I built my first ever esl today! It even works too! I'm using uninsulated stators 8mm apart (very far I know)
The bias voltage is around 3kv (I THINK)
I have about 10Mohms between it and the mylar.
The ESL is completely silent when idle. No flapping, no clicking, hissing or buzzing or anything.
What sound it makes seems very very good indeed. It even makes bass and the treble is fantastic. Truely a full-range speaker at close range.
I'm driving it with two 6v toroidal mains transformers. They sound very good and sound like they have a wide bandwidth.
The only problem is that it is very very quiet and the amp is having a horrible time. I was expecting sensitivity to be a little on the low side but not quite THIS low.
I don't think there are any leaks. I turned off the bias supply and the speaker still made sound about an HOUR later.
Oddly enough, I discharged the diaphragm by touching its wire against the stators and the speaker still worked??
Can anyone shed any light as to why my speakers are so quiet? It would be great if they were louder since they sound so promising!
The bias voltage is around 3kv (I THINK)
I have about 10Mohms between it and the mylar.
The ESL is completely silent when idle. No flapping, no clicking, hissing or buzzing or anything.
What sound it makes seems very very good indeed. It even makes bass and the treble is fantastic. Truely a full-range speaker at close range.
I'm driving it with two 6v toroidal mains transformers. They sound very good and sound like they have a wide bandwidth.
The only problem is that it is very very quiet and the amp is having a horrible time. I was expecting sensitivity to be a little on the low side but not quite THIS low.
I don't think there are any leaks. I turned off the bias supply and the speaker still made sound about an HOUR later.
Oddly enough, I discharged the diaphragm by touching its wire against the stators and the speaker still worked??
Can anyone shed any light as to why my speakers are so quiet? It would be great if they were louder since they sound so promising!
Regarding the bias: the panels are huge capacitors. The will store a charge for a long time. If you connect the supply and disconnect it they will retain the charge and play with no difference. So I assume that yo got the diaphragms charged up and then they stayed that way.
Your panels look very nice now you need some frames. Regards.
Your panels look very nice now you need some frames. Regards.
Hi,
welcome too the community ;-)
Well, it seems that nearly each and every noob makes the same old failures, of not believing the old cracks
ESLs just work with closest proximity of the electrodes. Close means in this case very very close.
Mankind managed to fly to the moon....but they have been unable to fly to Pluto.....same is with ESLs... 1mm works, 8mm don´t.
Since the field strength sinks quadratically with every doubling of distance its no wonder, that Your panels are quiet.
Second the 3kV are a good starting point for the Bias for a 2mm distance. The voltage gradient 0.75kV/mm of Bias is just the half of what You should be able to apply without probs.
It seems a good sign, that Your speaker plays so long after the Bias is off. The longer it plays the less leakage there is. But to be honest. To really see the real effect, You have to crank up efficiency of Your panels seriously.
Btw. The reason, that they play even after ´discharging´ is simply that You haven´t dischraged them completely! The very high resistance of the diaphragm coating resists the complete dicharge for quite a long time.
jauu
Calvin
welcome too the community ;-)
Well, it seems that nearly each and every noob makes the same old failures, of not believing the old cracks
ESLs just work with closest proximity of the electrodes. Close means in this case very very close.
Mankind managed to fly to the moon....but they have been unable to fly to Pluto.....same is with ESLs... 1mm works, 8mm don´t.
Since the field strength sinks quadratically with every doubling of distance its no wonder, that Your panels are quiet.
Second the 3kV are a good starting point for the Bias for a 2mm distance. The voltage gradient 0.75kV/mm of Bias is just the half of what You should be able to apply without probs.
It seems a good sign, that Your speaker plays so long after the Bias is off. The longer it plays the less leakage there is. But to be honest. To really see the real effect, You have to crank up efficiency of Your panels seriously.
Btw. The reason, that they play even after ´discharging´ is simply that You haven´t dischraged them completely! The very high resistance of the diaphragm coating resists the complete dicharge for quite a long time.
jauu
Calvin
I suspect you have only drained the static charge from a small area of the diaphragm. Due to the surface resistance on the diaphragm you would probably need to discharge from more than one spot or wait for the static to migrate to where you shorted the diaphragm over a longer time frame.bigwill said:
2 mm spacing won't give the diaphragm much travel if you want FR with decent bass.
If I'm not mistaken when I replace the 3M foam tape on my Martin Logans I used 1.5 mm x 2 = 3 mm.
Acoustat used...
approx. 1/16 of an inch stator spacers. With the right amp even 0ne plus 0ne's can surprise you with the bass that they can reproduce.
