| Other ESL build with simple materials available from local grocery stores - Click HERE for Original Thread |
| Audio_idiot |
Hi,
This is my virgin thread, I hope I didn't step on other threads on this subject.
I am presently collecting materials for my ESL. I wish to experiment with a small prototype before proceeding into fullrange and high SPL.
I had many restraint in allocating commonly suggested materials for ESL, especially bulky materials which will cost both me arms and legs to ship.
This is how far I had progressed,
The film:- there are no other alternative other then mail order some 6um mylar film,
Insulator:- Plexi glass seems to be the default material for insulator in my location as many thickness can be obtained easily,
##PS, small items shipping cost is ok.
Transformers:- The bias supply can be made easily using common EI step down transformer, the matching OPT needs more thought but can be arrange.
Stator:- Perforated plates are not off-the-shelf item in our local hardware store. Wire stators seems extreamly labour intensive and quality solid core copper wires are rare commodity.
I happened to come across some fine mesh metal mosquito nets and other stainless steel wire mesh for filter.
Question:
I wonder if we can fine mesh metal wire nets as stator conductor by glueing them onto the lighting grates, egg crates or other form of support for mechanical strength?
Cheers |
|
|
| SY |
| quote: | | I wonder if we can fine mesh metal wire nets as stator conductor by glueing them onto the lighting grates, egg crates or other form of support for mechanical strength? |
That can work well- but don't use a very fine mesh, since the percent open area can be low. Normal window screen size would be a good choice. |
|
|
| bigwill |
| I'm also trying to build a test ESL panel out of easily avaliable stuff. Will cling-film or 'saran-wrap' work for the diaphragm? |
|
|
| SY |
| Generally, they're much too thick. You want something about 6-10u. (1/4-1/2 mil in British units) |
|
|
| medum |
Have a look at this nice australian site:
www.eraudio.com.au
You can buy all kind of parts for your ESL speaker!
ROB MACKINLAY can give advice, too!
Jørn S. |
|
|
| BillH |
Hi, Audio_idiot.
The wire screen is an interesting product for stators and I haven't seen it used before. You will not have to deal with the problem of sharp edges like perforated material, but I wonder if all the wire ends on the screen will give you a leakage problem.
As you're finding out, getting materials for an ESL is one of the biggest challenges a DIYer faces. I had to order parts from three different places just to build new HV bias supplies.
| quote: | | Bigwill writes: Will cling-film or 'saran-wrap' work for the diaphragm? |
In addition to what SY wrote, the cling film seems too 'soft', especially if you're going to use graphite on it. I've got a package of window insulating film that I bought for the double sided tape inside. The film measures 0.005" thick, about 12 micron and is heat shrinkable. It's a bit thick, but if it's available in your area, it may be worth looking into.
Audio_idiot, you'll probably find that your ESL design will evolve as you build according to the materials available. Please let us know how it works out. |
|
|
| SY |
Bill, the Saran films from grocery stores these days don't even have Saran in them!
FWIW, I use a Saran film (real Saran, PVDC) which works very well, but film that thin is not trivial to obtain. |
|
|
| bigwill |
| Hmmm, I just tried to heat-shrink some cling film and it worked nicely, and it seemed to be pretty strong and robust. You could poke it about a lot and simply apply heat again to get rid of the crinkles. You had to stretch it a lot to make it unable to go back to its original shape leaving permanent creases. I think it would be strong enough to apply graphite to, and any "dents" you put in it whilest doing so can just be heated again. The stuff felt light enough to produce high frequencies, but that's an entirely intuitive feeling not an actual fact.. |
|
|
| Bazukaz |
Hi,
If you want to experiment , a thin baking film may be easy to get , and usually comes in around 12 um thick. A black shoe wax may work for coating :).Shoe wax is soluble in water , so it is easy to apply a thin layer of it.
Wire mesh is possible to use for stators , however , it might be hard to insulate it well , like perforated steel. You will have to glue it on support structure , which will not be that easy.
I'd suggest the wire method. It is not the cheapest , and consumes lots of time , but you can build a quality panel. PVC insulated or magnet wire is not difficult to get.To stretch the wires succesfully , you would need a jig. It can be built from threaded steel rods on a wooden frame.The louvers are placed between the rods.The wire is then wound on both stators at once. Then , the wire is additionally stretched. The next (and probably the most time consuming) step is glueing the wires.
Trying to build ESL in a fast way is very likely to result in failure.
Regards,
Lukas. |
|
|
| bigwill |
| Anyone know where you can get mylar in the UK? Or even cheap perforated metal for that matter? |
|
|
| Audio_idiot |
Hi,
Thanks for the great responds, I wouldn't mind trying out on Cling film for a start, I feel that it's much more importnt to get the engineering correct before we venture into more ideal materials. Shoe wax on cling film is most attractive! Love to see the face of my friends when I told them what the ESL are made from!!! :-)
I looking into combination of wire and perforated plate stator construction...
I recon we can stratch/tension the mesh using the bicycle tube stretch table and the glue the mesh on the louvre. Then apply the glue over the whole mesh with a roller brush (just like painting over em) so that the glue will double as insulator as well. Do you think this is workable?
As for the openings, I read somewhere in this forum that there are no fix and fast rule on %tage of the opening... thus, I think anything which i can blow a breath through without much restriction is goo enough... like i say, all depand on what I can get at my local hardware store. I also feel that finer wires of the wiremash cross knitted can provide a much even/stronger static field the wire or perforated stator...
BillH, I feel that the leakage problem will be worse in wire mash as the ends as all pointy sharp thin wires which is the most coducive environment for leakage to occur... but can we some how close that out???
Another question on the wire stator, will the louvre stay static after power down?? and give a unsuspected shock to anyone toucing em? as plastics can store some static charges...
