All things are relative, including the effectiveness and ease of application of high resistance coatings to diaphragms.
I have had good results with licron and it doesn't matter a bit if the coating is uniform or not. When have you ever seen a uniform graphite coating? When have you ever seen a graphite coating with resistance as high licron?
As long as the resistance is high, the coating will do its job just fine. You don't need a heavy coating. One pass with the spray is sufficient. If any of the licron is on the diaphragm it will work.
Licron is the easiest way to get a high resistance coating that seems to last. There are no "iffy" ingrediants like sheep's kidney extracts, etc., found in many "amateur" formulations. It sticks to polyester and provides high resistance. Sure, it would be nice if there were a way to apply it so it appears uniform, but are you going to look at the speakers or listen to them?
I_F
I have had good results with licron and it doesn't matter a bit if the coating is uniform or not. When have you ever seen a uniform graphite coating? When have you ever seen a graphite coating with resistance as high licron?
As long as the resistance is high, the coating will do its job just fine. You don't need a heavy coating. One pass with the spray is sufficient. If any of the licron is on the diaphragm it will work.
Licron is the easiest way to get a high resistance coating that seems to last. There are no "iffy" ingrediants like sheep's kidney extracts, etc., found in many "amateur" formulations. It sticks to polyester and provides high resistance. Sure, it would be nice if there were a way to apply it so it appears uniform, but are you going to look at the speakers or listen to them?
I_F
QUOTE:
I have had good results with licron and it doesn't matter a bit if the coating is uniform or not. When have you ever seen a uniform graphite coating? When have you ever seen a graphite coating with resistance as high licron?
Hi,
The particles in Licron are simply too large to ensure uniformity in ultrathin layers. From the limited experience with Licron there already have been several people with uniformity problems so it is definately not the easiest way. If it suits your needs that's fine, but it's physical and chemical properties are simply not optimised for the use in ESLs. Did you ever measure the added mass of Licron? Probably not! The same thing for the electrical resistance, which seems orders of magnitude lower than the (professional) EC-coating. I am not saying it is bad stuff, but it is not the best.
Without a careful examination of the physical and chemical properties, telling people that coating X or Y is the best or the easiest is just a personal statement.
I have had good results with licron and it doesn't matter a bit if the coating is uniform or not. When have you ever seen a uniform graphite coating? When have you ever seen a graphite coating with resistance as high licron?
Hi,
The particles in Licron are simply too large to ensure uniformity in ultrathin layers. From the limited experience with Licron there already have been several people with uniformity problems so it is definately not the easiest way. If it suits your needs that's fine, but it's physical and chemical properties are simply not optimised for the use in ESLs. Did you ever measure the added mass of Licron? Probably not! The same thing for the electrical resistance, which seems orders of magnitude lower than the (professional) EC-coating. I am not saying it is bad stuff, but it is not the best.
Without a careful examination of the physical and chemical properties, telling people that coating X or Y is the best or the easiest is just a personal statement.
I_Forgot said:
I have had very good results with Zelec, and I can tune the resistance with dilution. I’ve built panels with low and high resistance coatings and monitored SPL vs time as the panel charges up and I can see the effects of resistance on charge time. I did this as part of tuning the coating for use on the original Quad ESL, where there isn’t a series resistor, and the panel resistance is used as a series resistor of sorts. When applied in a thin film, it self levels as it dries increasing uniformity. Too bad you can’t get the stuff anymore. It came in a water based and a solvent based solution. I’m working on the 2 pint samples I got from Dupont.
The way I did graphite was to rub it in liberally (a knuckle busting job to be sure), it has to be ground on, not just wiped around. Then I took paper towel pieces wetted with alchohol and tried to wipe it off. You get a bit off and the loose stuff is removed that way, but you can never get it all off. This forms a higher resistance coating and gets rid of the loose stuff. I suppose that it is possible to remove it completely but I never wiped that much. Vacuuming in between steps helps too. But in general graphite is a total pain in the butt to apply (hand cramp fiesta), and isn’t particularly uniform, but in my experience, it is stable and does last a long time (hard to get more stable and less effected by humidity than carbon).
How much of the can did you use to build your panels, and how big are your panels? It would be nice to know how far a can goes. $55 is steep, but if it’s a lifetime supply, then that’s not too bad.
The iffy ingredients turn me off too. Licron and Zelec are designed to bond to polyester, as Martha would say, “it’s a good thing”.
