How long was the EHT turned off before you removed the connection?
This will have a big effect on SPL level when you remove the HV connection and connect diaphragm to CT.
It is not the charge stored in the stator tape that directly makes the sound. Rather it is the charge stored in the stator tape that charges up the air gap capacitance when the diaphragm is connected to the CT. The voltage across the air gap when its capacitance is charged is what produces the sound.
In a simplified lumped model, you can think of the air gap capacitance and stator tape capacitance as two capacitors in series. The voltages across each capacitor will be in the ratio of the leakage resistance associated with each capacitor. With PVC insulation, its leakage resistance is much less than that of air, so essentially all of the EHT voltage appears across the air gap capacitor. With Mylar, the leakage resistance is similar or larger than that of air so a large portion of the EHT voltage appears across the stator insulation capacitance and the remainder across the air gap capacitance. Remember SPL is proportional to the voltage across the air gap.
http://www.diyaudio.com/forums/plan...truct-cube-louver-acoustat-7.html#post2154621
The stator tape capacitance is much larger than the air gap capacitor. So, when the EHT is disconnected and diaphragm connected to CT, the larger stator capacitor now acts like a voltage source and discharges the air gap capacitance and then charges it with the opposite polarity to a voltage equal to its own. This is what causes the polarity swap.
I will make a few illustrations later today to better explain.
Perhaps 10 seconds. However I Think the EHT (which is a high voltage tester actually) discharges rather quickly when turned off. Perhaps I should retest with my ladder type EHT.
Any illustrations would be great, thanks!!
Attachment #1: For the best test, you would not discharge the EHT supply at all.
Step 1 ==> play pink noise thru ESL with EHT supply connected for 5 minutes.
Step 2 ==> with pink noise still playing, disconnect EHT supply.
Step 3 ==> with pink noise still playing, connect CT to diaphragm.
- If minimal charge is being stored in the stator insulation dielectric, the SPL will continue to decrease with time.
- If significant charge is being stored in the stator insulation dielectric, the SPL will decrease at first reaching a minimum and then start building back up in level with reversed polarity.
Attachment #2: Here is a schematic of a lumped model representation of an ESL bias circuit. Most wire stator ESLs use PVC insulation which is pretty leaky as far as insulation goes with bulk resistivity much less than air. In this case(shown on the left) the majority of the EHT supply voltage appears across the air gap capacitor where it contributes to sound output. If “better” insulation is used like Teflon or Mylar, the insulation resistance is similar in magnitude or even higher than that of air. In this case(shown on the right) a significant percentage of the EHT supply voltage will appear across the insulation dielectric, leaving only a portion of the EHT voltage across the air gap where it needs to be to contribute to sound output.
Attachment #3: This is a time history of voltages to show what is happening when you perform the test shown in Attachment #1. The voltage values and trends aren’t meant to be an exact model of anything in particular; they are for illustrative purposes only. Cd was about 20x greater than Ca, which is pretty typical. Rd was just a bit more than Ra. Note that it is the relative values of Rd and Ra that determine the voltages across Cd and Ca(not the values of Cd & Ca).
Prior to time = t1, the EHT supply voltage of 5kV is connected to the diaphragm and voltage has distributed between the air gap and the dielectric capacitors. In this case, the insulation resistance is a bit more than that of the air leaving only 2kV of the applied 5kV across the air gap.
Once the EHT connection has been removed and the diaphragm is connected to the CT, the current flow reverses direction as the energy stored in Cd charges Ca to an equal and opposite voltage during the time period t1-to-t2. An equilibrium is reached at t2 with ~2.5kV across each capacitor.
For times after t2, the resistance in parallel with each capacitor will slowly discharge them. Any imbalance in voltage across the capacitors will result in current flow from the Cd capacitor to restore equilibrium. How slowly the voltage and SPL decays will depend on the values of the leakage resistances and how much charge had been stored in Cd.
Note that ESLs with high resistance diaphragm coatings will behavior differently from the lumped model described above. However, the general voltage and SPL trends will be similar.
Attachments
Last edited:
That was very informative!
I did retest this, disconnecting the EHT and connecting the CT in under 1 second. No discharging at all. I forgot the microphone at home, but my impression is the same. The SPL goes down a lot and stays there. I may have some leakage that I have not considered. I Think the charge on the film disappears quickly. I dont get the spark when I connect the CT to the film.
I will try to remove the insulation on the stators in one of the four Squares, shouldnt be impossible. Will there be any problems with the film having more charge in one of the Squares?
I did retest this, disconnecting the EHT and connecting the CT in under 1 second. No discharging at all. I forgot the microphone at home, but my impression is the same. The SPL goes down a lot and stays there. I may have some leakage that I have not considered. I Think the charge on the film disappears quickly. I dont get the spark when I connect the CT to the film.
I will try to remove the insulation on the stators in one of the four Squares, shouldnt be impossible. Will there be any problems with the film having more charge in one of the Squares?
Based on your latest test description, I would not remove the insulation yet.
I just realized that your 4 layers of Mylar insulation have adhesive backing in between each layer. It is likely that the adhesive is not nearly as good an insulator as Mylar. This may be keeping the Mylar layers from storing as much charge as I had been thinking based on my experiments with Teflon insulation. Your measurement of polarity reversal back in post#19 does clearly show the Mylar is storing some charge, but it may be a small enough amount to not significantly affect your SPL output capability.
Concerning possible leakage:
Did the SPL start going down as soon as you removed the EHT connection? even before you connected the CT? or did the SPL stay roughly the same until you connected the CT. You should not need to rush between Steps 1 and 3. An ESL with minimal leakage should not lose SPL quickly when EHT connection is removed. I can play mine with EHT disconnected for a minute or more depending on humidity level and lose only -3dB of SPL.
You might consider adding a blinking neon light charge indicator in line with your diaphragm connection so that you can observe charging and discharging rates. This will help test for leakage and diaphragm coating conductivity.
http://www.diyaudio.com/forums/planars-exotics/76222-esl-build-thread-12.html#post952272
http://www.diyaudio.com/forums/planars-exotics/153773-esl-technical-questions.html#post1958015
http://www.diyaudio.com/forums/plan...ogan-sequel-ii-dead-panels-7.html#post3731257
I just realized that your 4 layers of Mylar insulation have adhesive backing in between each layer. It is likely that the adhesive is not nearly as good an insulator as Mylar. This may be keeping the Mylar layers from storing as much charge as I had been thinking based on my experiments with Teflon insulation. Your measurement of polarity reversal back in post#19 does clearly show the Mylar is storing some charge, but it may be a small enough amount to not significantly affect your SPL output capability.
Concerning possible leakage:
Did the SPL start going down as soon as you removed the EHT connection? even before you connected the CT? or did the SPL stay roughly the same until you connected the CT. You should not need to rush between Steps 1 and 3. An ESL with minimal leakage should not lose SPL quickly when EHT connection is removed. I can play mine with EHT disconnected for a minute or more depending on humidity level and lose only -3dB of SPL.
You might consider adding a blinking neon light charge indicator in line with your diaphragm connection so that you can observe charging and discharging rates. This will help test for leakage and diaphragm coating conductivity.
http://www.diyaudio.com/forums/planars-exotics/76222-esl-build-thread-12.html#post952272
http://www.diyaudio.com/forums/planars-exotics/153773-esl-technical-questions.html#post1958015
http://www.diyaudio.com/forums/plan...ogan-sequel-ii-dead-panels-7.html#post3731257
Last edited:
I had some time and made another, smaller panel. This one is 3x3" and easier to make. It uses perforated steel stators. It is one with 10mm Square holes and 5mm steel between the holes. Roughly 44% open area. I Think the holes are a bit big, but that was what I had and it was Quick to make.
I didnt insulate it. Only D/S spacing with tape. I used the same Shackman coating and High voltage source as Before.
Results are still quite bad. The efficiency is a bit better actually. Still 8-10dB off in the low range compared to simulation. I tested removing the EHT and they play quite nicely for a long time as long as the film is charged, the old ones lost SPL immediately. I also checked the polarity of the impulse when I connect the film directly to the transformers (removing EHT). Now I get an impluse that is the same polarity as when the EHT is connected. I suppose that means my charge is not trapped in the insulation (there is none...).
I suspected the coating, and made a panel without coating to see that there is a difference. And yes, without coating there is no output at all. I am still wondering if the coating is my problem.
I dont seem to have much leakage of the DC bias.
I connected the EHT directly to the film so I know there is no loss of voltage in the protective resistors.
I also realized that I put the EHT Cable on the film without any charging ring. Just 7mm or so of wire flat on the film. But the film is coated all the way to the edges.
Is the Shackman VLC07 a bad or good coating? I read some bad comments here on the forum.
I didnt insulate it. Only D/S spacing with tape. I used the same Shackman coating and High voltage source as Before.
Results are still quite bad. The efficiency is a bit better actually. Still 8-10dB off in the low range compared to simulation. I tested removing the EHT and they play quite nicely for a long time as long as the film is charged, the old ones lost SPL immediately. I also checked the polarity of the impulse when I connect the film directly to the transformers (removing EHT). Now I get an impluse that is the same polarity as when the EHT is connected. I suppose that means my charge is not trapped in the insulation (there is none...).
I suspected the coating, and made a panel without coating to see that there is a difference. And yes, without coating there is no output at all. I am still wondering if the coating is my problem.
I dont seem to have much leakage of the DC bias.
I connected the EHT directly to the film so I know there is no loss of voltage in the protective resistors.
I also realized that I put the EHT Cable on the film without any charging ring. Just 7mm or so of wire flat on the film. But the film is coated all the way to the edges.
Is the Shackman VLC07 a bad or good coating? I read some bad comments here on the forum.
So, I seem to finally be going somewhere.
This is the small panel I made that doesnt measure well. This is at 10cm (4") with 400V bias and 360Vrms on the stators. It is not insulated in any way.
It is quite far from as effeicient as it should be.
What I did next was to spray my Shackman coating on the uncoated panel, but I didnt polish it at all (you are supposed to polish like crazy accordning to the instructions)
I used a piece of copper tape to charge the film.
This looks much better. Now I am some dBs off from the simulations, which is what I was expecting. Woohoo!
This is the 2cm measurement. This also is quite Close to the simulation.
Either the coating it self has been the problem, or the termination of the EHT to the film has been bad. Another possibility has to do with leaking. The previous atempts have all had coating all the way to the edge of the film. This may have leaked charge.
Thanks for all the help I got in this thread!
To be continued...
This is the small panel I made that doesnt measure well. This is at 10cm (4") with 400V bias and 360Vrms on the stators. It is not insulated in any way.
It is quite far from as effeicient as it should be.
What I did next was to spray my Shackman coating on the uncoated panel, but I didnt polish it at all (you are supposed to polish like crazy accordning to the instructions)
I used a piece of copper tape to charge the film.
This looks much better. Now I am some dBs off from the simulations, which is what I was expecting. Woohoo!
This is the 2cm measurement. This also is quite Close to the simulation.
Either the coating it self has been the problem, or the termination of the EHT to the film has been bad. Another possibility has to do with leaking. The previous atempts have all had coating all the way to the edge of the film. This may have leaked charge.
Thanks for all the help I got in this thread!
To be continued...
First, congratulations on progress.
Concerning the coating and termination:
1) Did you try exhaling or breathing warm moist air on to the polished coating to see if the SPL went up as suggested in 1) of post #18?
If SPL didn’t increase, the termination is more likely the cause.
2) The copper tape you use for the termination, does it use conductive adhesive?
If not, you may not to apply addition coating along the edge of the tape after you lay it down.
Pretty cool confirmation of the principles involved in the behavior of stator insulation.
Thank you!
1. Yep, I tried breathing on it and couldnt notice any big difference to be honest. But then again we had really high humidity anyway when I did those tests. So while I also Think that the termination have been a problem, the coating hasnt been perfect.
This is the speaker with Lycron spray as coating. No polishing and easy to apply, I like the stuff. I Think the Shackman spray works, but it is much more difficult to use.
Now I did a complete copper ring around the film. The film is 9x9cm and I masked the edge so that only 8x8cm is coated. Copper ring is 8x8cm outer dimension.
2. My tape does have non-conductive adhesive, but I put the copper side against the film.
The result is that up to 3kHz or so the simulation and reality are very close. Very cool.
I checked the impulse just as a sanity check. If it was inverted I Think I would have picked up hand knitting dry suits for diving as a hobby instead of speakers...
Any thoughts on if there is a big difference in SPL/diaphragm size with perforated sheets versus wire stators? I want a small speaker that still fills a room with Music (reasonably loud).
1. Yep, I tried breathing on it and couldnt notice any big difference to be honest. But then again we had really high humidity anyway when I did those tests. So while I also Think that the termination have been a problem, the coating hasnt been perfect.
This is the speaker with Lycron spray as coating. No polishing and easy to apply, I like the stuff. I Think the Shackman spray works, but it is much more difficult to use.
Now I did a complete copper ring around the film. The film is 9x9cm and I masked the edge so that only 8x8cm is coated. Copper ring is 8x8cm outer dimension.
2. My tape does have non-conductive adhesive, but I put the copper side against the film.
The result is that up to 3kHz or so the simulation and reality are very close. Very cool.
I checked the impulse just as a sanity check. If it was inverted I Think I would have picked up hand knitting dry suits for diving as a hobby instead of speakers...
Any thoughts on if there is a big difference in SPL/diaphragm size with perforated sheets versus wire stators? I want a small speaker that still fills a room with Music (reasonably loud).
Excellent. I am glad to hear that your results are getting closer to what they should be.
Licron Crystal is Excellent, and it seems that maybe your original coating could have been less than perfect and not working as it should have.
SPL and Diaphragm size is the same as it is for regular dynamic speakers only you must compensate for dipole cancellations.
Displacement is the key factor.
As true with dynamic drivers and ESL's, in order to maintain the same SPL for the next octave lower the diaphragm has to move 4 times more the distance (Xmax * 4).
I have some spreadsheets to help calculate this as well for this.
One is from the Linkwitz Labs site and the other is from the Audio circuit.
https://www.google.com/url?sa=t&rct...xIGgAw&usg=AFQjCNH1QR1HeW1veDd2cgjFxR9EneC8Kg
Links to both spreadsheets can be found on this page as well in sectio 1.2.1 ,
Do It Yourself - Electrostatic Speakers - Project: ESL-220-30 by Marc Schroeyers
And this is the one I use for a reference for monopole systems,
Piston Excursion calculator
I have a few other versions as well but they are all pretty within each other.
I had been wanting to do a few tests on the differences of efficiency vs open area, But I just never got that far too build so many panels to test with.
Mainly my focus was to get the most efficiency I could get out of one size using my form of construction and then applying that to a larger panel, and it will just follow as size increases.
Just as it does when you start increasing the number of drivers in any other speaker system.
However with each successive build my open area got greater as well and I did notice that each next one sound better and better because of that.
My very First panel had something like 12% to 20% open area using some perforated Aluminium Soffit material that I found at Home Depot by the time I got it coated.
It muffled the sound a lot compared to later designs of the exact same size with window screen with at least double the open area.
I never got to measure the difference though.
My very First panel was about 3.5" X 9" and my very First larger panel was about 4 times as big at 19" x 7.5" (overall dimensions) with this concept in mind and is why I choose the two sizes.
This should have given me 6db (or maybe 12db, I forget) more efficiency compared to my single little panel.
But those bigger panels are not operable anymore nor can they run at the same voltage level that I have now gotten my little panels to run at, So I never tested them for the difference.
FWIW
jer
Here are a couple more threads on this subject,
http://www.diyaudio.com/forums/multi-way/1035-max-spl-calculation.html#post7136
http://www.diyaudio.com/forums/multi-way/165765-simple-driver-spl-equation-needed.html#post2165845
Licron Crystal is Excellent, and it seems that maybe your original coating could have been less than perfect and not working as it should have.
SPL and Diaphragm size is the same as it is for regular dynamic speakers only you must compensate for dipole cancellations.
Displacement is the key factor.
As true with dynamic drivers and ESL's, in order to maintain the same SPL for the next octave lower the diaphragm has to move 4 times more the distance (Xmax * 4).
I have some spreadsheets to help calculate this as well for this.
One is from the Linkwitz Labs site and the other is from the Audio circuit.
https://www.google.com/url?sa=t&rct...xIGgAw&usg=AFQjCNH1QR1HeW1veDd2cgjFxR9EneC8Kg
Links to both spreadsheets can be found on this page as well in sectio 1.2.1 ,
Do It Yourself - Electrostatic Speakers - Project: ESL-220-30 by Marc Schroeyers
And this is the one I use for a reference for monopole systems,
Piston Excursion calculator
I have a few other versions as well but they are all pretty within each other.
I had been wanting to do a few tests on the differences of efficiency vs open area, But I just never got that far too build so many panels to test with.
Mainly my focus was to get the most efficiency I could get out of one size using my form of construction and then applying that to a larger panel, and it will just follow as size increases.
Just as it does when you start increasing the number of drivers in any other speaker system.
However with each successive build my open area got greater as well and I did notice that each next one sound better and better because of that.
My very First panel had something like 12% to 20% open area using some perforated Aluminium Soffit material that I found at Home Depot by the time I got it coated.
It muffled the sound a lot compared to later designs of the exact same size with window screen with at least double the open area.
I never got to measure the difference though.
My very First panel was about 3.5" X 9" and my very First larger panel was about 4 times as big at 19" x 7.5" (overall dimensions) with this concept in mind and is why I choose the two sizes.
This should have given me 6db (or maybe 12db, I forget) more efficiency compared to my single little panel.
But those bigger panels are not operable anymore nor can they run at the same voltage level that I have now gotten my little panels to run at, So I never tested them for the difference.
FWIW
jer
Here are a couple more threads on this subject,
http://www.diyaudio.com/forums/multi-way/1035-max-spl-calculation.html#post7136
http://www.diyaudio.com/forums/multi-way/165765-simple-driver-spl-equation-needed.html#post2165845
Last edited:
I Think the coating I used Before is actually usable, if you are very careful. My dad used it on his Audiostatics and they perform quite well now. Hard to tell if they are as good as new, and I dont have any new ones to test for comparison. The key difference is that his diaphragm was big, so he just smudged the coating over the diaphragm. Polishing it off (like I probably did) would take too much time for him. I also Think that I did the termination incorrectly. So I am not saying the Shackman coating is bad, but it wasnt easy to use compared to the Licron.
Your builds have been a big inspiration, and I started messing with ESLs because I saw how you seemed to get good SPLs from small speakers. And I did use some of the spreadsheets and simulations to figure out what was possible. After that I started to actually build something.
I havent thought much about open area, that is also something to test.
My concern so far has been insulation. I just painted my perforated stators with acrylic clear coat. I would like to use high voltages and I wonder if the perforated or wire stators will handle that better. I like the Construction of the perforated stators. But since I already made wire stators I will make a new test panel with Licron coating and good termination. The stators will also be painted with acrylic clear.
I am having a lot of fun, and thanks to the tiny perforated stators I make a speaker in under an hour. And I dont use much material either...
Next step is to use insulated stators. That is going to take a bit more time, but I feel more Confident with every step.
I appreciate all help, links and thoughts in this thread, very helpful! I have understood things much quicker than I thought I would.
Good Deal.....My soul intent is to inspire DIYer's to give this technology a try!!
And, take the mystery out of how, why and what makes them work better.
The one thing that smaller panels of going against them is the smaller surface area.
So this is why I had to bring the voltages up so high in order to get good efficiency out of them.
As you know this opens up another can of worms and another can of paint!! He,he,he
But it shows that it can be done and just imagine that once you get that far and apply the techniques to a larger panel what a world of difference this would be rather than just having a big system that looks good and soundsss.....Well,..... okay!
I was looking for that Big sound!!!
I figured that if my 4" midrange drivers (4 total, 2 per side) can approach +110 in my room than My little panel ought to be able to do that too!!
It had the same surface area of not more and it definitely had more Xmax than the dynamic drivers did!!!!!?
This would also keep me from having to keep replacing those expensive little buggers every time I got antsy and fried them!
It was a challenge, But you know much to my amazement, they really can perform at that level !!!!
When it comes to Clear Acrylic I would suggest to use only the Red Primer as well if you can tolerate the color.
Use only one color coat if you have to, on the primer, then the rest clear.
Just the red stuff, as it is loaded with Talc and it has a high Dielectric Constant as well this is different than Dielectric strength.
I have not made a A&B test for performance with using it or not yet, but technically it is beneficial besides its qualities for proper painting technique.
The bond it creates for the paint to adhere to is much greater than just the clear on bare metal.
In one of my stress tests the static force at 13.8kv was so high that it pulled the clear clean off of the 1/16" TIG rod sample.
I never made a picture of this and I was going to, sorry about that.
But not only was there a perfectly round hole about a 1/32" in diameter or so in the 8-10mil layer of paint, the inside wall of the hole was straight and not burn't and the top edge as a sharp as a piece of broken glass.
Not to mention that there were stress cracks leading form the edge of the cavity some 1/8" to 1/4" long!
Just like a piece of safety glass looks in a bullet goes through it.
This didn't happen with any of my samples that had a primer base.
Also when it come to Wire vs Perforated, wire has little to no sharp edges at all.
The holes in metal all have a sharp edges, primer helps to better seal and round out those sharp edges on the holes.
On the side that punch went through the metal the holes edges will be more rounded and this is the side you want to have towards the diaphragm.
It is imperative to use a primer on this material as helps to coat and round out the edges.
See these pictures and thread for more info,
http://www.diyaudio.com/forums/plan...tric-coatings-fact-fiction-2.html#post2893839
and here,
http://www.diyaudio.com/forums/plan...tric-coatings-fact-fiction-2.html#post2902754
I never have tried perforated metal except for the soffit material I mentioned due to the cost, But I now have some that was sent to me left over from a fellow DIYer's previous build, So I will be making some "4 x 48" panels out of those sometime soon.
I always say soon but it has already been 2 years since I built my last panel and I have yet to mount a diaphragm and show the test results!!!! He,he,he,he
Sorry, I just have too many projects going on at once.
When you say tiny holes, Remember that the coating is going to make them smaller and you have to take that in to account for your open area.
That is why my soffit material didn't work so good and the holes ended up considerably smaller than they were when it was just a fresh piece of metal.
For reference they were 3/32" staggered holes on 7/64" centers.
Cheers!!
jer
And, take the mystery out of how, why and what makes them work better.
The one thing that smaller panels of going against them is the smaller surface area.
So this is why I had to bring the voltages up so high in order to get good efficiency out of them.
As you know this opens up another can of worms and another can of paint!! He,he,he
But it shows that it can be done and just imagine that once you get that far and apply the techniques to a larger panel what a world of difference this would be rather than just having a big system that looks good and soundsss.....Well,..... okay!
I was looking for that Big sound!!!
I figured that if my 4" midrange drivers (4 total, 2 per side) can approach +110 in my room than My little panel ought to be able to do that too!!
It had the same surface area of not more and it definitely had more Xmax than the dynamic drivers did!!!!!?
This would also keep me from having to keep replacing those expensive little buggers every time I got antsy and fried them!
It was a challenge, But you know much to my amazement, they really can perform at that level !!!!
When it comes to Clear Acrylic I would suggest to use only the Red Primer as well if you can tolerate the color.
Use only one color coat if you have to, on the primer, then the rest clear.
Just the red stuff, as it is loaded with Talc and it has a high Dielectric Constant as well this is different than Dielectric strength.
I have not made a A&B test for performance with using it or not yet, but technically it is beneficial besides its qualities for proper painting technique.
The bond it creates for the paint to adhere to is much greater than just the clear on bare metal.
In one of my stress tests the static force at 13.8kv was so high that it pulled the clear clean off of the 1/16" TIG rod sample.
I never made a picture of this and I was going to, sorry about that.
But not only was there a perfectly round hole about a 1/32" in diameter or so in the 8-10mil layer of paint, the inside wall of the hole was straight and not burn't and the top edge as a sharp as a piece of broken glass.
Not to mention that there were stress cracks leading form the edge of the cavity some 1/8" to 1/4" long!
Just like a piece of safety glass looks in a bullet goes through it.
This didn't happen with any of my samples that had a primer base.
Also when it come to Wire vs Perforated, wire has little to no sharp edges at all.
The holes in metal all have a sharp edges, primer helps to better seal and round out those sharp edges on the holes.
On the side that punch went through the metal the holes edges will be more rounded and this is the side you want to have towards the diaphragm.
It is imperative to use a primer on this material as helps to coat and round out the edges.
See these pictures and thread for more info,
http://www.diyaudio.com/forums/plan...tric-coatings-fact-fiction-2.html#post2893839
and here,
http://www.diyaudio.com/forums/plan...tric-coatings-fact-fiction-2.html#post2902754
I never have tried perforated metal except for the soffit material I mentioned due to the cost, But I now have some that was sent to me left over from a fellow DIYer's previous build, So I will be making some "4 x 48" panels out of those sometime soon.
I always say soon but it has already been 2 years since I built my last panel and I have yet to mount a diaphragm and show the test results!!!! He,he,he,he
Sorry, I just have too many projects going on at once.
When you say tiny holes, Remember that the coating is going to make them smaller and you have to take that in to account for your open area.
That is why my soffit material didn't work so good and the holes ended up considerably smaller than they were when it was just a fresh piece of metal.
For reference they were 3/32" staggered holes on 7/64" centers.
Cheers!!
jer
I have one layer of grey primer and 4mils of clear coat on my perforated stator now. I am out of paint, so I Think I will build the speaker tomorrow.
One problem is that I cant find red primer. We used to have it here, but nowadays it seems all primers are grey. The only red ones we have are rust remedies/primers. I dont know what is in that. We dont have labels with all ingredients on the cans here. Only the safety stuff.
Should be interesting to see how it works. I will measure and then raise the voltage, remeasure, raise the voltage and so on until the panel doesnt like it anymore.
Updates will come along the way.
One problem is that I cant find red primer. We used to have it here, but nowadays it seems all primers are grey. The only red ones we have are rust remedies/primers. I dont know what is in that. We dont have labels with all ingredients on the cans here. Only the safety stuff.
Should be interesting to see how it works. I will measure and then raise the voltage, remeasure, raise the voltage and so on until the panel doesnt like it anymore.
Updates will come along the way.
The only gray Primer that you should use is a metal etching type for the first coat on bare metal and after that use the red stuff.
Only because I have yet to find that type that doesn't have a gray pigment.
Gray primers have Titanium Dioxide for a pigment and it will conduct at high voltages, Same goes for Black primers that use Carbon Black for a pigment.
Red Primer doesn't have any of these pigments so red is okay.
It is just loaded with Talc and this is even better.
The very reason that I can't get my original panels to not arc at bias and signal voltages above about 3.5kv is because I used a white pigmented paint to coat those.
No matter how thick I made it (in some areas the holes were almost completely closed) it would still arc even in the thickest spots.
The rust remedies are okay as I think that it is just a sales pitch, or it may contain an acid, like the metal etching types do, is all.
You can sometimes find the MSDS and technical data sheets on line for whatever you find on the shelf, and, it will list most all of the ingredients especially the fillers and pigments like the ones I mentioned as they are considered dangerous, so they have to list them.
4 mil of clear will be good for about 6-7kv.
I have measured as high as 1900v to 2100v per mil at flashover for the clears and as high as 2200v per mil for just the red primer alone.
So as use 1500v per mil as a general guarantied safe limit for no arc over.
The sample test screen that I coated survived 13.8Kv (the limit of my regulated supply), and it's coating thickness is a about 8 mil or so, I was very lucky to get that impressive picture.
I still have the very same sample from 3 years ago and I get it out every once in a while and test it and it still works just the same.
jer
Only because I have yet to find that type that doesn't have a gray pigment.
Gray primers have Titanium Dioxide for a pigment and it will conduct at high voltages, Same goes for Black primers that use Carbon Black for a pigment.
Red Primer doesn't have any of these pigments so red is okay.
It is just loaded with Talc and this is even better.
The very reason that I can't get my original panels to not arc at bias and signal voltages above about 3.5kv is because I used a white pigmented paint to coat those.
No matter how thick I made it (in some areas the holes were almost completely closed) it would still arc even in the thickest spots.
The rust remedies are okay as I think that it is just a sales pitch, or it may contain an acid, like the metal etching types do, is all.
You can sometimes find the MSDS and technical data sheets on line for whatever you find on the shelf, and, it will list most all of the ingredients especially the fillers and pigments like the ones I mentioned as they are considered dangerous, so they have to list them.
4 mil of clear will be good for about 6-7kv.
I have measured as high as 1900v to 2100v per mil at flashover for the clears and as high as 2200v per mil for just the red primer alone.
So as use 1500v per mil as a general guarantied safe limit for no arc over.
The sample test screen that I coated survived 13.8Kv (the limit of my regulated supply), and it's coating thickness is a about 8 mil or so, I was very lucky to get that impressive picture.
I still have the very same sample from 3 years ago and I get it out every once in a while and test it and it still works just the same.
jer
Ok, I measured, raised the voltage and measured again. I had no way of measuring the bias voltage this time, I could only Count the number on diodes in my ladder. With low bias I had similar efficiency to the panel without insulation 7-800V with two diodes. Great!
Then I tried six diodes, the efficiency went up as expected. I tried twelve diodes and got another six dB or so. After this I played a bit of Music again, and it works quite well. At this time I was happy with the results, thinking that this was probably the safe bias voltage for this panel. I Think it was close to 4kV.
Next step I tried 16 diodes and the efficiency went up a Little again. But when I started playing Music I got a spark. So the panel i scrap now. But I got the result I was after. I am using a 1kW 4ohm amp, but output from my laptop means I had a bit more than 20Vrms from the amp, so 2500-3kVrms out from the transformers. 4kV bias +2kVpeak =6kV between film and stator. If I had a thin spot or any dirt in the paint I am not suprised that it sparked. I was surprised by the SPL I got from this relatively low bias and tiny panel.
I still have the wire stators with the tape insulation. I will use them to test the tape again now that I have the coating and bias termination under control. It is a relatively simple test to do. Spray painting stators is a problem for me since I have nowhere to do it. Ventilation is a problem...
Perhaps heat shrink/tubing or a cable stator can work for me.
Then I tried six diodes, the efficiency went up as expected. I tried twelve diodes and got another six dB or so. After this I played a bit of Music again, and it works quite well. At this time I was happy with the results, thinking that this was probably the safe bias voltage for this panel. I Think it was close to 4kV.
Next step I tried 16 diodes and the efficiency went up a Little again. But when I started playing Music I got a spark. So the panel i scrap now. But I got the result I was after. I am using a 1kW 4ohm amp, but output from my laptop means I had a bit more than 20Vrms from the amp, so 2500-3kVrms out from the transformers. 4kV bias +2kVpeak =6kV between film and stator. If I had a thin spot or any dirt in the paint I am not suprised that it sparked. I was surprised by the SPL I got from this relatively low bias and tiny panel.
I still have the wire stators with the tape insulation. I will use them to test the tape again now that I have the coating and bias termination under control. It is a relatively simple test to do. Spray painting stators is a problem for me since I have nowhere to do it. Ventilation is a problem...
Perhaps heat shrink/tubing or a cable stator can work for me.
That sounds about right and along the same results I got, Congtrat's on your work!!
The more stages you have there is also a bit of loss that increases in the divider per number of stages as well due to internal component resistances and impedance's.
Here is a calculator that I find quite useful,
Cockcroft walton Voltage Multiplier Calculator
Like I had said before there is a practical limit.
As you keep increasing the bias this increases the efficiency and this is great.
But, increasing the bias also brings your combined voltage of Bias+Signal closer to your insulation's point of limit for arc breakdown.
So your maximum voltage is still limited by the insulation material and the ionization of the air.
The cool thing is that if the resistance of the stator coating is high enough then you can still run the voltages high enogh to get the maximum eficiency you can to the point of ionizing the air in the gap and not have any high current arc over flash's that cause damage to the stator coating and burn holes or even completely burn't diaphragms.
Much to my amazement I have done this and found no burn holes in one of my diaphragms even after careful inspection with a magnifying glass, and, I had the thing lit up like Neon bulb for quite some time!!
Now at the time I was running so high to have ionization in the gap I didn't have a SPL meter at the time, but I am suspecting that some form of compression is starting to occur with the sound at this point.
Since ESL's work because of static charge they take little or no current to operate.
The resistance of the coating and the resistance of the air form a voltage divider as explained in some earlier thread, and it is said that the ionization of the air is the absolute limit.
The nice thing about using a good coating is that yes maybe you do have to raise the voltages up a bit more compared to that of a bare stator but using the coating acts as a current limiting resistor should a flashover occur and this keeps the current in the arc low enough to keep it from destroying things.
This is something that is a flaw in the older ESL's back in the day (57's and 63's?).
I have been wanting to test this theory of coated vs non-coated stators and efficiency and Peak SPL.
But, I have already experienced the difference of two exact sized panels already coated and one coating that can't get past a 3.5kV of bias plus a moderate signal and one that can tolerate as high as a 8.5kv and even a 10 kv of bias plus the very same moderate signal.
The difference was more than just night and day!!
It was more like a cloudy night and a sunny day at the beach!!!
Also the dynamics of the sound seemed to be much greater (IMHO) with a higher bias voltage even when the signal was reduced to the same SPL to compensate for increased efficiency in that test.
These are my observation and once I had seemed to find the Holy Grail I decided that it was to costly and time consuming to keep making more panels to test all of this again right now.
Since I burned up my best panels doing this, This is why I made a new one that can handle these type of voltage limit's and start enjoying them for a change and take a break from my 11 years of off and on research.
Besides this gives me the base platform for my original goal, A Direct Drive Amplifier.
Tubes that are suitable for this phase of my project aren't exactly readily available anymore as they were when I started this project and semiconductors definitely won't be able to survive a panel that arcs either!!
This is essential for a design the uses a high bias voltage and a solid state amplifier at supply voltages that aren't too over extreme in about a 1.5kv to 3.5kv range considering you can get a decent sound (+90db) with only a 500v to 700v drive signal to the stators.
I figure it would take a push pull amp with a 6kv supply for 12kv p-p for an over the edge type of system.
It is true!!!
It is just very hard to beat those transformers at times!!
But it can be done !!
Cheers !!!
jer
The more stages you have there is also a bit of loss that increases in the divider per number of stages as well due to internal component resistances and impedance's.
Here is a calculator that I find quite useful,
Cockcroft walton Voltage Multiplier Calculator
Like I had said before there is a practical limit.
As you keep increasing the bias this increases the efficiency and this is great.
But, increasing the bias also brings your combined voltage of Bias+Signal closer to your insulation's point of limit for arc breakdown.
So your maximum voltage is still limited by the insulation material and the ionization of the air.
The cool thing is that if the resistance of the stator coating is high enough then you can still run the voltages high enogh to get the maximum eficiency you can to the point of ionizing the air in the gap and not have any high current arc over flash's that cause damage to the stator coating and burn holes or even completely burn't diaphragms.
Much to my amazement I have done this and found no burn holes in one of my diaphragms even after careful inspection with a magnifying glass, and, I had the thing lit up like Neon bulb for quite some time!!
Now at the time I was running so high to have ionization in the gap I didn't have a SPL meter at the time, but I am suspecting that some form of compression is starting to occur with the sound at this point.
Since ESL's work because of static charge they take little or no current to operate.
The resistance of the coating and the resistance of the air form a voltage divider as explained in some earlier thread, and it is said that the ionization of the air is the absolute limit.
The nice thing about using a good coating is that yes maybe you do have to raise the voltages up a bit more compared to that of a bare stator but using the coating acts as a current limiting resistor should a flashover occur and this keeps the current in the arc low enough to keep it from destroying things.
This is something that is a flaw in the older ESL's back in the day (57's and 63's?).
I have been wanting to test this theory of coated vs non-coated stators and efficiency and Peak SPL.
But, I have already experienced the difference of two exact sized panels already coated and one coating that can't get past a 3.5kV of bias plus a moderate signal and one that can tolerate as high as a 8.5kv and even a 10 kv of bias plus the very same moderate signal.
The difference was more than just night and day!!
It was more like a cloudy night and a sunny day at the beach!!!
Also the dynamics of the sound seemed to be much greater (IMHO) with a higher bias voltage even when the signal was reduced to the same SPL to compensate for increased efficiency in that test.
These are my observation and once I had seemed to find the Holy Grail I decided that it was to costly and time consuming to keep making more panels to test all of this again right now.
Since I burned up my best panels doing this, This is why I made a new one that can handle these type of voltage limit's and start enjoying them for a change and take a break from my 11 years of off and on research.
Besides this gives me the base platform for my original goal, A Direct Drive Amplifier.
Tubes that are suitable for this phase of my project aren't exactly readily available anymore as they were when I started this project and semiconductors definitely won't be able to survive a panel that arcs either!!
This is essential for a design the uses a high bias voltage and a solid state amplifier at supply voltages that aren't too over extreme in about a 1.5kv to 3.5kv range considering you can get a decent sound (+90db) with only a 500v to 700v drive signal to the stators.
I figure it would take a push pull amp with a 6kv supply for 12kv p-p for an over the edge type of system.
It is true!!!
It is just very hard to beat those transformers at times!!
But it can be done !!
Cheers !!!
jer
I guess my Question is that if the ionization of the air is what limits our Peak SPL, can the efficiency still increase as long as we can contain the bias voltage without arcing when it is raised?
I have made it to 12kv with some leakage starting but it ran and it was even more efficient.
But can we take it to say 15Kv or even 20kv or so of bias and get away with only using like 100v to 400v of drive signal or less and still have an ample amount of SPL?
FWIW
jer
I have made it to 12kv with some leakage starting but it ran and it was even more efficient.
But can we take it to say 15Kv or even 20kv or so of bias and get away with only using like 100v to 400v of drive signal or less and still have an ample amount of SPL?
FWIW
jer
You had 12kV bias with how much D/S?
Simultating it at quadesl.nl seems to give a maximum when the drive voltage is a bit below the bias voltage. Say 3k5V drive and 5kV bias. I havent thought of the actual voltage between the film and the stator, the drive is in RMS and shared on the two stators. It should be easy to calculate, but time for bed now!
Simultating it at quadesl.nl seems to give a maximum when the drive voltage is a bit below the bias voltage. Say 3k5V drive and 5kV bias. I havent thought of the actual voltage between the film and the stator, the drive is in RMS and shared on the two stators. It should be easy to calculate, but time for bed now!
Believe it or not about .072" -+ 3mil !!
http://www.diyaudio.com/forums/planars-exotics/109789-esl-diaphragm-coating-8.html#post2160477
This link shows the two different sized panels I had made,
http://www.diyaudio.com/forums/planars-exotics/158115-material-esl-2.html#post2075441
When I state voltages I am usually referring to peak voltages as that is what matters when it comes to the puncture voltage.
jer
http://www.diyaudio.com/forums/planars-exotics/109789-esl-diaphragm-coating-8.html#post2160477
This link shows the two different sized panels I had made,
http://www.diyaudio.com/forums/planars-exotics/158115-material-esl-2.html#post2075441
When I state voltages I am usually referring to peak voltages as that is what matters when it comes to the puncture voltage.
jer
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.