Cheers
AI/ |
|
|
| Gustava |
You can make do without fancy jigs to string your wire on polystryrene louvres from flourescent lights. Just use small cable ties at each end to hold each loop of wire.
I used plexiglass strips for the spacers. You can just see the cable ties behind the spacer at the end of each loop.
The wire is teflon-coated 30 gauge wire-wrap.
After stretching the wire tight I held it to the louvre with dots of epoxy. Once the epoxy was dry I cut cut the cable ties off. |
|
|
| Gustava |
I have a whole roll of brass screen that I haven't got round to trying yet. I'm not worried about leakage because the screen is on the outside of the spacers, away from the membrane.
Not to worry about residual charge on the membrane. The resistance is about 100k ohms and the voltage bleeds off over 4-5 minutes.
Cheers. |
|
|
| Audio_idiot |
Hi Gustave,
Thanks for the respond, with 30swg wire wrap, the must be a lot of wire wrap!!! may I know what sort of spacing are you using?
Do you mean that if the looss ends are outside of the membrane area, leakages is not a concern anymore?
Cheers |
|
|
| SY |
| Gustava, I cheated and used stators from Acoustats, but I'd love to build more. How did you do your wire tensioning, just by hand or with a jig? |
|
|
| Audio_idiot |
Hi guvasta,
The spacing between wires looks very far apart... Helps with speed of construction.
What do you mean when you said "each loop of wire" , I was in the impression that the wire stator is made out of a single length of wire.
Thanks for the pix, looks much easier then I thought.
Cheers |
|
|
| Mutley666 |
I have found some mylar available on ebay at a reasonable price, and the shipping is just $5 to the EU. Would be a step up from cling film. ;)
mylar |
|
|
| MJ Dijkstra |
| it is not mylar (r), but hostaphan |
|
|
| Mutley666 |
| Thats the German/Dutch equivalent is it not? Is it not produced under licence from Dupont? |
|
|
| MJ Dijkstra |
Mylar is a registered trademark of Dupont only.
Hostaphan is a registered trademark of Hoechst only.
Mylar and Hostaphan both are polyester films. There are several different subtypes of both species. Some are even custom made. |
|
|
| tade |
lay the window screen down onto the eggcrate and then at every cross in the eggcrate use your hot soldering iron to force the screen into the plastic. Works like a charm and is easy but labor intensive.
Where can i source very thin heat shrinkable mylar and an HV bias supply? Will that aussie site ship to the US? That 3.5 micron mylar looks tempting.
I built my ESLs so long ago i have forgotten where that info is.
Thanks |
|
|
| medum |
| quote: | Originally posted by tade
Will that aussie site ship to the US? That 3.5 micron mylar looks tempting.
|
Guess so - Rob shipped to me in Denmark! |
|
|
| Calvin |
Hi,
sorry to disillusionize a bit, but You won´t be able to build a very good ESL panel just with material You get at the grocery store ;) At least not in German DIY-Stores -and we Germans know DIY because we spend a lot of our lifetime at those places :D
It is quite easy to get the material for wire stators. PVC coated wire, plastics for spacers and frames, Aluminium bars for frame construction, glues, wood for mounting frames, paints, silicone, screws and nails and some more interesting things, but You won´t find a good diaphragm material and it will be hard to find good material for other kinds of stators too. Since Wire stators of the kind of Audiostatics are very cheap from a viewpoint of costs, I´d recommend to build a smaller wire stator ESl first.
jauu
Calvin |
|
|
| Audio_idiot |
Hi Calvin,
I agree that we can't have everything from the local shops but no harm to at least a try at it;) ,
Looks like wire stator is the way to go if we insist on local hardware stores... and still get away with quality ESL build.
I had made an order for some the 6uf mylar film through mail.
If we go for the wire stator, what size do you recommand as "small" monitor? and still with ego boosting result?
Cheers
KenC |
|
|
| Calvin |
Hi,
since You have to partner the panel with a bass, I´d choose the width of the panel roghly the same as the bass. The height is Your choice, depending on what You want to reach, but I´d suggest a panel with a ratio W/H of 1/4 or more. A panel of lets say 20x100cm will work from ~200Hz on and ensures enough dynamic headroom. With smaller panels the lower crossover freq rises. I would use a panel with just 10cm width not below 1kHz.
jauu
Calvin |
|
|
| Audio_idiot |
Hi calvin, guys,
Thanks for the suggestion. 200mmx1000m @ 200hz looks managable.
I got the impression that we can avoid lots of hassle if we keep the crossover point outside of congested frequency (i.e. within 100~150hz, below which the low freq starts to become non-directional,..etc etc...)... many speakers gurus recommand that, what is your view.
Does it mean that if I wish to increase the efficiency, I can parallel more then two panel of the same size with the same transformer? 200mm x 1m is about the right size to pair or quadra up in reasonable space...
Another question, has there been many "true" single fullrange ESL working successfully from 60hz to 30khz?? I mean can they play matallica type of music??? or what type of music can be played???
Cheers |
|
|
| Calvin |
Hi,
You might use the panel down to its ground resonance. Something ML has done. The pros are:
- easy crossover design -just a high Q-second order Highpass
- high output and
- largest possible bandwidth
- low crossover point acoustically and much higher electrically!
- use of highest mechanical tension -> stability and efficiency
The cons:
- to my experience -and I´m not alone- the high Q of the resonance (slow decay) is sonically detectable like kind of a ´footprint´
- the distortions are quite high.
- You have to control the mechanical stretching of the diaphragm quite carefully for low tolerances
The pic shows the results with my panel (ML Sequel size) and a comparison to an original ML Prodigy panel
Even though You get close to perfection measurement results it sounds better when the panel is not used down to its ground resonance. Now I use a filter that cuts off @350Hz acoustically.