Sheldon
you can experiment with...
dilute PVA adhesive (white glue). I have used Weldbond but there are hundreds to choose from. When diluted 5 or 6 to 1 (water to PVA) you can apply it with a very fine spray mister bottle. In this dilute form it will wet onto the surface very well and more importantly stick very well. If wetting is an issue you can use any number of agents to improve wetting. The PVA is hydroscopic in nature and so works via moisture layer conduction. You can also dope the mix with some "Quats" which you can easily extract from an antistatic dryer sheet that you have soaked in water. This type of coating has a thin milky look to it not totally clear but very see through. If you use a fine mister you can achieve a very thin coating. Something for those who want to play and don't want to spend a lot of money on things. Low tech yes but it works. You can also play with adding a very small amount of Borax powder to this mix. You can start with a tiny pinch of Borax and work down from there. The Borax will make the PVA cross link and will not only give it far better stretch qualities but also make it bond better just dont use too much.
That should keep some of you guys busy for a while. Regards Moray James.
dilute PVA adhesive (white glue). I have used Weldbond but there are hundreds to choose from. When diluted 5 or 6 to 1 (water to PVA) you can apply it with a very fine spray mister bottle. In this dilute form it will wet onto the surface very well and more importantly stick very well. If wetting is an issue you can use any number of agents to improve wetting. The PVA is hydroscopic in nature and so works via moisture layer conduction. You can also dope the mix with some "Quats" which you can easily extract from an antistatic dryer sheet that you have soaked in water. This type of coating has a thin milky look to it not totally clear but very see through. If you use a fine mister you can achieve a very thin coating. Something for those who want to play and don't want to spend a lot of money on things. Low tech yes but it works. You can also play with adding a very small amount of Borax powder to this mix. You can start with a tiny pinch of Borax and work down from there. The Borax will make the PVA cross link and will not only give it far better stretch qualities but also make it bond better just dont use too much.
That should keep some of you guys busy for a while. Regards Moray James.
I would think any thickness under 20 micron would suffice in order to outperform a dynamic speaker easily, wouldn't anyone agree?
Getting it as light and thin as possible would in other words only improve things in comparison to other ESLs, wouldn't it?
I wonder how powerfull the "motor" of an ESL actually is....
And how weight/drive ratio is between different speakers and different ESLs....Aren't ESLs astronomically superior (in weight/drive ratio) to dynamic drivers allready?
The esl club site shows a HV amplifier, they claim the audio transformer in an ESL remains the weakest link in the chain by far......
Though the ESL-57 treble panels I have sound really ultra sweet, the goal -for me- is to outperform them or at least copy the performance..
I'm thinking of a full-range ESL, but fear I will be more satisfied with a dynamic bassdriver.. I dumped the bass panels of the ESL-57 because their sound didn't quite satisfy me. The bass speakers I'm using right now are not exactly "audio nirvana" either though....
If I use a dynamic bass speaker would it be best to build dipoles?
I was looking at the latest QUAD ESLs, they are strutted by a triangular piece. They appear to gain more steadyness that way, looks interesting..
Maybe it's a good idea indeed to start with a small mid-treble panel , say 6"x24" like my QUADs. That would be a fair comparison, wouldn't it?
Is coppertape and pertinax really the way to go? Or are there more possibilities like strips of metal and plastics? I guess it makes sense to get every part as rigid as possible, for durability IS my goal...
Thanks for all your thoughts...
Getting it as light and thin as possible would in other words only improve things in comparison to other ESLs, wouldn't it?
I wonder how powerfull the "motor" of an ESL actually is....
And how weight/drive ratio is between different speakers and different ESLs....Aren't ESLs astronomically superior (in weight/drive ratio) to dynamic drivers allready?
The esl club site shows a HV amplifier, they claim the audio transformer in an ESL remains the weakest link in the chain by far......
Though the ESL-57 treble panels I have sound really ultra sweet, the goal -for me- is to outperform them or at least copy the performance..
I'm thinking of a full-range ESL, but fear I will be more satisfied with a dynamic bassdriver.. I dumped the bass panels of the ESL-57 because their sound didn't quite satisfy me. The bass speakers I'm using right now are not exactly "audio nirvana" either though....
If I use a dynamic bass speaker would it be best to build dipoles?
I was looking at the latest QUAD ESLs, they are strutted by a triangular piece. They appear to gain more steadyness that way, looks interesting..
Maybe it's a good idea indeed to start with a small mid-treble panel , say 6"x24" like my QUADs. That would be a fair comparison, wouldn't it?
Is coppertape and pertinax really the way to go? Or are there more possibilities like strips of metal and plastics? I guess it makes sense to get every part as rigid as possible, for durability IS my goal...
Thanks for all your thoughts...
Re: you can experiment with...
Hi
Please note the safety sheets of Borax powder. Its use in most of the dutch faramacist production units is prohibited because it is denoted as a dangerous substance. Maybe you can use another strong base here.
