I’ve had plans to try to build a pair of ESL speakers for some years now.
Finaly I have got all the material I need and have done some experiments with different coatings, spacings etc.
The plan for the speakers are as follows:
Use two ESL-panels/speaker and 2 (or 4) 12” woofers for the low end.
I have built the ”mid”-panels from 1000x400mm alu.sheets, 2mm spacing and 12µm Hostaphan film- polymercoated.
The ”tweeter”-panels are 1000x200mm, 4.8µm Hostaphan – but not shure what spacing to use and if i should divide the panel into 2 or 4 cells?
Using two toroids(80VA) as audiotransformer (ratio 1:80) and have abt. 4kV on the stators.
If done right, what SPL can one expect from a panel of this size and how low should it be able to play?
/R
Finaly I have got all the material I need and have done some experiments with different coatings, spacings etc.
The plan for the speakers are as follows:
Use two ESL-panels/speaker and 2 (or 4) 12” woofers for the low end.
I have built the ”mid”-panels from 1000x400mm alu.sheets, 2mm spacing and 12µm Hostaphan film- polymercoated.
The ”tweeter”-panels are 1000x200mm, 4.8µm Hostaphan – but not shure what spacing to use and if i should divide the panel into 2 or 4 cells?
Using two toroids(80VA) as audiotransformer (ratio 1:80) and have abt. 4kV on the stators.
If done right, what SPL can one expect from a panel of this size and how low should it be able to play?
/R
It's hard to predict actual SPL levels with any accuracy. Theoretical SPL levels can be calculated (see for example my simulator at quadesl.nl).
How low it can play depends on the size and tension of the diaphragm. It's a balancing act to get good SPL, sensitivity and low frequencies. Not much options but to go trial and error on this one.
For the tweeter panel 1mm spacing (stator to stator) would be enough. IMO 200mm is too wide to get good horizontal dispersion at high frequencies. You'd rather have something 1/10 of that.
I wonder what kind of polymer coating you are using. Can you tell us a little more about that?
How low it can play depends on the size and tension of the diaphragm. It's a balancing act to get good SPL, sensitivity and low frequencies. Not much options but to go trial and error on this one.
For the tweeter panel 1mm spacing (stator to stator) would be enough. IMO 200mm is too wide to get good horizontal dispersion at high frequencies. You'd rather have something 1/10 of that.
I wonder what kind of polymer coating you are using. Can you tell us a little more about that?
I taught I had abt. 4kV on the stators, but practical tests point to a much lower bias!
Have a HV-probe on its way, so will measure it shortly.
I'm not looking for an absolute SPL, but rather in what region DIY panels tend to land!?
The coating used on these panels are from MT Audio.
I have experimented with graphite, soap, fabric softner, glue etc.
Next test will probably be Elvamide doped.
The "tweeter panel" was planed to be divided into 2 or 4 sections verticaly.
/R
Have a HV-probe on its way, so will measure it shortly.
I'm not looking for an absolute SPL, but rather in what region DIY panels tend to land!?
The coating used on these panels are from MT Audio.
I have experimented with graphite, soap, fabric softner, glue etc.
Next test will probably be Elvamide doped.
The "tweeter panel" was planed to be divided into 2 or 4 sections verticaly.
/R
Do you have 4KV of audio signal on the stators, or do you have 4KV of bias voltage on the diaphragm?
If you hook up two toroids of 1:80 in series (is that what you are doing?) then you can get about 3KV of audio drive (depends on the amplifier of course, I assumed 50W in 8 Ohms)
So how high is the bias voltage?
You can devide the tweeter panel into multiple cells, but if you are going to drive them with the same signal, then it won't help you to improve dispersion. I think a strip of 3cm wide along the side of the panel in the picture would work nicely as a tweeter.
If you hook up two toroids of 1:80 in series (is that what you are doing?) then you can get about 3KV of audio drive (depends on the amplifier of course, I assumed 50W in 8 Ohms)
So how high is the bias voltage?
You can devide the tweeter panel into multiple cells, but if you are going to drive them with the same signal, then it won't help you to improve dispersion. I think a strip of 3cm wide along the side of the panel in the picture would work nicely as a tweeter.
Hi,
if done correctly, You could expect SPLs of >100dB@4m with the smaller panel in a freq-range >300Hz.
You should get lower SPLs with the larger panel in the same frequency range, because of the larger d/s. If done properly the frequency range won´t extend to lower frequencies anyway! The thicker membrane will in fact lead to a higher fo because of its superior tensile strength and therefore much higher posible mechanical tension! So the larger d/s is the wrong way to go, since You can´t use it for good! This is a very typical failour nearly every beginner makes. Smallest possible d/s needed for the lowest sensible fo is the way to success
jauu
Calvin
if done correctly, You could expect SPLs of >100dB@4m with the smaller panel in a freq-range >300Hz.