You can increase efficiency by paralleling panels, Yes, But(!) the impedance curve drops and might reach very low killer levels.
Too the bandwidth of the Trannie-Panel combination sinks. You might not reach 20kHz any more.
If there are series resistances to correct for the freq-response in the HF-region the freq-response will drop even more.
I´d prefer to build a single panel rather than paralleling several smaller ones.
And Your last question: There are a few...and No!
For high SPL purposes build a Hybrid! Much better much easier and more compact!
jauu
Calvin |
|
|
| Audio_idiot |
Calvin,
I have to chew on what you had just said... as I am very new to this :xeye: :xeye:
Do you mean that 200~250mm x >1000mm panel, (Taller the better) @200~300hz is about as far as we can ask for with (compromised) good results and without excessive complications?
BUT I still have to limit the size of the panel within the dimension of "bicycle tyre tube stretching table". Or is there other simple stretcher that I can use?
And I'll just have to increase the tension and bias oltage to raise the SPL.
This may sound stupid but I'll ask anyway... Can we series some panel without ill effects? Since capacitance reduced when connected in series...?
AND do we still need to measure the tension of the diapharm?
Cheers |
|
|
| Calvin |
Hi,
the prime goal in ESL construction is efficiency.
To get high efficiency You can:
- build big panels
- use smallest stator/diaphragm distances (d/s) possible
- use the highest possible voltages
- use high mechanical tension --> high ground resonance --> high dynamic stability
A d/s of ~1mm is still practical to handle and allows for frequencies down to ~200Hz. With this distance the diaphragm should be supported every 70-100mm to not to be drawn into the stators.
Stretching as hard as possible, You can get ground resonances from ~150Hz with a 3µm film to 250Hz with a 12µm film. Stretching hard raises the stability treshold and the efficiency significantely. Having the crossover not lower than one octave above Fs means a crossover point between 300 to 500Hz. When You measure the distortion You will see that they rise -rather shoot up- below 300Hz anyway. Additionally installing the panel in an open baffle You have to account for phase cancellation. So 300Hz is a good starting point for a panel (150Hz Fs) of handeable size. You can still use a passive crossover to equalize the phase cancellation.
One of the biggest mistakes of ESL-Noobs is, that they try to get down in frequency range as low as possible.
But You loose on nearly every important parameter in doing this!
You don´t necessarily need to tension mechanically. You can use heat treatment very well, but You will have to support the diaphragm with spacers closer apart than with high mechanical tension.
You can upsize the tyre stretcher by cutting the tyres and gluing together as many as You need.
Connecting ESLs in series....I have to admit that I never had that idea! :bigeyes: A symmetrical ESL could be considered as two series connected capacitors.
The problem is that the driving signal voltages for each panel would be halved, so efficiency wouldn´t be raised as with parallel connection. Since the capacity of this arrangement is halved too, the impedance the amp sees would be raised. While this could be desirable in the HF-range the impedance of the single panel/tranny is quite high (and doesn´t give the anp any stability problem) in the mid-band. A positive effect could be a rise in the upper bandwidth limit. But You need a higher transformation factor for the needed higher signal voltages which will impose the state of affairs again.
jauu
Calvin |
|
|
| Audio_idiot |
Hi Calvin,
Thanks for the in-depth, More questions
| quote: | | A d/s of ~1mm is still practical to handle and allows for frequencies down to ~200Hz. With this distance the diaphragm should be supported every 70-100mm to not to be drawn into the stators. |
Do you mean support by dividing the panel by strips of 1mm thk insulators of 70~100mm apart laid vertically on both sides of the diapharm? or using "dot" double-sided tape (or "dot" insulators) spread in random at spacing of 70~100mm from each double-sided tape point?
| quote: | | You don´t necessarily need to tension mechanically. You can use heat treatment very well, but You will have to support the diaphragm with spacers closer apart than with high mechanical tension. |
I read from other threads that tension is limited when heat stretch, mechanical cna stretch harder but film will sag over time... can we mechanically stretch the film then heat shrink the film and get the advantage of higher tension and no sagging in the future?
The series connection will have to wait until I build a high output voltage swing power amp before I can try em on...
| quote: | | You can upsize the tyre stretcher by cutting the tyres and gluing together as many as You need. |
Why didn't I think of tha;)
Cheers |
|
|
| Calvin |
Hi,
the supporting points or strips can be made from tape, foam rubber, spacer material or silicone. The last beeing probabely the best because of easy handling and the adhesive and hygrophobic properties.
since this panel is curved You need horizontal supporting spacers
here a flat panel with foam rubber
You can get the highest tension by mechanical stretching. Heat treatment allows for much lower tension, but You might position the supporter points a bit closer to each other instead.
With curved panels there´s only the possibility of mechanical stretching, while flat panels allow both methods. Heat treatment is easy and gives very constant and equal tension and You don´t need to break in the panel.
For the first tries I´d recommend a flat panel with lots of silicon dots or foam rubber placed relatively close (to give a 1:70 ratio) and heat treat the diaphragm.
Heat treatment always means that the film tensions to a specific degree. If it was too lose it will tighten, if it was tensioned to hard it will lose a bit. So in practise I´d tension just a little bit by hand to get all the wrinkles out of the film, glue the stator onto the film and heat treat the thing after drying. Then place it with the film side down on a flat and clean peace of glass and press some silicone dots between stator and film. After drying You can finish the stators and build them together.
jauu
Calvin |
|
|
| Audio_idiot |
Hi Calvin,
Thanks for the pix, now I understand exactly how.:)
So heat shrink the film, 1mm gap between stator and diapharm is good for high efficiency from 200~300hz up? what sort of Bias voltage are we looking? and any indiction of SPL efficiency guild?
BTW, the curve stator is pre-rolled; the diapharm follows the foam contour which follows the stator contour?