I would get some one mm spacer material and stretch your diaphragms taught (about 2-3% elogation is good). what are the dimensions of your panels and what thickness film are you using?
approx. 1/16 of an inch stator spacers. With the right amp even 0ne plus 0ne's can surprise you with the bass that they can reproduce.
I would get some one mm spacer material and stretch your diaphragms taught (about 2-3% elogation is good). what are the dimensions of your panels and what thickness film are you using?
Hi,
before anyone could give a real serious answer, one should know more about the concept itself. Reading something of 18" drivers in a dipole cabinet and ESL.....well it rather sounds like a good horror story to me with raped victims and hopefully loads of electric blood ;-)
Very probably will You experience serious problems with that kind of combination. What You need as a partner for a couple of such huge basses is a large panel with high efficiency and high SPLmax. While 18" width is reasonable, You should rather think of using 60" in height (or stacking 2 of Your panels). Still though even such a big panel will only be good down to ~300Hz-250Hz. Of course could You dimension it such it plays lower, but than You must accept vastly rising distortion levels and the risk of the membrane blown into the stator. For frequencies above 200Hz the excursion is already so low that a D/s of 1mm (1/25") would be sufficient for high SPLs. Now add 30-50% to provide a safety margin against acoustic coupling. So now we´re somewhere between 1.0 and 1.5 mm. Since You want to use huge basses, lets choose rather safe 1.5mm d/s (3mm stator/stator). Applying the ´rule of thumb´ of 70 to 100*d/s for the proximity of the membrane spacers we get ~100mm of distance.
When You use horizontal spacers like ML or vertical ones like Final, the 100mm high (or wide) segments and a 6µ-membrane will give a Fs of >100Hz. Using thermal tensioning the Fs will be closer to 100Hz, using strong mechanical tension You can end up between 150 and 200Hz. I´d suggest strong mechanical tension of ~150Hz which You can reach by pulling the membrane to 2-3% elongation. Original Mylar and Hostaphan will hold this tension for years.
A Fs of 150Hz leads to a reasonable cutoff-frequency of >250Hz. You will have to kill the Fs-resonance with a Notch-Filter, because it will be very strong ~+10dB! A simple HighPassFilter alone is not enough.
A panel of 60"*30" and 1.5mm d/s will have a capacitance value of 1.5-2nF. This means a ohmic value varying between 4megs and 4kohms between 20Hz and 20kHz.
A d/s of 1.5mm can safely handle a peak(!) Voltage of ~2.5kV. An amp providing solid 100W@8ohms gives ~40V peak. The audio tranny will need than a transformer-ratio of 1:62.5
A pair of Power Toroids for 230V/6V (50 to 80VA is ok) will theoretically give a ratio of 1:76, but recalculated with their efficiency values (+90-95%) You end up with realistic 1:68.
Without stray-inductance effects and the secondary ohmic resistance the amplifier now sees an impedance value @20kHz of 4kOhms/(68²)< 1Ohm. Added to this the ohmic resistance of the tranny´s secondary this should give a safe value for Your amp. But the highs will be attanuated by this voltage divider. Since You have to attenuate the highs this effect isn´t by far as problematic as it sounds. Additionally if the tranny has a somewhat higher stray inductance this will beef up the highs by the electrical resonance.
Rather more problematic is, that toroids have typically very low losses which leads to phase angles that can reach more than 80° around 2kHz to 6kHz! If Your amp can safely handle this load, this is imo the most straightforward and best sounding solution. If Your amp can´t handle it, You better try using other types of trannies with higher losses.
The Bias voltage with a 1.5mm d/s can be between 2kV and 3kV. Of course higher is better. Above a value of 1kV the effiviency won´t raise much more, but the distortion values will reduce further and sonically it sounds more impulsive (with decreasing Bias the sound softens up, so You can tune the sonic character of Your panel with the Bias a bit).
So now we have an excellent panel with good efficiency and a remarkable dynamic range, crossed over by ~300Hz using a Notchfilter and a rather soft 6dB or 12dB circuit (acoustically the panel will show rather 24dB to 36dB!). If You use a 12dB filter with a higher Q-factor, You can elegantly equalize the losses due to phase cancellation with low parts number count.
Lets see how we can add a bass partner to this
Using a classical CB or BR or TL is easy frequencywise, but reports on bad integration of Panel and Sub are countless. There are imo two reasons for this. First a lot of combinations show strongly differing acoustical filter curves at the crossover point and secondly a strongly changing distribution character between panel (cylindrical dipole) and bass (globe).