You should get lower SPLs with the larger panel in the same frequency range, because of the larger d/s. If done properly the frequency range won´t extend to lower frequencies anyway! The thicker membrane will in fact lead to a higher fo because of its superior tensile strength and therefore much higher posible mechanical tension! So the larger d/s is the wrong way to go, since You can´t use it for good! This is a very typical failour nearly every beginner makes. Smallest possible d/s needed for the lowest sensible fo is the way to success
jauu
Calvin
The voltage was meant as 4kV on the diaphragm!
In reality i suspect it only to be ~2kV, but I'll find out when I get my HV-probe.
The choice of spacing was due to avalible double sided tape with good glue.
I now also have 1,5mm PVC and 2mm old Bakelite strips avalible.
I'm tempted to try the Bakelite as it would make a real "retro" ESL - I hope the smell after cutting the strips fade with time!
Is there a good formula for the tension off the diaphragm or is it a trade secret and eaquals trial and error for me?
The info I have found point to everything from 0.8kg/5cm to "tight as
" ?
On my present panel i tensioned it to ~3.5kg/10cm.
With 2mm spacing I can get it to touch the stators on some bass notes at medium SPL levels, but with i HP-filter I can crank my GainClone almost to max (~50W).
Right now I'm using two 1:40 toroids - so the 1:80 is for a pair!
/R
In reality i suspect it only to be ~2kV, but I'll find out when I get my HV-probe.
The choice of spacing was due to avalible double sided tape with good glue.
I now also have 1,5mm PVC and 2mm old Bakelite strips avalible.
I'm tempted to try the Bakelite as it would make a real "retro" ESL - I hope the smell after cutting the strips fade with time!
Is there a good formula for the tension off the diaphragm or is it a trade secret and eaquals trial and error for me?
The info I have found point to everything from 0.8kg/5cm to "tight as
" ?On my present panel i tensioned it to ~3.5kg/10cm.
With 2mm spacing I can get it to touch the stators on some bass notes at medium SPL levels, but with i HP-filter I can crank my GainClone almost to max (~50W).
Right now I'm using two 1:40 toroids - so the 1:80 is for a pair!
/R
There is no 'best' tension, it depends on what you want. There are two opposing factors involved:
1. the higher the tension, the higher you can make your bias voltage (membrane stability), the higher the sensitivity of the panel.
2. the higher the tension, the higher the resonance frequency, so you lose low frequency output.
Stator spacing also effects sensitivity a lot.
So you have to balance these things to get what you want from your speaker.
1. the higher the tension, the higher you can make your bias voltage (membrane stability), the higher the sensitivity of the panel.
2. the higher the tension, the higher the resonance frequency, so you lose low frequency output.
Stator spacing also effects sensitivity a lot.
So you have to balance these things to get what you want from your speaker.
Ok - have just finished painting a couple of new test stators!
They are 350 x 1000 x 1.5 mm
This time I'm going to try my 4.8µm film and 1.5mm spacing.
Is there any gain in splitting the area into a "tweeter cell" of abt. 20-30mm along the side and the rest off the area into cells off abt. 125x300mm ?
/R
They are 350 x 1000 x 1.5 mm
This time I'm going to try my 4.8µm film and 1.5mm spacing.
Is there any gain in splitting the area into a "tweeter cell" of abt. 20-30mm along the side and the rest off the area into cells off abt. 125x300mm ?
/R
Yes a tweeter cell will give you benefit, especially if the stator is electrically separated from the rest. Then you can start playing with attenuation resistors in series with the stators (of the big section).
It's hard to tell from the picture, but how open is your stator material? It looks like it may be a little bit too open.
It's hard to tell from the picture, but how open is your stator material? It looks like it may be a little bit too open.
Hi,
I disagree with arend-jan about best tension values. To my taste and experience the best value is always the highest possible, just before the membrane material starts to flow.
For the few usable membrane materials this means a elongation of typically 1-2%. Above this value the material starts to flow. The resonance freq won´t rise any more.
Since ESL panels are normaly used in open baffle montage, there is no sense in trying to reach a very low crossover-frequency, hence a very low fo. Instead its better to place the crossover-freq not lower than a point where a max. equing of ~6dB is needed. Even with a 40cm wide panel this point is reached around 200Hz. Stretching a 4.8µm film to 1-2% elongation will result in a fo of ~120-150Hz (with properly distanced spacers). A d/s is 1mm should already be ok, but with a certain safety headroom to cope with uneven stators and acoustic coupling 1.5mm is fine.
jauu
Calvin
I disagree with arend-jan about best tension values. To my taste and experience the best value is always the highest possible, just before the membrane material starts to flow.