A friend found a perforated shop, I think this is good news. I hope there are sufficient veriety to choose from. While we are at it, what spec of perforated plate should I be looking for? i.e. material, %open, thiskness, hole size, holes arrangement pattern? There are too many specs mentioned in too many articles which I'd read, could you please point me the right direction?
Cheer |
|
|
| Calvin |
Hi,
I used standard HV 3/4 sheets. Mans 3mm diameter spaced 4mm apart. The sheet´s thickness should be 1mm or less.
Before coating the sheets were curved.
Looks like this
after fluidized bed coating
and after top coating
Since I didn´t electropolish the sheets a thicker insulation was necessary. The advantage is, that the holes have nicely rounded ´walls´, which prevents bandwidth limitation and the thickness is such that I can touch both stators at full loudness without getting shocked or even worse.
Next step was to glue the spacers on. While I used 1.0mm to the front, the back stator got 1.1mm thick spacers. The difference in spacer thickness is to cope for the always present bowing of the diaphragm.
After that the diaphragm is mounted in the stretching frame, stretched to a very high value of mechanical tension and glued onto the back stator.
jauu
Calvin |
|
|
| Bazukaz |
| quote: | Originally posted by Audio_idiot
Hi Calvin,
Thanks for the pix, now I understand exactly how.:)
So heat shrink the film, 1mm gap between stator and diapharm is good for high efficiency from 200~300hz up? what sort of Bias voltage are we looking? and any indiction of SPL efficiency guild?
BTW, the curve stator is pre-rolled; the diapharm follows the foam contour which follows the stator contour?
A friend found a perforated shop, I think this is good news. I hope there are sufficient veriety to choose from. While we are at it, what spec of perforated plate should I be looking for? i.e. material, %open, thiskness, hole size, holes arrangement pattern? There are too many specs mentioned in too many articles which I'd read, could you please point me the right direction?
Cheer |
Hi,
For 1 mm air gap , the bias voltage should not be very high.For uncoated stators , i'd recommend experimenting in the 1.5 - 2.5 kV range , while well coated stators may work with higher voltages without arcing.Too much bias voltage also tends to "stick" diapraghm to one of stators , if there is not enough tension.
RV 3-4 perforated steel is probably one of the best choices. If you are going to build a flat panel , the minimum steel sheet thickness is around 1mm.It also depends on panel dimensions.For curved panels , 0.7 mm thickness should be ok.
Regards,
Lukas. |
|
|
| Calvin |
Hi,
dammit, I knew I forgot something :smash:
The polarizing voltage is ~1.7kV. This is high enough for good ouput and quiet working .
jauuu
Calvin |
|
|
| Audio_idiot |
Hi guys,
Thnks for sharing your eperience, greatly appricated.
Calvin,
What is fluidise bed coating? I wonder if i can do the coating at home as well?? and Electropolish??? I feel that coating the stator is essensial for the well being and safety of people around more then others... any additional labour/expenses for safety alway justify itself.
Bazokaz,
1.5~2.5Kv should be reasonable easy to build and use safely:)
Could you eleborate on the spec of the RV3-4 perforated steel plate?
Calvin, Bazokaz,
You guys mentioned curved stator one time too many to get me interested:D and the calvin's pix are just irresistable... I understand this is one way to "cure" the directivity characteristic of ESL, Since I am quite confident to get the stators rolled accurately after checking out with local metal fabricators, the devil in me just took over control of my rational senses with ease.... :devilr: Could you kindly teach me how to make curve panels?
Cheers |
|
|
| Calvin |
Hi,
RV3-4 specs are given before. Have a look at suppliers data too.
A standard size of sheet metal is 1.250x 2.500mm.
There is a outer rim of app 10mm that is not perforated.
I had my stator sheets cut in peaces of 1.250mm height and 250mm width, see second pic. (Btw. take care of how the cut is made, there should be no sharp edges or pins left). I have a unperforated part at the top and bottom of my sheets, so there´s a part where the glue for the spacers finds anough surface to hold on to.
Curving the panel is easy...just tell Your metal worker to do it for You :D I used 15°-20°. The reason for curving is not(!) a wider dispersion character in first place, as it is often claimed. The reason to curve is higher stability of the sheet with less resonance and rattle. It is imo the only way to manufacture stable and stiff sheet metal panels without the use of addditional bracing.
Have the sheets curved before coating.
Electropolish means a process where the sheets are put into a bath with acids. Current is led through the metal. The effect is, that sharp edges are etched away, leaving smoothly rounded holes.
The reason for this is, that field strength is high at needle like points or sharp edges. When the sheets are stamped and not drilled, they have one side with sharp edges. So these edges are very prone to flash overs. Besides higher field strength every coating is thin at such edges (data sheets sometimes state the min radius of edges for constant coating thickness). While the coating thickness may be toatally sufficient for Your use on the face sides of the sheet , it will not be at the edges! In other words. The insulation has to be thicker with sharp edges than with rounded electropolished ones.
But this process can´t be done in DIY and is quite costly.
You can coat with several polymeres and laquers. Polymeres can be applied in different ways, the best known are electrostatic spray coating and fluidized bed coating.
With the first procedure, the polymer powder is sprayed on the metal with the use of an electrostatic spray gun. After spraying the metal is heated in an oven, thereby melting the powder. This way You can reach app 200µm thickness. This isn´t thick enough for our use.
Fluidized bed means, that the powder is held in a box. When air is blown through the bottom of this box, the powder ´floats´ like a liquid.
The metal is heated up and and is dipped into the powder. The powder melts around the metal. This way You can get very even and thick coatings of up to 800µm. The big problem with perforated sheets is, that they cool down very quick...in fact too quick in most cases. The powder does not melt fully, leaving small unmelted balls of powder sticking loosely together (You can hardly see this effect, but the isolatory effect is drastically reduced)
So You have to find a coater who´s able to coat thin perforated metal sheets. Anyhow it´s very tricky to get coatings with more than ~300µm thickness.