Nevertheless, Moray is right in his procedure to make a good HC- system. The only comment is that doping with Quats may be temporarily so I would add a relatively large amount of it, so keeping the resistance on the safe (low) side.
moray james said:
Hi
Please note the safety sheets of Borax powder. Its use in most of the dutch faramacist production units is prohibited because it is denoted as a dangerous substance. Maybe you can use another strong base here.
Nevertheless, Moray is right in his procedure to make a good HC- system. The only comment is that doping with Quats may be temporarily so I would add a relatively large amount of it, so keeping the resistance on the safe (low) side.
I'm still a bit of a noob on esl practice, some say the large capacitance of an ESL makes it hard to drive, some claim it is the impedance dip.
What precautions can I take best to build it in a way that it is easier to drive (not going below 1 ohm preferably)? It would seem as if an ESL is better for low freq use If you'd ask an amp...
I'm driving my ESLs with a chip based amp (audiodigit mc4x100, though nowadays I would've bought AMP9 from 41hz.com). I feed this amp with 15V instead of 27V, this way it can drive serious loads. Though I have two class A hybrid (vacume-solid) monoblocks, I prefer the sound of the Tripath chip based amp.
The amp has overcurrent protection, but never went into protection-mode with the ESLs connected. The amplifier is never even warm, where it heated up quite a bit in my car (on 24V). My guess is it's no problem, it sounds very good upto high listening levels. Or could there be problems I'm overlooking?
Any thoughts much appreciated...
What precautions can I take best to build it in a way that it is easier to drive (not going below 1 ohm preferably)? It would seem as if an ESL is better for low freq use If you'd ask an amp...
I'm driving my ESLs with a chip based amp (audiodigit mc4x100, though nowadays I would've bought AMP9 from 41hz.com). I feed this amp with 15V instead of 27V, this way it can drive serious loads. Though I have two class A hybrid (vacume-solid) monoblocks, I prefer the sound of the Tripath chip based amp.
The amp has overcurrent protection, but never went into protection-mode with the ESLs connected. The amplifier is never even warm, where it heated up quite a bit in my car (on 24V). My guess is it's no problem, it sounds very good upto high listening levels. Or could there be problems I'm overlooking?
Any thoughts much appreciated...
Hi,
Using electrical segmentation (e.g. audiostatic) will raise the inpedance in a significant degree. Your amplifier will like this. Your step-up transformer will like it too. In fact you're creating a semi-crossover situation which can be criticised by some purist, but the sound will improve most of the time (better dispersion as well).
Another way is to use a resistor of say 1,5 ohm at the output of the amp, but this can affect the high freq. response, so you should use the lowest value acceptable. You could also add some impedance increasing net work like described in the book of E. Fikier. Did you find it?
In case of amplifiers, power is not enough. It must be stable under high capacitative load. If not, there will be overshoot, sometimes a burned amplifier.
MartinJan
Using electrical segmentation (e.g. audiostatic) will raise the inpedance in a significant degree. Your amplifier will like this. Your step-up transformer will like it too. In fact you're creating a semi-crossover situation which can be criticised by some purist, but the sound will improve most of the time (better dispersion as well).
Another way is to use a resistor of say 1,5 ohm at the output of the amp, but this can affect the high freq. response, so you should use the lowest value acceptable. You could also add some impedance increasing net work like described in the book of E. Fikier. Did you find it?
In case of amplifiers, power is not enough. It must be stable under high capacitative load. If not, there will be overshoot, sometimes a burned amplifier.
MartinJan
The load the ESL will place upon the amp is very dependent on stepup ratio of the audio transformer. Higher stepups will be much more difficult to drive.
Assuming a well designed transformer with high enough Lprim, the capacity seen by the amplifier is approx. equal to (Cesl+Ctransformer) * N^2
where
Cesl= the capacitance of the esl
Ctranformer = the combined internal capacitances in the transformer refelcted to the secondary
N=turns ratio
With increasing N, Ctransformer will also increase (more or less in a linear way, when all else is equal which it is not). So doubling stepup ratio will increase capacity by at least 4 times, probably much more.
So if you're after easy drive reqs go for a very low stepup, in the 1:50-1:70 range. Will also allow for more bandwidth in the transformer. The price you pay is less output. For which you can compensate by either limiting spacing (increasing field strength, also limiting max swing) or using an amplifier that will provide more voltage swing (higher rail voltages, go for at least +/- 50-60V).
Basically, by lowering stepup you're trading current requirements for voltage requirements.