If the problem of the combination lies within the filters, one should try completely different branches, e.g 24 dB LP and 12dBHP (+Notch)
If this isn´t enough than one should think of using dipole basses and preferably stacked ones. The problem with dipoles is their limited frequency range. E.g 6.5"-8" drivers in a small U-shaped cabinet can reach up to 300Hz and slightly above, but with 18"ers You are restricted to below 150Hz.
There are two solutions to the prob.
a) use just a tower of smaller basses ( Daytons 6.5" Aluminium-drivers come to mind) or
b) have a look at MLs site how they built their Statement. A tower of small basses reaching down to 60-80Hz and a CB-subwoofer with big drivers providing the deep bass.
Usually solution a) will be more than enough for most users, but b) is for very large rooms and it is imo THE State(ment) of Art in speaker building
jauu
Calvin
before anyone could give a real serious answer, one should know more about the concept itself. Reading something of 18" drivers in a dipole cabinet and ESL.....well it rather sounds like a good horror story to me with raped victims and hopefully loads of electric blood ;-)
Very probably will You experience serious problems with that kind of combination. What You need as a partner for a couple of such huge basses is a large panel with high efficiency and high SPLmax. While 18" width is reasonable, You should rather think of using 60" in height (or stacking 2 of Your panels). Still though even such a big panel will only be good down to ~300Hz-250Hz. Of course could You dimension it such it plays lower, but than You must accept vastly rising distortion levels and the risk of the membrane blown into the stator. For frequencies above 200Hz the excursion is already so low that a D/s of 1mm (1/25") would be sufficient for high SPLs. Now add 30-50% to provide a safety margin against acoustic coupling. So now we´re somewhere between 1.0 and 1.5 mm. Since You want to use huge basses, lets choose rather safe 1.5mm d/s (3mm stator/stator). Applying the ´rule of thumb´ of 70 to 100*d/s for the proximity of the membrane spacers we get ~100mm of distance.
When You use horizontal spacers like ML or vertical ones like Final, the 100mm high (or wide) segments and a 6µ-membrane will give a Fs of >100Hz. Using thermal tensioning the Fs will be closer to 100Hz, using strong mechanical tension You can end up between 150 and 200Hz. I´d suggest strong mechanical tension of ~150Hz which You can reach by pulling the membrane to 2-3% elongation. Original Mylar and Hostaphan will hold this tension for years.
A Fs of 150Hz leads to a reasonable cutoff-frequency of >250Hz. You will have to kill the Fs-resonance with a Notch-Filter, because it will be very strong ~+10dB! A simple HighPassFilter alone is not enough.
A panel of 60"*30" and 1.5mm d/s will have a capacitance value of 1.5-2nF. This means a ohmic value varying between 4megs and 4kohms between 20Hz and 20kHz.
A d/s of 1.5mm can safely handle a peak(!) Voltage of ~2.5kV. An amp providing solid 100W@8ohms gives ~40V peak. The audio tranny will need than a transformer-ratio of 1:62.5
A pair of Power Toroids for 230V/6V (50 to 80VA is ok) will theoretically give a ratio of 1:76, but recalculated with their efficiency values (+90-95%) You end up with realistic 1:68.
Without stray-inductance effects and the secondary ohmic resistance the amplifier now sees an impedance value @20kHz of 4kOhms/(68²)< 1Ohm. Added to this the ohmic resistance of the tranny´s secondary this should give a safe value for Your amp. But the highs will be attanuated by this voltage divider. Since You have to attenuate the highs this effect isn´t by far as problematic as it sounds. Additionally if the tranny has a somewhat higher stray inductance this will beef up the highs by the electrical resonance.
Rather more problematic is, that toroids have typically very low losses which leads to phase angles that can reach more than 80° around 2kHz to 6kHz! If Your amp can safely handle this load, this is imo the most straightforward and best sounding solution. If Your amp can´t handle it, You better try using other types of trannies with higher losses.
The Bias voltage with a 1.5mm d/s can be between 2kV and 3kV. Of course higher is better. Above a value of 1kV the effiviency won´t raise much more, but the distortion values will reduce further and sonically it sounds more impulsive (with decreasing Bias the sound softens up, so You can tune the sonic character of Your panel with the Bias a bit).
So now we have an excellent panel with good efficiency and a remarkable dynamic range, crossed over by ~300Hz using a Notchfilter and a rather soft 6dB or 12dB circuit (acoustically the panel will show rather 24dB to 36dB!). If You use a 12dB filter with a higher Q-factor, You can elegantly equalize the losses due to phase cancellation with low parts number count.
Lets see how we can add a bass partner to this
Using a classical CB or BR or TL is easy frequencywise, but reports on bad integration of Panel and Sub are countless. There are imo two reasons for this. First a lot of combinations show strongly differing acoustical filter curves at the crossover point and secondly a strongly changing distribution character between panel (cylindrical dipole) and bass (globe).
If the problem of the combination lies within the filters, one should try completely different branches, e.g 24 dB LP and 12dBHP (+Notch)
If this isn´t enough than one should think of using dipole basses and preferably stacked ones. The problem with dipoles is their limited frequency range. E.g 6.5"-8" drivers in a small U-shaped cabinet can reach up to 300Hz and slightly above, but with 18"ers You are restricted to below 150Hz.
There are two solutions to the prob.
a) use just a tower of smaller basses ( Daytons 6.5" Aluminium-drivers come to mind) or
b) have a look at MLs site how they built their Statement. A tower of small basses reaching down to 60-80Hz and a CB-subwoofer with big drivers providing the deep bass.
Usually solution a) will be more than enough for most users, but b) is for very large rooms and it is imo THE State(ment) of Art in speaker building
jauu
Calvin
Thanks for your comprehensive reply Calvin.
I built a panel with 2mm spacing since I only have 2mm foam... but with IDENTICAL RESULTS!
It is SLIGHTLY louder but still very quiet! I really don't know what's going on
Also worth noting that with both panels, if I directly touch the bias lead to the output of my bias supply (bypassing the 10Megs of resistance) I hear a pretty loud high pitch whine from the speaker. There is also an arc to the lead! Does this mean my ESL is leaking? Maybe my mylar coating is MUCH too conductive? I couldn't measure a resistance on it with my DMM so I guessed it would be slightly suitable!
I built a panel with 2mm spacing since I only have 2mm foam... but with IDENTICAL RESULTS!
It is SLIGHTLY louder but still very quiet! I really don't know what's going on
Also worth noting that with both panels, if I directly touch the bias lead to the output of my bias supply (bypassing the 10Megs of resistance) I hear a pretty loud high pitch whine from the speaker. There is also an arc to the lead! Does this mean my ESL is leaking? Maybe my mylar coating is MUCH too conductive? I couldn't measure a resistance on it with my DMM so I guessed it would be slightly suitable!
Did you apply some sort of slightly conductive coating to the diaphragm? What did you use? How are you making electrical contact with the diaphragm?
I have found that with aluminum stators you sometimes need to burn through the layer of oxide that naturally forms on the aluminum before your wires actually make good electrical contact with them. Buring through the oxide is usually accomplished by cranking the volume very high. You'll hear a little sizzle then all of a sudden the speaker will start playing loudly and will continue to work at any volume level afterwards.
Can you post a schematic of your wiring?
I_F
I have found that with aluminum stators you sometimes need to burn through the layer of oxide that naturally forms on the aluminum before your wires actually make good electrical contact with them. Buring through the oxide is usually accomplished by cranking the volume very high. You'll hear a little sizzle then all of a sudden the speaker will start playing loudly and will continue to work at any volume level afterwards.
Can you post a schematic of your wiring?
I_F
I_Forgot said:
I used graphite for the electrical coating. I could not measure its resistance because my dmm only measures up to around 50Mohms but I suspect that the resistance of my coating is a bit on the low side even though I could not measure it. I suspect that I put on way too much to be honest
Connection to the diaphragm is made by a path of copper tape all around the spacers.
Connection to the stators is achieved by........... taping a wire to them with copper tape
The resistance measured ok but maybe it doesn't work like that?
Hi Bigwill,
1) If your coating wouldn't charge at all, you would just hear heavy distortions. You mentioned it sounds clear, so it charges at least a bit.
2) if your ESL doesn't become louder instantaneously when connecting to HV, it charges not enough. First option too high resistance of coating, second not enough charge of the capacitors (at least 0,033µF) each of the HV-supply.
3) But my favorite is, that you have significant leakage. So much that it makes no difference if on HV or not. Your Stators are not insulated and i guess full of sharp edges surrounding the holes. You loose all the charge into the air, since field density on the sharp edges exceeds insulation capability of air easily !
Make a test: Run your ESL on high input and put a humid breath into the speaker. (Be carefull not to touch the stators with your mouth - its lethal !) Doing so it might confirm my guess. If it smells like ozone, you have heavy leakage.
Capaciti
1) If your coating wouldn't charge at all, you would just hear heavy distortions. You mentioned it sounds clear, so it charges at least a bit.
2) if your ESL doesn't become louder instantaneously when connecting to HV, it charges not enough. First option too high resistance of coating, second not enough charge of the capacitors (at least 0,033µF) each of the HV-supply.
3) But my favorite is, that you have significant leakage. So much that it makes no difference if on HV or not. Your Stators are not insulated and i guess full of sharp edges surrounding the holes. You loose all the charge into the air, since field density on the sharp edges exceeds insulation capability of air easily !
Make a test: Run your ESL on high input and put a humid breath into the speaker. (Be carefull not to touch the stators with your mouth - its lethal !) Doing so it might confirm my guess. If it smells like ozone, you have heavy leakage.
Capaciti
I think you're getting ahead of yourself.
Graphite will not give a coating that is more than 50 Meg Ohms, especially if you "used too much". Graphite will typically read in the 100k-2 Megohm range. How did you do the measurement? How did you apply the graphite? Does the diaphragm look gray (it should). Are you sure you connected the right side of the diaphragm to the bias supply? Did you connect the other side of the bias supply to the center tap of the audio transformer(s)?
ESLs will play at very low volume with no bias supplied because the diaphragm will pick up charges just from being handled. The film may also retain some charge from the film manufacturing process.
I_F
Graphite will not give a coating that is more than 50 Meg Ohms, especially if you "used too much". Graphite will typically read in the 100k-2 Megohm range. How did you do the measurement? How did you apply the graphite? Does the diaphragm look gray (it should). Are you sure you connected the right side of the diaphragm to the bias supply? Did you connect the other side of the bias supply to the center tap of the audio transformer(s)?
ESLs will play at very low volume with no bias supplied because the diaphragm will pick up charges just from being handled. The film may also retain some charge from the film manufacturing process.
I_F
Hi,
there are two measures that can give You an idea about how leaky Your system is, or roughly which value the resistance of Your coating is.
The first is to build this little circuit:
It works as a blinker. If the panel is ok it´ll blink just every few seconds (~10secs). If it doesn´t the flash rate is much higher. At start up the flashrate is always high but quickly reducing after a few seconds. After ~ one minute it should have reached a stable slow blinking rate.
Function is:
C1 is charged up by the charging current of the Panels membrane. The voltage measurable over C1´s pins rises. As soon the ignition voltage of the neon flashlight is reached the lamp fires, thereby discharging C1 and the procedure starts again. If the panel is leaky, the charging current is high and C1´s voltage rises quickly and so the flashrate is high. If the membrane is charged up and low in leakage the current is tiny and it takes much longer for C1s voltage to reach the ignition point. Hence the lamp fires at a low rate.
Another advantage of this neat circuit is that it works like a on/off switch, thereby isolating the diaphragm completely from the HV and associated noise while in off-mode. Cost will be less than 1$.
It can stay in the ESL as a ´power indicator´ and I find it extremely useful.
The second method is using Your DMM in a series connection with the diaphragm. You use the internal resistance of Your DMM, connected in series to the unknown diaphragm resistance, so it forms a voltage divider with which You can (roundabout) exstimate the coating resistance.
jauu
Calvin
there are two measures that can give You an idea about how leaky Your system is, or roughly which value the resistance of Your coating is.
The first is to build this little circuit:
An externally hosted image should be here but it was not working when we last tested it.
It works as a blinker. If the panel is ok it´ll blink just every few seconds (~10secs). If it doesn´t the flash rate is much higher. At start up the flashrate is always high but quickly reducing after a few seconds. After ~ one minute it should have reached a stable slow blinking rate.
Function is:
C1 is charged up by the charging current of the Panels membrane. The voltage measurable over C1´s pins rises. As soon the ignition voltage of the neon flashlight is reached the lamp fires, thereby discharging C1 and the procedure starts again. If the panel is leaky, the charging current is high and C1´s voltage rises quickly and so the flashrate is high. If the membrane is charged up and low in leakage the current is tiny and it takes much longer for C1s voltage to reach the ignition point. Hence the lamp fires at a low rate.
Another advantage of this neat circuit is that it works like a on/off switch, thereby isolating the diaphragm completely from the HV and associated noise while in off-mode. Cost will be less than 1$.
It can stay in the ESL as a ´power indicator´ and I find it extremely useful.
The second method is using Your DMM in a series connection with the diaphragm. You use the internal resistance of Your DMM, connected in series to the unknown diaphragm resistance, so it forms a voltage divider with which You can (roundabout) exstimate the coating resistance.
jauu
Calvin
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.