Since The thickness of the coating wasn´t enough yet, I sprayed with an air gun several layers of an clear PU-laquer on.
If You want to coat the sheets all by Yourself You can use acrylic or PU-laqueres. I´m not a talented painter and always had problems with acrylic laqueres (they dry up quickly). The advantage is, they are water soluable. I prefer PU-laquer with which I get much smoother and better surfaces. The electrical properties of PU are excellent for ESL use either. But be prepared to coat many times thin layers till You reach the desired thickness (>500µm).
Use an air gun for coating...rolling won´t work well.
Since a professional coater provides for good reliable results (at least should he ;) ) and the PU-laquer isn´t cheap either, I´d have first the panels coated professionally and check if there´s a need for an additional top coat later on. You not only want working results but good optics too, right? :D
jauu
Calvin |
|
|
| Audio_idiot |
Calvin,
Thanks for your very detail explaination and instructions.
Now the specs get much more detail,
250mm x1250mm curve panel
spacer 1mm at front and 1.1mm at the back.
I will source for these now
perforated plate,
0.7~1mm thick, dia3mm 4mmc/c,
cureve 15~20degree or using 955mm ~715mm curve radius.
spacer of 1mm and 1.1mm. (ehat width are we looking at? 25mm?)
and try finding good painter for the PU coating.
This will take some time :xeye:
As for the electropolishing, I read about acid etching (w/o passing electricity) of the sharp edges, are they workable?
Cheers |
|
|
| MJ Dijkstra |
[QUOTE]Originally posted by Calvin
[B]Hi,
Curving the panel is easy...just tell Your metal worker to do it for You :D I used 15°-20°. The reason for curving is not(!) a wider dispersion character in first place, as it is often claimed. The reason to curve is higher stability of the sheet with less resonance and rattle. It is imo the only way to manufacture stable and stiff sheet metal panels without the use of addditional bracing.
***
To my experience curving a panel will improve dispersion to a significant degree. I used 60 degrees (total) curvature, just like Martin Logan. What a difference with a flat panel!!!!
15 - 20 degrees (total or is it 30 - 40 total???) may be insufficient.
I doubt that the guys from ML curved their panels for increased rigidity, because their are much easier and less time consuming methods to achieve this.
Just my thougts. |
|
|
| Audio_idiot |
Hi,
Thanks for sharing you experience.
On the cureve plate, do you have to make a curve form to hold the stator in shape. I couldn't convince myself that the shape will hold by it self... A part from this, I feel that curve panels are really any more difficult to construct then a flat one...
Another very stupid question,
Does the nerrow dispersion of the ESL really matters in small listening areas?
Cheers |
|
|
| I_Forgot |
Some like to argure otherwise, but my experience with my self-built flat panel speakers and my Quad ESL-63s is that the narrow dispersion afforded by flat panel speakers results in better "imaging". The drawback is that the sweet spot really is a spot, but that spot is a slice of heaven. When listening to speakers with wider dispersion (such as the Quads), you're less likely to hear complaints from people who aren't in the spot, but even in the spot, there is some reduced imaging and transient quality, I suspect due to multipath resulting from wider dispersion of mid and high frequencies. There may also be some effect due to the aiming of the speakers at the ears. The left ear, slightly off axis with the right speaker, will experience a slightly different frequency response for the right channel, and vice-versa.
A flat speaker with nothing done to broaden dispersion has a predictable frequency response- 6 dB per octave rise in level if I recall correctly. This is very easy to make a filter to flatten for on axis listening. As soon as you do things to broaden dispersion, if you do it sucessfully, you now have a more complex on-axis response pattern to try to even out.
I_F |
|
|
| Calvin |
Hi,
there might be a positive effect with a strong curvature (btw. ML used to use 30°, i.e. +-15° on older panels. The new panels use less curvature) but You might experience greater problems when mounting the film. Stability of the panel is increased substantially even with such a slight curvature. It will be strong enough as soon as the panel is fixed along its sides with a mounting frame. I use an aluminium profile that has an 10mm wide slot in which the Panel (~6mm thick) is pushed. Between panel and slot walls is a layer of damping foam.
The process of curving is imo the easiest and cheapest way to improve stability! There is no time consuming or extra costing difference in the other building steps. So the choice is to curve the sheets, which is done in a few seconds on a machine, or to fix some bracing elements onto at least one stator (how for example?).
ML´s choice was the curved stator since ML not only built good speakers but speakers with terrific looks and lowest production costs.
Sonically there is a difference between panels where the complete film area is driven (wether they feature a curvature or be they flat) and panels with electrical segmentation. Dynamics in the upper freq-range improve with complete panels, but it will cost seriously in the distribution character department. Imho the sonic result is a matter of taste. I like the strong dynamics of a complete panel, but since I wasn´t prepared to fix my head on a tiny little spot area I curved the panel to give at least a slightly better dispersion. So for me curving payed off twice ;) If You want at least a second person to have a good listening, You have to care for the distribution character of the panel.
jauu
Calvin
ps: simple etching just puts off material of the sheet, so sheet thickness will be reduced and hole diameter rises generally. Electropolishing on the other hand works best at sharp edges and small points (the points of hig filed strength). All etching processes have the advantage that the surface area becomes bigger and that the bond of the coating to the stator improves. |
|
|
| MJ Dijkstra |
I made a curved wire stator with a curvature of + and - 30 degrees, so 60 degrees total and it matched the curvature of the sequel2, so I think we have 60 degrees total indeed.
I don't know about the newest models like the vantage and the summit since I haven't seen them in reality.