For the few usable membrane materials this means a elongation of typically 1-2%. Above this value the material starts to flow. The resonance freq won´t rise any more.
Since ESL panels are normaly used in open baffle montage, there is no sense in trying to reach a very low crossover-frequency, hence a very low fo. Instead its better to place the crossover-freq not lower than a point where a max. equing of ~6dB is needed. Even with a 40cm wide panel this point is reached around 200Hz. Stretching a 4.8µm film to 1-2% elongation will result in a fo of ~120-150Hz (with properly distanced spacers). A d/s is 1mm should already be ok, but with a certain safety headroom to cope with uneven stators and acoustic coupling 1.5mm is fine.
jauu
Calvin
New panel now playing!
Painted 1.5mm alusheet 3mm holes, spacers 1.5mm tape, 4.8µm Hostaphan, MTAudio coating.
Sounds realy good - have not measured frequency response yet, but they seem to have better highs than previous.
Divided the area into 6 cells as my first try ended with the tape slipping when i atached the film!
(stuck to the film very good, but not to the paint om the stators)
First tensioned to 2.5kg/10cm but reduced to 2kg/10cm and the tape now sticks to stators good.
I also have the panel "under preasure" from the wooden frame slots.
Time will tell if it holds, and if it does, I'm going to have some of this aluminum anodized.
/R
Painted 1.5mm alusheet 3mm holes, spacers 1.5mm tape, 4.8µm Hostaphan, MTAudio coating.
Sounds realy good - have not measured frequency response yet, but they seem to have better highs than previous.
Divided the area into 6 cells as my first try ended with the tape slipping when i atached the film!
(stuck to the film very good, but not to the paint om the stators)
First tensioned to 2.5kg/10cm but reduced to 2kg/10cm and the tape now sticks to stators good.
I also have the panel "under preasure" from the wooden frame slots.
Time will tell if it holds, and if it does, I'm going to have some of this aluminum anodized.
/R
I anodize for mainly two reasons:
1. Durability - scratch resistant, no problems with chemicals if I want to desolve glue etc.
2. Esthetics - looks good as color is even and it gets in everywhere.
The oxide is very thin, so I do'nt know if the insulation it provides is of any use with these high voltages?
The panel with the painted stators sound realy nice, so I think I'm going to bild a more permanent frame for them.
Have a 10" woofer on the shelf in the garage, that perhaps could be a mate for a complete speaker!?
/R
1. Durability - scratch resistant, no problems with chemicals if I want to desolve glue etc.
2. Esthetics - looks good as color is even and it gets in everywhere.
The oxide is very thin, so I do'nt know if the insulation it provides is of any use with these high voltages?
The panel with the painted stators sound realy nice, so I think I'm going to bild a more permanent frame for them.
Have a 10" woofer on the shelf in the garage, that perhaps could be a mate for a complete speaker!?
/R
Hi,
the thickness of the oxide layer is typically ~25µm, which is by far too less to have any remarkable insulative effect. Don´t ever touch Your panels when playing music! Could be the last thing You did!
Expect to need a insulator thickness of ~0.5mm for the high voltage levels of Your design if You don´t have the the metal sheets thoroughly deburred and ~0.3mm with well deburred sheets (rounded hole edges). You could use a standard clear PU-laquer and spray the panel at least 4 times each side. This is a very durable and scratch resistent material used as coating for staircases or boats. It´s electrical values are very suitable as ESL coating.
Since Your panels don´t look optimally designed yet, You might do this with the next panels and think about curving the panels
see the nicely rounded hole edges, which btw. increase the bandwidth of the system.
jauu
Calvin
the thickness of the oxide layer is typically ~25µm, which is by far too less to have any remarkable insulative effect. Don´t ever touch Your panels when playing music! Could be the last thing You did!
Expect to need a insulator thickness of ~0.5mm for the high voltage levels of Your design if You don´t have the the metal sheets thoroughly deburred and ~0.3mm with well deburred sheets (rounded hole edges). You could use a standard clear PU-laquer and spray the panel at least 4 times each side. This is a very durable and scratch resistent material used as coating for staircases or boats. It´s electrical values are very suitable as ESL coating.
Since Your panels don´t look optimally designed yet, You might do this with the next panels and think about curving the panels
An externally hosted image should be here but it was not working when we last tested it.
see the nicely rounded hole edges, which btw. increase the bandwidth of the system.
jauu
Calvin
Hrmpf......I'm deeply offended!
The eternal "problem" with DIY - You always come up with a better version along the way => never ending story!
How did you round the edges Calvin - etching?
Curved stators is a project for the future, have enough factors to sort out with my flat ones!
I have measured the current test panels and have a relativly flat frequency response from 100Hz to ~7kHz where they start to fall off.
Project for the wekk will be to paint a couple of tweeter stators in the same material.
/R
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.