The best way is to carefully balance it all, there is no standard recipe
Thanks again for the valuable info, I think I can best start with a hybrid system. This way the ESL can be kept small and the step-up minimal.
This also keeps things at a more workeable level.
How well would an ESL work in a sort of enclosure? I know ribbons are often placed in closed cabinets, but never seen this done with ESLs. Would it result in high distortion? I read ESLs have very high Vas (equivalent volume), but a smaller panel going from 200 hz and up might work in a reasonably small enclosure, wouldn't it?
I was planning on building a 65V amplifier with the TK2350 chipset from Tripath, don't know if it will drive an ESL though....
It may sound crazy but I just love the detailed sound of these chips...
http://www.41hz.com sells very nice kits with these chips....
At least they stay well within my budget.
Another thing I wondered, if I wind my own step-up transformer the primary windings should be kept minimal in windings in order to lower C, but to increase Z and L it should be high in windings. Isn't that contradictory?! I figured it to be a good idea to use pretty thin wire in order to increase Z and decrease C, and wind it all in not to many layers.
Or would it be a good idea to use more windings and thicker wire with thicker insulation in order to lower C?
Still I dream of building a direct drive amplifier for the project. (from 200 Hz and up, still using a dynamic bass speaker). Only thing I'm afraid of is it won't sound any better...
I think I will start with a low power single and very small graphite coated panel, stepped up by a small mains transformer and an SMPS supply on it like here:
http://sound.westhost.com/project105.htm
Thanks again for your thoughts..
This also keeps things at a more workeable level.
How well would an ESL work in a sort of enclosure? I know ribbons are often placed in closed cabinets, but never seen this done with ESLs. Would it result in high distortion? I read ESLs have very high Vas (equivalent volume), but a smaller panel going from 200 hz and up might work in a reasonably small enclosure, wouldn't it?
I was planning on building a 65V amplifier with the TK2350 chipset from Tripath, don't know if it will drive an ESL though....
It may sound crazy but I just love the detailed sound of these chips...
http://www.41hz.com sells very nice kits with these chips....
At least they stay well within my budget.
Another thing I wondered, if I wind my own step-up transformer the primary windings should be kept minimal in windings in order to lower C, but to increase Z and L it should be high in windings. Isn't that contradictory?! I figured it to be a good idea to use pretty thin wire in order to increase Z and decrease C, and wind it all in not to many layers.
Or would it be a good idea to use more windings and thicker wire with thicker insulation in order to lower C?
Still I dream of building a direct drive amplifier for the project. (from 200 Hz and up, still using a dynamic bass speaker). Only thing I'm afraid of is it won't sound any better...
I think I will start with a low power single and very small graphite coated panel, stepped up by a small mains transformer and an SMPS supply on it like here:
http://sound.westhost.com/project105.htm
Thanks again for your thoughts..
Hi, v-bro. I tried something like an enclosure to see if I could absorb the ESL back wave to get them closer to the wall than the ~1m I have them at now. Instead of building a box and stuffing it right away, I used heavy blankets covering the whole back of the ESL to get some idea if it was worth going ahead and building the box.
Unfortunately, the experiment was not successful. It turned the magical sound of the ESL into something very ordinary. I guess that's why you don't see it done except by Janszen. I had to find out for myself.
I would consider my testing crude at best, but I'm convinced that I don't want to give up the back wave.That sounds like a good place to start. You can probably reuse the bias supply on a larger driver if you want to go big later.
Thanks for the very quick response Bill 
It's exactly what I figured, I tried various absorption materials behind my ESLs, but ended up using only a small bent piece of foam. More absorption always resulted in distortion and less imaging-width.
Only thing I wondered, can an ESL be designed in a way that it would work well in an enclosure. This would perhaps make it possible to decrease the size of a bass system. I would perhaps build an ESL bass system after I successfully built treble panels, this way I would still have fullrange ESL sound in the end...
It's exactly what I figured, I tried various absorption materials behind my ESLs, but ended up using only a small bent piece of foam. More absorption always resulted in distortion and less imaging-width.
Only thing I wondered, can an ESL be designed in a way that it would work well in an enclosure. This would perhaps make it possible to decrease the size of a bass system. I would perhaps build an ESL bass system after I successfully built treble panels, this way I would still have fullrange ESL sound in the end...
Thinner wire won't increase Z. Z is determined completely by L, Rdc is neglectible (<0.1 ohm in my own trannies). But you're on the right track.
You need a certain minimal number of prim turns, depending on required Lprim and core parameters. Required Lprim in turn depends on lowest frequency. So with a given core, lower bottom frequency limit = more turns
Combined with turns ratio this gives the # of sec turns. It will be huge. To fit it all on the core you'll probably need wire as thin as youcan handle to keep number of layers (=capacitance and leakage) reasonable so bandwidth at the high end won't be compromised too much. Not much to choose on the sec side.
So: More bandwidth at the low end = less bandwidth at the high end. The only way around is lower stepup.
Leakage can be reduced by interleaving but this increases capacitance
Comprimises, compromises... transformer design is true art 
The bottom line: unless you're prepared to spend incredible amounts of time in designing, winding, measuring, rejecting and rewinding transformers, buy a ready-made one
As for direct drive: make life easy on yourself and forget about it. Not something for a first try.
Aiming for a hybrid system is a good idea, makes live easier on the panel, transformer, construction, .... You might even get away with 6/230V toriods for stepup.
And oh, don't put the woofer in an enclosure either, dipole bass is as addictive as esl-sound
again very much appreciated,
I was thinking indeed to concentrate on the panel first and test different mains transformer types. This way I can encorporate my findings in a transformer that can be created afterwards.
I was just fantasizing an ESL panel shaped round where the treble comes from the centre and the bass from around it (bit like an ESL-63).
I was thinking perforated sheet stators, slightly cone shaped. But than, how would one divide the frequencies in such a design well?
1:well partitioned membrane resistance?
2:separated stators with different V drive ratios from the step-up circuit (tapped 2nds or resistors)?
Main question in this, would it be possible to have lower frequencies gradually decrease towards the centre?
I do plan on making a whole series when things go well.
I was thinking indeed to concentrate on the panel first and test different mains transformer types. This way I can encorporate my findings in a transformer that can be created afterwards.
I was just fantasizing an ESL panel shaped round where the treble comes from the centre and the bass from around it (bit like an ESL-63).
I was thinking perforated sheet stators, slightly cone shaped. But than, how would one divide the frequencies in such a design well?
1:well partitioned membrane resistance?
2:separated stators with different V drive ratios from the step-up circuit (tapped 2nds or resistors)?
Main question in this, would it be possible to have lower frequencies gradually decrease towards the centre?
I do plan on making a whole series when things go well.
Hi, v-bro. How about something like the attached drawing? Tapered perforated stators with center tweeter sections.
For a first try, I'd run the four large sections in parallel, both small sections in parallel, and a single resistor between the paralleled large and small sections.
Attachments
While I was working on the drawing in the last post, I was thinking about using six separate perforated sections for each stator, which I hope helps explain the paralleling of the sections. It's relatively difficult to build the insulators to carry the sections, though.
An easier way would be to use double sided tape to section a single piece of perf as shown above and run the whole ESL full-range. You won't have electrical control of the sections' rolloffs this way, but the mechanical rollofs will still be there.
Maybe a blend of the two ideas? One large piece of perf with the center cut out for a smaller piece of perf and double sided tape to form the sections?
An easier way would be to use double sided tape to section a single piece of perf as shown above and run the whole ESL full-range. You won't have electrical control of the sections' rolloffs this way, but the mechanical rollofs will still be there.
Maybe a blend of the two ideas? One large piece of perf with the center cut out for a smaller piece of perf and double sided tape to form the sections?
wire winding...
strikes me as the most practicle way of building such a segmented stator. While more work than a perf version the wire wound version keeps everthing as small as possible on a single substrate. A good example of this is the Dutch fellow who wound a 63 style stator.
A perf fabrication looks good on a drawing but as soon as you trie to fit it all together and try to maintain some sort of physical integrity it falls apart in my view. Just my 2 cents. Regards.
strikes me as the most practicle way of building such a segmented stator. While more work than a perf version the wire wound version keeps everthing as small as possible on a single substrate. A good example of this is the Dutch fellow who wound a 63 style stator.
A perf fabrication looks good on a drawing but as soon as you trie to fit it all together and try to maintain some sort of physical integrity it falls apart in my view. Just my 2 cents. Regards.
It works the other way round: The center ring of the 63 plays fullrange, the outer rings receive a signal in which high frequencies are filtered out. The additional top and bottom panels play only bass.
Signal to the outer rings is also delayed in time which is vital to create the wide dispersion pattern of the 63. It won't work very well with simple resistors for filtering (although this will give better dispersion, it will compromise impulse response). Delay lines are the solution.
To get it right you need to take everything into account that connects to the delay line, transformer, panel capacitances, parasatic capacoitances between the panels etc etc. It took Peter Walker 18 years
For instance, exchanging the transformer in the esl63 for something different will likely completely ruin the dispersion pattern.
It's all explained in detail in the various publications of Peter Walker, you'll find them with a bit of googlin'
www.quadesl.org is a good starting point.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.