Cheers, |
|
|
| Capaciti |
Hi,
i think there are always arguments pro and con dispersion.
For my experience it need to be a combination of both dispersion options.
The segmentation of my flat panels is optimized in such way, that dispersion is very good between +/15° and decreases rapidly at higher angles.
The mentioned better image of narrow dispersion panels has nothing to do with the dispersion itself, but with the interaction of early room reflexions. As long the dispersion is limited to avoid early reflections, preferably from side walls, it should have identical image.
On the other hand the discussion need to consider the frequency range we are talking about. I would say that all ESL show comparable dispersion up to 2 Khz and just differ at higher frequency. And high frequencies will be better absorbed by air, thus minimizing wall reflexions.
capaciti |
|
|
| Audio_idiot |
Hi,
Thanks guys,
Looks like dispersion is up to individual taste... No quarms with that;)
Curve plate will definately stiffen the panels, but my question is the how well will the stator hold up its profile during construction? please do elaborare more.
I really like the idea of curve panel, I am tempted to make a tube to hold the panels, and the cavity behind the panel (in the tube) can be filled with fibre..etc resistive materials to absorb the back wave. Like a vented box... the looks is appealing (to me at least)What do you think?
Cheers |
|
|
| Calvin |
Hi,
MartinLogan manufactures curved panels by the thousands.
1mm thick sheets are already quite strong and there are no such great forces while manufacturing that the curvature could be endangered in any way apart from maybe stomping on them :rolleyes:
I strongly recommend not to use any kind of box or chamber.
One of the biggest advantages of an ESL is that You can get rid of it, because every compartement is an unwanted energy reservoir. And thats audible!
Since an ESL is constructed -and works- different to a dynamic driver, the demands for an compartement differ significantely anyway.
PJ Walker gave some hints of what could be used in praxis -and those compartments definitely look different :att'n: - but as we know, he never gave up on building open/boxless ESLs, for thats the best way to build an ESL.
jauu
Calvin |
|
|
| Audio_idiot |
Hi,
Thanks for clearing that out.
Another question, when you mentioned 15~20degree, that is 15~20degree dispersion off the axis? to phisically the panel will cover double off that?
And the panel width can be same as flat panels considerations without any additional considerations?
Cheers |
|
|
| wolf_moses |
| Roger Sanders' ESL book states it is very hard to keep the diahragm parallel to curved stators, as over time it will tend to be pulled into the back stator, causing non-linearity. I'm sure manufacturers can get around this, but what about the average DIYer? |
|
|
| Calvin |
Hi,
there are quite a few mistakes and failures in Sanders Book.
The one about curving is imo one of those ;-)
jauu
Calvin |
|
|
| Bazukaz |
Hi,
Since this thread is about easily obtainable methods as well , i'd like to continue it a bit to this direction :).
Currently , i am experimenting with conductive coatings. I found a few methods to do this , each having its own disadvantages.
One method is to use PVA based adhesives.Some types of these adhesives are hygroscopic , and will attract moisture. However , PVA , being water based , does not wet mylar well.Adhession to film is not very good either. The resistance is on the low side.
Another method is shoe cream rubbing.This method probably works because the cream contains carbon black(?). The resistance is too low as well , though it is possible to rise it by cleaning with wet cloth.
I tried some other methods , like dissolving nylon fishing line in acetic acid. This method gives very high resistance, but is hard to work with due to vapors of acid.
Any other ideas ?
Regards,
Lukas. |
|
|
| wolf_moses |
Calvin on curved ESLs
Hi Calvin,
Could you share your method of building curved ESLS which avoids Sanders' mistakes?
Thanks,
Wolf |
|
|
| moray james |
Lukas if you want PVA adhesive like Weldbond to stick to mylar you only need follow the directions. The trick is to dilute with water (5 parts water to 1 part weldbond) and then simply use a small hand held mister bottle to apply a lifht even coating. Let dry and you will hav excellent adhesion. You can also dope the water that you use to make it work better try a non sented dryer antistatic sheet soaked in the water. You can search under Quats
here is part of a theradWeldbond Post #5
Something that I will post to stimulate experimentation. Weldbond white glue which is a modified PVA adhesive is made by Frank T. Ross & Sons Ltd. in Toronto. (www.weldbond.com) when diluted one part glue to five parts water can be sprayed or brushed onto mylar and dries to a fine cloudy thin film. This bonds extreemly well to mylar. You can experiment with quaternium ammonium salts to dope the adhesive to achieve the desired resistivity. The most simple source would be to take a non sented Downy dryer sheet and soak it in some water. Then use the water to dope the adhesive. Many anti static products like hair conditioner or shampoo with added hair conditioner contain such salts and could also be experimented with. Easy to play with and almost free. Weldbond can be found at most any craft store as it is considered a universal adhesive. By the way Weldbond is totally non toxic. Best regards Moray James. |
|
|
| Calvin |
Hi,
with curved stators I wouldn´t use heat tensioning but solely mechanical tensioning with the highest possible value..of course just tension in the straight direction. The tension in curved direction should be as low as possible....just enough to straighten out any folds. But when You apply very high tension in the straight direction, You even get away with quite strong tension in the curved direction (obeying the 1:70...1:100 rule for d/s)!
You can do this with a mechanical stretching frame with curved ends.
I use only instant bonding glue or tape. After application of the glue or tape to the backside stator I stretch the membrane and lay ist on the stator. The stretching frame presses the membrane onto the glue joint by its weight. With 3M-Tape the instant bond is very strong but increases over some hours to its final value. Then I release the pressure of the tensioning frame and cut the panel free leaving some diaphragm material left. Now I glue some thin doublesided tape on the outside (backside) of the stator and turn the diaphragm over its rim and glue it on the thin tape. This way You have much less trouble with tapes that might not have the shear strenth to hold the constantly high diaphragm tension over a long time for You have much more glued area. If there were some tiny wrinkles left on the frontside of the stator, You could even pull those flat since the foam tapes I use as spacers allow for those small final corrections.
The soft foam introduces some damping to the membrane which I regard as positive for less standing wave-Q and less stress on the membrane itself at this joint. CSD Measurement showed very good results. MartinLogan used to glue the membrane on the stators backside too..but only in the straight direction.
jauu
Calvin |
|
|
| Audio_idiot |
Hi,
This is super informative!! Thanks guys!!
More stupid questions...:D
MJ,| quote: | | Weldbond white glue which is a modified PVA adhesive is made by Frank T. Ross & Sons Ltd. in Toronto. |
Don't think I'll find them in my local stores..:bawling: Could you help me nerrow down the search by tell me what are their "normal" application? then I can start reading label and contents in the store like I'm in a book shop...:rolleyes: :rolleyes:
Calvin,
| quote: | | But when You apply very high tension in the straight direction, You even get away with quite strong tension in the curved direction (obeying the 1:70...1:100 rule for d/s)! |
I have been told many time about tension high and low but I don't have a feel for thatis ther any indication of "high-tension":( , is there any indiction that I can fall on? eg weight for stretching, pressure applied? something? anything...
| quote: | | (obeying the 1:70...1:100 rule for d/s)! |
I am getting confused with the d & s value.. is it,
d=Panelwidth, &
s= Panel length?
| quote: | | You can do this with a mechanical stretching frame with curved ends. |
Are you saying that the bicycle tube stretch table is no use in your technique? In stead you have a rectangular frame to do the work?
Cheers |
|
|
| Calvin |
Hi,
d/s means diaphragm-to-stator-distance....usually ~ 0,5---5mm
To ensure that the diaphragm will not touch one of the stators when playing music the membrane should be supported at distances of 70 to 100 times the d/s. For a d/s of lets say 1,0mm supports should be distanced between 70 and 100mm apart. You see those supports by MartinLogan as clear horizontal spacers, or with Audiostatic as simple silicon dots.
When tensioning a material it is first just tensioned. When You stretch harder there will be a point where the material starts to ´flow´ and the tension won´t increase any more but the material starts to become thinner. With a film You can easily feel this point, since the stretching force doesn´t increase anymore but decreases. Furthermore if You listen to the ground resonance by e.g letting fall a pingpong ball on the film the resultant tone doesn´t get higher. Depending on the film this point is reached between 2% and 4% of length increase. I´d first test the film on this property.
I usually stretch around 2% which gives strong tension.
The good point about the tubed stretcher is, that it can develop a very even tension over the complete area and in every direction. With curved panels You need to stretch only in one direction. You could probabely use a tubed stretcher when You ensure that it stretches only in the straight direction as the mechanical stretcher does. The tube doesn´t have an advantage over the mechanical stretcher in this case as it has in the case of a flat panel with tension in 2 directions
jauu
Calvin |
|
|
| Audio_idiot |
Hi Calvin,
Thanks for clearing my cloud over the tension part:) Now I'm crystal clear.
But| quote: | | d/s means diaphragm-to-stator-distance....usually ~ 0,5---5mm |
But in the earlier suggestion of having 250mm x 1,200mm panel, using 1mm gap (stator to diapharm), has already exceeded the 1:100 ratio, am I understand properly that we can get over this by havingpoint support or strip sectionalised the panels to keep the 1:100 d/S ration and we can go as big a panel as we wish with still keeping the 1mm gap?
Cheers |
|
|
| Zvon |
Hi Calvin,
Your method of Mylar tensioning may be too high. While the high membrane tension is desirable for stability, Mylar service life in vicinity of spacers may be considerably shortened (as seen on many failed DIY and commerical ESL panels).
You tension Mylar to yield ("flow") point and the resulting stresses (membrane stresses) are uniform across the membrane thickness. Close to spacers the membrane is subject to significant curvatures and bendig stresses that should be superimposed to the membrane stresses.
Cyclic loading (read fatigue) of this magnitude (close to yield) are responsible for delamination of coatings in this area (read loass of bias) and ultimately tears parallel to the spacers (read panel rebuilding).
Regards |
|
|
| moray james |
If you are into experimenting with coatings weldbond is an amazing product to play with. You can find information on weldbond and where you can buy it. Your best possible chance of finding weldbond would be in a craft and hobby store. Regards Moray James.
http://www.franktross.com
I am sure if you email weldbond they will inform you of where you can purchase thier product. I hav even seen it on ebay. |
|
|
| wolf_moses |
Thanks for your very helpful response, Calvin. Plenty of food for thought.
What about the tendency for foam to rot over time, as in woofer surrounds?
Regards, |
|
|
| Audio_idiot |
Hi Zvon,
I think your concern is correct theoritically. working in plastic zone of any material isn't that predictable... There is a way around it.
BTW, please kick me if I'm wrong...
In tech term, what Calvin did was working the film into it's plastic zone just over the yield point (where the film becomes permanently elongated) If you are worried about the longivity of the film/panel, experiment with the force or elongation lenth and you will find the "yield point", then work within the yield point.
When I start on the project. This is the plan: I'll find the weight required at yield point by hanging weight at the stretching ends. This is how i imagine, Take a speciment of say 100 wide 1m length of film, I will use rectangle frame on a table. Double sided tape one end to a table (fix end) and double side tape the other end (free end) on a stick and hang weight on it, then I'll bounce a ping pong ball (like Calvin suggested) on it, then record the weight where the tone of the film stated to stay constant. That should be the yield point.
The next is your choice, one can ease off a few grams to work safely within the elastic zone or do what Calvin did and work in the plastic zone.
Just my two cent worth
Cheers |
|
|
| Calvin |
Hi,
so is that ;) The Hostaphan film I use starts to flow above ~4%.
So with just 2% I stay well away from this border. But as Zvon remarked right, You have to keep such things in mind.
I try to destress the joint spacer-film a bit by using soft materials (even when I use hard spacers I add some soft and damping material at those points) and up till today I had no diaphragm failure so far because of this reason. I have to admit that I have no 10 Years+ experience on this point, but all panels work for several years by now. A bit funny to me that the harder stretched panels all work easier, better and with less flaws than the softly tensioned panels :confused:
@AI
You´re right. The panel can be as big as You wish. As long as the rule for the membrane support applies, the panel works unconditionally stable to very high spl-levels, wether it is a very small d/s or a rather big d/s value. When You measure the dimensions of a ML-panel You will easily verify this rule ;)
@WM
aging is a natural problem, especially for certain plastics. UV-Light is the ´destroying´ factor. So You might capsule Your panel to achieve a longer life, and/or You have to use materials which are long lasting by their chemical formulation. Foam surrounds of drivers have been problematic, but so did PVC- and rubber-surrounds too. The foam tapes of 3M are specified as UV-resistant , but I´m convinced, that they won´t hold forever :cannotbe: On the other hand I think that a average lifespan of 10-15years is an acceptable value for a speaker.
jauu
Calvin |
|
|
| Audio_idiot |
Hi calvin,
Thanks again for clearing yet another puzzle for me!!
| quote: | so is that The Hostaphan film I use starts to flow above ~4%.
So with just 2% I stay well away from this border. But as Zvon remarked right, You have to keep such things in mind.
I try to destress the joint spacer-film a bit by using soft materials (even when I use hard spacers I add some soft and damping material at those points) and up till today I had no diaphragm failure so far because of this reason. I have to admit that I have no 10 Years+ experience on this point, but all panels work for several years by now. A bit funny to me that the harder stretched panels all work easier, better and with less flaws than the softly tensioned panels |
I wouldn't commend on how much tension is desireable but from your experience, I recon it is better would keep at about 3~4% elongation or at exactly the point before the film "flows"..... But I recon it is not as easy to determine this point as it may seem
I think your soft materials really gave that extra elasticity the film need.
Excellent work Calvin!!! |
|
|
| Audio_idiot |
Hi guys,
I'd finally ordered my perforated plate.
I will then send for rolling after collection.
I planned to acid-etch the plates at home using battery acid to remove sharp edges/burrs and oil/dirt before applying multiple layers of Polyeutelyne coat.
I recon its a good time to discuss about the film coating. The PVA glue application inspire me into thinking, can we use any other advesive, blend with some graphite powder (or any resistive elements)and apply onto the film?
Is shoe shine wax any bad as film coat?
Cheers |
|
|
| Calvin |
Hi,
sorry for keeping on repeating :cannotbe:
I don´t recommend to stretch the film to the flow point! I stay well off of that point, stretching just ~2% where 4% is possible.
Acid etching might reduce the sharp edges problem a bit, but also reduces the thickness of the metal sheet and rises hole diameter!
I recommend using Polyurethane laquer too and not Polyethylene alone (it just has very high flashover values..all other parameters are less good). A topcoat of PU shows better electrical characteristics for overload conditions (lower surface resistance and higher epsilon).
PU laquer is easily available as a 2-component coating for boats and staircases, so its very rugged, it handles easily and gives very glossy and nice even surfaces.
jauu
CAlvin |
|
|
| moray james |
| you need to remember that mylar has a surface tension of about 45 dynes per cm Sq which is about the same as polished plate glass in other words very smooth. Not many adhesives will grip on a surface this smooth. Adhesives grip or grab onto and into any small surface irregularity they can find and so have dificulty on very smooth surfaces. Most people think that glues chemically bond materials together but that is no normally the case. Unless you really want to spend your time experimenting and don't mind rebuilding your speakers when they fail you would be best recommended to use one of the several methods that are proven to work and are available. These will cost you some extra money but they will insure that you only have to build your speaker once. You can search to find this information on this forum. Regards Moray James. |
|
|
| SillyCone |
Hi,
| quote: | | I recommend using Polyurethane laquer too |
I do not think it's a good idea to cover a high resistivity
protective coating with a lower resistivity coating.
The reason is that you have no electrical path for eliminating
the "playing loud" charge build up in both coatings.
You must use a low resistivity coating from the metal and up,
or You will get a very moody Esl that sometimes will play loud
and sometimes not at all.
I also wonder if Polyurethane has the same low required resistivity as Nylon66 or PVC, shall do some testing. |
|
|
| I_Forgot |
The glue to use is Scotchgrip 4693 or 4693H which can be purchased in 5 oz. tube from McMaster-Carr for about $8. A 5 oz tube will be enough glue to make a pair of speakers, if you make a lot of mistakes along the way. If you don't, you'll have plenty left over.
This glue is formulated to bond to low surface energy plastics, such as polyester (mylar). It works. It bonds to the mylar so tightly that the film will tear before the glue lets go.
I_F |
|
|
| Calvin |
Hi,
@silly:
You´re right, the basecoat should have a lower resistivity too. Idon´t recommend PE as insulator.
PU has -at least the infos I´ve got about it- quite similar parameters as PVC and Nylon (Nylon6) with regard to resistance values and epsilon and is easily available as laquer on the market.
jauu
CAlvn |
|
|
| Calvin |
Hi,
@silly:
You´re right, the basecoat should have a lower resistivity too. Idon´t recommend PE as insulator.
PU has -at least the infos I´ve got about it- quite similar parameters as PVC and Nylon (Nylon6) with regard to resistance values and epsilon and is easily available as laquer on the market.
jauu
Calvin |
|
|
|
|
|
