencies, this is an axiom.
It's another matter if
Then, superimposed, these now two resonances smear, smooth out the edges of one of them and as a result, the frequency response in low frequencies becomes smoother in a certain room.
The same Martin Logan in his full-range electrostatics in the bass panel uses the division of the common driver into four different in size, and therefore in resonance, and thus smears one common one, that is, the entire bass panel.
Quad
(with a sock) drivers.The next stage of progress will be the influence of the protective sock as a compression element in the driver design, and here even without measurements it can be calculated that the greater the mass and the less compliance it has, the greater the compression will be in low frequencies.
I will not stop asserting that electrostatics are a very, very weak motor in low frequencies, by artificially creating compression elements on its path we reduce the output of low frequencies, this is an axiom.
It's another matter if
Quad
uses this sock for acoustic purposes, that is, creating another resonance in addition to the main one created by the working membrane.Then, superimposed, these now two resonances smear, smooth out the edges of one of them and as a result, the frequency response in low frequencies becomes smoother in a certain room.
The same Martin Logan in his full-range electrostatics in the bass panel uses the division of the common driver into four different in size, and therefore in resonance, and thus smears one common one, that is, the entire bass panel.
Last edited:
This is too complicated a question, I asked a simpler question - is it true that the covers stretched over these speakers serve as protection from dust, maybe they are for other purposes😀?
The dust covers are single side panels that are clip fixed to surrounding rails on all four sides. Plastic rails top and bottom, aluminium rails on the sides.
So the dust panel Mylar is fixed all around the four panels!
Is it protecting from dust? Yes, it is. Although the speaker is not 100% closed like the individual panels of the original Quad ESL model.
Is there another purpose? Maybe, maybe not.
Red = round plastic or aluminium profile
Black = plastic clip-on strip
Yellow = Mylar
The latter eventually can be a source for failure.
Largest part of failure of Quad ESL-63 and later is due to stator glue problems.
The stretching is different and the size is different of both inner membrane panel and dust cover panel.
Quad ESL-63 and later models only have two different size bass panels.
So the dust panel Mylar is fixed all around the four panels!
Is it protecting from dust? Yes, it is. Although the speaker is not 100% closed like the individual panels of the original Quad ESL model.
Is there another purpose? Maybe, maybe not.
Red = round plastic or aluminium profile
Black = plastic clip-on strip
Yellow = Mylar
It is not direct failing, more degrading in sound and higher risk of arcing due to dirt building up on the charged Mylar.
The latter eventually can be a source for failure.
Largest part of failure of Quad ESL-63 and later is due to stator glue problems.
The resonance of the panel membrane is different from the resonance of the dust cover panel.
The stretching is different and the size is different of both inner membrane panel and dust cover panel.
Quad ESL-63 and later models only have two different size bass panels.
- On small size segment on the edges of each ring panel, two of these on each panel, total 4 in each speaker. Surface size 120 cm2 each, 480 cm2 per speaker
- The full size bass panels, surface size 910 cm2 each. Depending on the model (4 or six panels) you have two or four of these in each speaker.
Attachments
14 years ago, when I got into electrostatics, I took the full-range Martin Logan as a basis to make my own, only larger in size, based on it.
There, the bass panel is implemented as follows: two isobaric membranes and three stators, one large membrane, but separated by insulators in such a way that the 4 segments were of different physical sizes and, accordingly, each segment had its own main resonance, the audible frequency range that was fed to them from the step-up transformers was 30-1000 Hz.
This solid low-frequency emitter with four segments did not have one pronounced resonance, and therefore no harmonics emanating from it, but in my modern low-frequency emitter there is such a pronounced resonance, therefore there is only one segment, but not four like in the old one.
Therefore, I admit that the protective sock, also known as a cover, also known as a film, also known as Mylar, can serve as such an additional segment of its own resonance that partially destroys the resonance of the working, active membrane, and if they are mounted separately, then the front cover can be set up with one resonance, and the back cover with another. But this is one positive side of the matter, while the negative side is the compression role of these protective films (socks, covers), which will necessarily smooth out the peaks of the useful, musical signal, in short, they will cut off the dynamic range in the lower part of the register.
Attachments
A fully assembled Quad63 sandwich is a quite complex acoustical structure, and certainly hard to model. Therefore maybe it could be interesting to specifically real-life test (please by measurements ...) all these assumptions and theories, especially the one about dynamic loss for low frequencies on a dust protected Quad63 vs. a non-dust protected Quad63. It will not be me testing this one. As for me, I will test single and isolated films very basic acoustical behaviors only.
Ahhh - that's the confusion!
I always considered the dust cover the, ehh, dust cover, build from film, pressed on the frame as described by wout31 above.
While the sock would be the cloth that is wrapped around the speaker. Definitely not film.
Two different things.
Jan
So it's good that this is happening on this forum, in other words, it's a search for truth.
And so, why are passive films installed on speakers?
After all, before measuring something, we must know what exactly to measure and for what purpose.
And so, why are passive films installed on speakers?
After all, before measuring something, we must know what exactly to measure and for what purpose.
As said, I plan to open another thread to publish data about materials used and found in speakers and in other acoustic constructs. I plan to test these materials for different properties such as reflection, transmission, internal damping and the like. Including films used in ESL's of course. The results might be presented like this following one, as an example for induced noise within films and sheets of different materials like aluminium (red), a sheet of baking paper (green), Mylar 4.2um (blue) and Vliesseline Stickflies (black). In this case noise was induced by a normalized stimulus from a 6mm diameter steel sphere falling laterally onto the non-tensioned films/sheets.
Upper graph is the induced impulse response, lower graph shows the frequency sprectrum of the induced noise. I think these graphs are self-explanatory, so I will not comment further on.
Upper graph is the induced impulse response, lower graph shows the frequency sprectrum of the induced noise. I think these graphs are self-explanatory, so I will not comment further on.
Last edited:
Do as you like and thank you, this is a kind idea.
I anyway will start a new thread without the terms "Quad63" and "dust protection" in it's title. The new thread should be about materials properties and consecutive specific suitablilty for audio use. Bare materials basics, a kind of iterative materials functional assessment. Nothing else.
Think e.g. about a prototype tensioned film and rods based "harmonica", lightweight combined 1D diffusor/absorber? You will be quite happy to know which materials you may choose to start with. You certaily will not choose a film that has a very low reflections-to-transmission ratio ... On the contrary, if you would like to hide your speakers behind a curtain, then you may choose a product with the opposite properties. And by the way, of course you may also make a more educated choice for a dust protection by better knowing your candidate's material's properties.
See what I mean with the planned new thread?
I anyway will start a new thread without the terms "Quad63" and "dust protection" in it's title. The new thread should be about materials properties and consecutive specific suitablilty for audio use. Bare materials basics, a kind of iterative materials functional assessment. Nothing else.
Think e.g. about a prototype tensioned film and rods based "harmonica", lightweight combined 1D diffusor/absorber? You will be quite happy to know which materials you may choose to start with. You certaily will not choose a film that has a very low reflections-to-transmission ratio ... On the contrary, if you would like to hide your speakers behind a curtain, then you may choose a product with the opposite properties. And by the way, of course you may also make a more educated choice for a dust protection by better knowing your candidate's material's properties.
See what I mean with the planned new thread?
Last edited: