hi all.
I have made some els but i have a problem, the sensivity is very very low, i guess it´s about max 90db with 150w🙁 my own guees is that i maybe have way too much tension on the diaphragm because i can not see any movement from the diaphragm at all then music playing and thats with no croosover and the sound is bright, the size of the panel are 100cm*24cm i have made them curved and D/S is 1.5mm and the distance between the spacers are 12cm, mylar is 6um and stepup trafo have 1:80 ratio, stators are made of 1.5mm steel with 58% air and isolation is made with emalje paint and 2 component polyester resin.
Then i remove the biasvoltage wire the speaker keeps playing in about 20-25 min so i chould not have any leakage from the coating (varmason esd floor polish) and then i turn up the bias voltage from about 2kv to 6,5kv there is not much difference in sound level.
The diaphragm and it´s here i think the problem maybe is, are stretched so much that i right between 2 of the spacers need to put 50g of weight before the diaphragm touch the stator.
hope i have give enough information to your guys to maybe help me with my problem...
regards
uglykid
I have made some els but i have a problem, the sensivity is very very low, i guess it´s about max 90db with 150w🙁 my own guees is that i maybe have way too much tension on the diaphragm because i can not see any movement from the diaphragm at all then music playing and thats with no croosover and the sound is bright, the size of the panel are 100cm*24cm i have made them curved and D/S is 1.5mm and the distance between the spacers are 12cm, mylar is 6um and stepup trafo have 1:80 ratio, stators are made of 1.5mm steel with 58% air and isolation is made with emalje paint and 2 component polyester resin.
Then i remove the biasvoltage wire the speaker keeps playing in about 20-25 min so i chould not have any leakage from the coating (varmason esd floor polish) and then i turn up the bias voltage from about 2kv to 6,5kv there is not much difference in sound level.
The diaphragm and it´s here i think the problem maybe is, are stretched so much that i right between 2 of the spacers need to put 50g of weight before the diaphragm touch the stator.
hope i have give enough information to your guys to maybe help me with my problem...
regards
uglykid
Sorry to hear about your ESL problem-- it can be very frustrating but it's probably something simple, since you're configuration sounds about right to give very high output.
I really doubt the problem could be too much tension on the diaphragm.
I found out recently that a panel can play at very low volume even with a faulty bias supply and/or non-conducting diaphragm, since there can be a small potential between the diaphragm and stators in any case.
I could be totally wrong and sound like an idiot here--but I'm going to guess that, since you said the volume doesn't change much when you more than doubled the bias voltage, it sounds like the bias supply could be faulty, or it could be the diaphragm coating is not conducting sufficiently.
I would check the bias supply first. If you don't have a HV probe to measure it's output, you should be able to touch the leads together and see a good spark--- unless of course the supply uses a hugely high-value output resistor, in which case it may be perfectly functional but the high value resistor would not pass enough current to give a good spark. (my 2.7 KV bias supply uses a 20 Meg output resistor but it still passes enough current to give a pretty good spark when I touch the leads together)
If your bias supply checks good, and you can disassemble the panel without destroying it, you could see if the coating is the culprit by re-coating the diaphragm with liquid dish detergent containing sodium laurel sulfate (most brands do, and it works very well indeed)-- and if you then get good volume, that was it.
Good luck with it!
I really doubt the problem could be too much tension on the diaphragm.
I found out recently that a panel can play at very low volume even with a faulty bias supply and/or non-conducting diaphragm, since there can be a small potential between the diaphragm and stators in any case.
I could be totally wrong and sound like an idiot here--but I'm going to guess that, since you said the volume doesn't change much when you more than doubled the bias voltage, it sounds like the bias supply could be faulty, or it could be the diaphragm coating is not conducting sufficiently.
I would check the bias supply first. If you don't have a HV probe to measure it's output, you should be able to touch the leads together and see a good spark--- unless of course the supply uses a hugely high-value output resistor, in which case it may be perfectly functional but the high value resistor would not pass enough current to give a good spark. (my 2.7 KV bias supply uses a 20 Meg output resistor but it still passes enough current to give a pretty good spark when I touch the leads together)
If your bias supply checks good, and you can disassemble the panel without destroying it, you could see if the coating is the culprit by re-coating the diaphragm with liquid dish detergent containing sodium laurel sulfate (most brands do, and it works very well indeed)-- and if you then get good volume, that was it.
Good luck with it!
Last edited:
Tension does not affect sensitivity above resonance much, only below it.
What is the resonance frequency of the diaphragm ? I guess for your size and support spacing(if there is sufficient tension) it should be somewhere between 100-200 hz.
Bright sound can be because of beaming. Have you measured the frequency response(on/off axis) ? Normally non-segmented larger ESL elements tend to have a rise in highs.
One possible reason for low output is too high resistance of stator coating.
There is a theory that resistor divider is formed between air and coating and significant part of voltage drop is across the coating.
What hole size is ? if holes are too large, sensitivity could be reduced.
What is the resonance frequency of the diaphragm ? I guess for your size and support spacing(if there is sufficient tension) it should be somewhere between 100-200 hz.
Bright sound can be because of beaming. Have you measured the frequency response(on/off axis) ? Normally non-segmented larger ESL elements tend to have a rise in highs.
One possible reason for low output is too high resistance of stator coating.
There is a theory that resistor divider is formed between air and coating and significant part of voltage drop is across the coating.
What hole size is ? if holes are too large, sensitivity could be reduced.
Hi,
As soon as You switch on the bias supply You should notice a deflection of the diaphragm. This deflection can more easily be seen if the the ESL-stators are not yet married. You should always test the diaphragm bias as long as You have access to the open film membrane, i.e. as long as the panel is not finished.
How did You constuct the diaphragm contacts? Try to evaluate if the membrane surface is fully charged and that all segments of the membrane are biased. The output voltage of the bias supply alone is not a reliable indicator.
The second possibility lies in crossed audio tranny connections. Keep in mind that You have 3 contacts of which the middle is common connected to the diaphragm via the bias supply (and/or gnd-referenced). The outer connections voltages must be out of phase referenced to the middle. If they are connected in phase the output will be close to 0 (and just due to winding differences/tolerances).
High membrane tension certainly would not hinder easy deflection and output, but raise the lower bandwaith limit. High tension is in fact positive in any case as it raises efficiency and dynamic range.
jauu
Calvin
Thats a rather clear indicator not for too high mechanical tension but for no charge on the diaphragm or crossed audio tranny connections.
As soon as You switch on the bias supply You should notice a deflection of the diaphragm. This deflection can more easily be seen if the the ESL-stators are not yet married. You should always test the diaphragm bias as long as You have access to the open film membrane, i.e. as long as the panel is not finished.
How did You constuct the diaphragm contacts? Try to evaluate if the membrane surface is fully charged and that all segments of the membrane are biased. The output voltage of the bias supply alone is not a reliable indicator.
The second possibility lies in crossed audio tranny connections. Keep in mind that You have 3 contacts of which the middle is common connected to the diaphragm via the bias supply (and/or gnd-referenced). The outer connections voltages must be out of phase referenced to the middle. If they are connected in phase the output will be close to 0 (and just due to winding differences/tolerances).
High membrane tension certainly would not hinder easy deflection and output, but raise the lower bandwaith limit. High tension is in fact positive in any case as it raises efficiency and dynamic range.
jauu
Calvin
Last edited:
If you are sure that you have done everything right, the most likely cause could be that your coating material isn't very good.
Wachara C.
Wachara C.
Hi,
had a look at the Statguard floor finish 220521.
It says it is a aqueous acrylic emulsion with a solids content of 18%. That sounds like EC-powder linked in a polymer matrix (EC: electro-conductive) similar to Zelec Millikens ECPs.
The specced surface resistance obviously varies with coating thickness, ranging up from 10e7Ohms. A specced reach of 49m²/L translates to a wet coating thickness of ~20µm which is similar to a thin coating of a typical color laquer. Depending on the procedure of coating the wet thickness may be too thin, though I regard this aspect rather improbable, since ECP-based coatings may be very thin (dried-coating-thickness of >0.5µm).
Anyway, the transition from a highly conductive copper strip to a low conductance material will give (lowlevel) corona that may eat away the coating over time. As such the design of the contacts influences the longtime stability of the panel. Talking about a new panel I regard this possibilty also rather improbable.
If the Statguard is a ECP-based antistatic it should perform well and stable (conductivity is not affected by humidity/water), its useablity rather a matter of how well it wets and sticks to the diaphragm material (which would be a matter of the resins used, the diaphragm material and its preparation)
jauu
Calvin
had a look at the Statguard floor finish 220521.
It says it is a aqueous acrylic emulsion with a solids content of 18%. That sounds like EC-powder linked in a polymer matrix (EC: electro-conductive) similar to Zelec Millikens ECPs.
The specced surface resistance obviously varies with coating thickness, ranging up from 10e7Ohms. A specced reach of 49m²/L translates to a wet coating thickness of ~20µm which is similar to a thin coating of a typical color laquer. Depending on the procedure of coating the wet thickness may be too thin, though I regard this aspect rather improbable, since ECP-based coatings may be very thin (dried-coating-thickness of >0.5µm).
Anyway, the transition from a highly conductive copper strip to a low conductance material will give (lowlevel) corona that may eat away the coating over time. As such the design of the contacts influences the longtime stability of the panel. Talking about a new panel I regard this possibilty also rather improbable.
If the Statguard is a ECP-based antistatic it should perform well and stable (conductivity is not affected by humidity/water), its useablity rather a matter of how well it wets and sticks to the diaphragm material (which would be a matter of the resins used, the diaphragm material and its preparation)
jauu
Calvin
If you turn the bias up from 2kv to 6.5Kv you should easily hear close to a 6db increase of output level.
If there is no change then I would suspect the supply.
Since you are getting at least a 90db output this is a substantial amount of volume and your coating appears to to working.
What happens after you leave the bias supply on for a while?
Your coating is specified at around 10^9 and this is a bit high ,but it should still work,and by your report it does,so I don't suspect the coating unless you didn't get a good coverage.
How well did it wet the mylar when you applied it?
Just a few questions,But you definitely should have heard a difference if you increase the bias!!
You might try doubling your transformation ratio also and see if it make a difference as it should.
But recheck your bias though,many times I have seen this as the cause of low output of a panel.
If your supply is putting out 6.5kv then it should be blowing you out of the room even much more so with a 1:160 transformation ratio.
I run my little panels at those same voltages and they are 1/4 the size of the ones you have and I get +99db at one meter with a 6.8Kv bias.
Only I run a 1:256 transformation ratio on them.
Jer 🙂
If there is no change then I would suspect the supply.
Since you are getting at least a 90db output this is a substantial amount of volume and your coating appears to to working.
What happens after you leave the bias supply on for a while?
Your coating is specified at around 10^9 and this is a bit high ,but it should still work,and by your report it does,so I don't suspect the coating unless you didn't get a good coverage.
How well did it wet the mylar when you applied it?
Just a few questions,But you definitely should have heard a difference if you increase the bias!!
You might try doubling your transformation ratio also and see if it make a difference as it should.
But recheck your bias though,many times I have seen this as the cause of low output of a panel.
If your supply is putting out 6.5kv then it should be blowing you out of the room even much more so with a 1:160 transformation ratio.
I run my little panels at those same voltages and they are 1/4 the size of the ones you have and I get +99db at one meter with a 6.8Kv bias.
Only I run a 1:256 transformation ratio on them.
Jer 🙂
Last edited:
Hi,
Jer, You´re really settle for 90dB@150W if You easily gain the same SPL with just 1W? 😕
Man, Your ears must be exceptionally sensitive
😀
jauu
Calvin
uglykid
geraldfryjr
Jer, You´re really settle for 90dB@150W if You easily gain the same SPL with just 1W? 😕
Man, Your ears must be exceptionally sensitive
😀 jauu
Calvin
No,Calvin,Actually I like it quite loud but 90db is for good nearfield listening and is about average really.
Anymore than than that ear fatigue starts to set in quickly for me as I have listen to very loud systems for many years.
After the fatigue starts to set in it has to go louder in order for me too feel the same effect,Actually at least 105db is more comfortable for me.
He,he,he,he !!
Your right Calvin he should be getting 90db with only a 5V to 10v peak signal input to the step up transformer,Like I said I run those same voltages and get +99db at 1 meter and 110db at the diagphram and that wasn't even pushin' it yet.
As I have mentioned many times before, Stator coating failure was the demise of one of those little panels before I got to see the ultimate performance of them agian to measure.
They did it before but they didn't that day.
:Cheers:
jer 🙂
Anymore than than that ear fatigue starts to set in quickly for me as I have listen to very loud systems for many years.
After the fatigue starts to set in it has to go louder in order for me too feel the same effect,Actually at least 105db is more comfortable for me.
He,he,he,he !!
Your right Calvin he should be getting 90db with only a 5V to 10v peak signal input to the step up transformer,Like I said I run those same voltages and get +99db at 1 meter and 110db at the diagphram and that wasn't even pushin' it yet.
As I have mentioned many times before, Stator coating failure was the demise of one of those little panels before I got to see the ultimate performance of them agian to measure.
They did it before but they didn't that day.
:Cheers:
jer 🙂
Last edited:
Hi.
Charlie.
that was actually your homepage that inspired me to made some els again, i did a couple for about 15 years ago but they failed completed because off bad stator isolation and bad diaphragm coating at that time. I have tryed to separate my panel but with no luck, so i have now made a little test panel and that panel act like the big one with very low sensitivity but now i hopefully can figure out what there was wrong with the big panel.
I have tried my biassupply and i can not get a good spark even with 6.5kv i get a tiny tiny little spark then the wires almost touching each other, thats bad so i made another biassupply with a "good" spark and connected it to the test panel but no change.
bazukas.
my hole size is 4mm (3.5mm after coating) and 58% open area, i have tried another stator isolation on the little testpanel with same result so i dont think it is the isolation on the stators thats the problem, actually then i have coated the stators i tested them by putting them together with kitchen alu foil between (sandwich) connected the stepup trafo and turned the amp (150w) fully up to its clipping limit with a 1khz sinustone and the foil vibrated so much that i have to add earprotection that was insane loud.
calvin.
my diaphragm contact is made by cutting a long thin piece (5mm)of kitchen alu foil and put very thin dobbelsided tape on one side and place it on one of the stators so it made contact to the coating then i press the stators together, the foil goes up in both side of the panel.
you say a high conduktive copper foil over time eat up the coating... is it a good idea and use graphite (spray) instead? i mean add coating and then with masking tape made a connector of graphite and then maybe add a layer of coating on top of the graphite (sandwich).
you say that i have to test the coating before i put the panel together... how to? i have a x100 probe to my scope and i have a couple of good dvm and a couple of different powersupplys, actually i have a lot of measurement equipment but not a megger, can i use some of this?
the statguard polish is very easy to apply and it cover the diapgragm very nicely it´s a little soapy to use and it takes about a hour or so to dry, it sticks very well to the diaphragm i can not rip it of with the tapetrick and a fingernail.
i have tried to apply a very thick layer on the testpanel but still no change in soundlevel.
I´am sure i not have wrong trafo connection, have worked 22 years on transformer factory with testing trafos 🙂 offcouse i could have made a mistake but the trafo connection should be ok.
chinsettawong.
i think i have done like many others so i am very confused what is wrong, but maybe the varmason actually is a bad coating i really dont know, so i will try to use the dish soap on the test panel when i get home tommorow to see if it make any difference.
geraldjer.
like i wrote in my first post, the speaker keeps playing for about 20-25 minutes after i have disconnected the biassupply so the coating and isolation should be ok.
i will also try to raise the stepup ratio, but again i am confused other have a working speaker with the same D/S and stepup ratio like i have...hmmm maybe i am more deaf than i thought 🙂
bentoronto.
my test amp clips with 150w but honest i havent mesurement the input voltage on stepup trafo... will do then home again.
i will be back soon.
uglykid
Charlie.
that was actually your homepage that inspired me to made some els again, i did a couple for about 15 years ago but they failed completed because off bad stator isolation and bad diaphragm coating at that time. I have tryed to separate my panel but with no luck, so i have now made a little test panel and that panel act like the big one with very low sensitivity but now i hopefully can figure out what there was wrong with the big panel.
I have tried my biassupply and i can not get a good spark even with 6.5kv i get a tiny tiny little spark then the wires almost touching each other, thats bad so i made another biassupply with a "good" spark and connected it to the test panel but no change.
bazukas.
my hole size is 4mm (3.5mm after coating) and 58% open area, i have tried another stator isolation on the little testpanel with same result so i dont think it is the isolation on the stators thats the problem, actually then i have coated the stators i tested them by putting them together with kitchen alu foil between (sandwich) connected the stepup trafo and turned the amp (150w) fully up to its clipping limit with a 1khz sinustone and the foil vibrated so much that i have to add earprotection that was insane loud.
calvin.
my diaphragm contact is made by cutting a long thin piece (5mm)of kitchen alu foil and put very thin dobbelsided tape on one side and place it on one of the stators so it made contact to the coating then i press the stators together, the foil goes up in both side of the panel.
you say a high conduktive copper foil over time eat up the coating... is it a good idea and use graphite (spray) instead? i mean add coating and then with masking tape made a connector of graphite and then maybe add a layer of coating on top of the graphite (sandwich).
you say that i have to test the coating before i put the panel together... how to? i have a x100 probe to my scope and i have a couple of good dvm and a couple of different powersupplys, actually i have a lot of measurement equipment but not a megger, can i use some of this?
the statguard polish is very easy to apply and it cover the diapgragm very nicely it´s a little soapy to use and it takes about a hour or so to dry, it sticks very well to the diaphragm i can not rip it of with the tapetrick and a fingernail.
i have tried to apply a very thick layer on the testpanel but still no change in soundlevel.
I´am sure i not have wrong trafo connection, have worked 22 years on transformer factory with testing trafos 🙂 offcouse i could have made a mistake but the trafo connection should be ok.
chinsettawong.
i think i have done like many others so i am very confused what is wrong, but maybe the varmason actually is a bad coating i really dont know, so i will try to use the dish soap on the test panel when i get home tommorow to see if it make any difference.
geraldjer.
like i wrote in my first post, the speaker keeps playing for about 20-25 minutes after i have disconnected the biassupply so the coating and isolation should be ok.
i will also try to raise the stepup ratio, but again i am confused other have a working speaker with the same D/S and stepup ratio like i have...hmmm maybe i am more deaf than i thought 🙂
bentoronto.
my test amp clips with 150w but honest i havent mesurement the input voltage on stepup trafo... will do then home again.
i will be back soon.
uglykid
Aha! That's a perilous way to assess output voltage (which is what we should be talking about, not power). Yet that is the basic starting point in your query.
Quite possibly your amp is able to pump only a small voltage into your transformers.
Try to 'scope the voltage - computer software oscilloscopes are pretty good.
Ben
As I had mentioned I believe that your coating seems to be functioning properly.
If you actually have 6.5kv being produced by your supply you should be able to draw an arc of at least 1/4" long even with a current limiting series resistor.
See this thread for reference,
http://www.diyaudio.com/forums/plan...tor-insulation-mylar-coating.html#post2767712
The only thing a current limiting resistor will due is limit the current and the thickness of the arc as there is still 6.5kv and still should jump the gap only not as often as it would without the resistor.
The test that is shown on that post was at about 5kv as shown on the meter.
I must mention the ruler that I was using was segmented in 1/5 inch increments as to avoid any confusions.
jer 🙂
If you actually have 6.5kv being produced by your supply you should be able to draw an arc of at least 1/4" long even with a current limiting series resistor.
See this thread for reference,
http://www.diyaudio.com/forums/plan...tor-insulation-mylar-coating.html#post2767712
The only thing a current limiting resistor will due is limit the current and the thickness of the arc as there is still 6.5kv and still should jump the gap only not as often as it would without the resistor.
The test that is shown on that post was at about 5kv as shown on the meter.
I must mention the ruler that I was using was segmented in 1/5 inch increments as to avoid any confusions.
jer 🙂
Last edited:
Hi,
if the bias supply puts out good 6.5kV there should be a clean visible sparc over a few mm (if there are no or small valued current limiting resistors).
If You hold the wire of the supply down onto the membrane coating (the second bias-supply connection to gnd and the respective stator) the membrane should immediately and clearly recognizable be drawn towards the stator. If not, then either the bias-supply voltage is pulled down (leakage) or there´s no contact at all to the coating, or the coating simply isn´t working.
Thin strips of Aluminum foil are perfectly ok, though pure Aluminum tends to build up a insulating oxide surface layer. I always used the coating to ´glue´ the strips to the diaphragm. Then I additionally ´painted´ the strips with another thick layer of the coating, thereby creating a large contact surface and a environmental sealing of the strips. A press-on contact as You´ve done will also work, but may fail over longtime periods, because of corrosion of the contact metal and/or lowlevel corona eating up the coating.
Anyway, if You got high SPLs when testing the panels with an alu-diaphragm, the stators as well as trafo and bias-supply and their connection seem ok.
Then the missing output with the Statguard coating can only be due to a missing contact to the coating, or Statguard is unsuitable as coating or, as last possibility, the tranny might have died by internal flashovers whilst testing.
jauu
Calvin
if the bias supply puts out good 6.5kV there should be a clean visible sparc over a few mm (if there are no or small valued current limiting resistors).
If You hold the wire of the supply down onto the membrane coating (the second bias-supply connection to gnd and the respective stator) the membrane should immediately and clearly recognizable be drawn towards the stator. If not, then either the bias-supply voltage is pulled down (leakage) or there´s no contact at all to the coating, or the coating simply isn´t working.
Thin strips of Aluminum foil are perfectly ok, though pure Aluminum tends to build up a insulating oxide surface layer. I always used the coating to ´glue´ the strips to the diaphragm. Then I additionally ´painted´ the strips with another thick layer of the coating, thereby creating a large contact surface and a environmental sealing of the strips. A press-on contact as You´ve done will also work, but may fail over longtime periods, because of corrosion of the contact metal and/or lowlevel corona eating up the coating.
Anyway, if You got high SPLs when testing the panels with an alu-diaphragm, the stators as well as trafo and bias-supply and their connection seem ok.
Then the missing output with the Statguard coating can only be due to a missing contact to the coating, or Statguard is unsuitable as coating or, as last possibility, the tranny might have died by internal flashovers whilst testing.
jauu
Calvin
Hi all.
Ok there at least one big fool in Danmark...it´s actually me...i have now tried every thing all your great guys have mention and nothing have work for me!!! so i have done some brainstorming to figure out that was wrong, then one night it hits me that i was trying to made.. a MID/HIGH els speaker..and not a full range, so i connected a crossover to my speakers in my living room ( a couple of old system audio 1070) and removed all frequences under 300hz with 24db/octave and magic happens🙂 i got very low level from them too and thats with the same amp i was using to test the els panels, the only difference was that the els play so much much better🙂
So my speakers actually work like they should and i will go on and finish them.
I will post pictures of the stretcing jig and the speaker panels in a couple of days.
Thank you all very much for solutions and answers and for trying to help me🙂🙂 the only thing there was wrong was me😱
Kind regards
uglykid
Ok there at least one big fool in Danmark...it´s actually me...i have now tried every thing all your great guys have mention and nothing have work for me!!! so i have done some brainstorming to figure out that was wrong, then one night it hits me that i was trying to made.. a MID/HIGH els speaker..and not a full range, so i connected a crossover to my speakers in my living room ( a couple of old system audio 1070) and removed all frequences under 300hz with 24db/octave and magic happens🙂 i got very low level from them too and thats with the same amp i was using to test the els panels, the only difference was that the els play so much much better🙂
So my speakers actually work like they should and i will go on and finish them.
I will post pictures of the stretcing jig and the speaker panels in a couple of days.
Thank you all very much for solutions and answers and for trying to help me🙂🙂 the only thing there was wrong was me😱
Kind regards
uglykid
I am glad you figured it out !!!
It is amazing how how much sound there is once you add the lower registers !
I used to think the same thing with my little panels, even with just some little 5.25" woofers they just exploded in sound !!!
Then they exploded !!! He,he,he,he
jer 😉
It is amazing how how much sound there is once you add the lower registers !
I used to think the same thing with my little panels, even with just some little 5.25" woofers they just exploded in sound !!!
Then they exploded !!! He,he,he,he
jer 😉
Last edited:
- Status
- Not open for further replies.