Hello Guys,
First post and my first step into the DIY ESL panel journey.
I am sorry if this has been asked before. I did a forum search already, but has not made much progress.
I have two questions:
1. Regarding spacer ration 80:1 to 100:1 is this both horizontal and vertical direction? So if I have a .080" D/S spacing and choose a 100:1 ratio then my spacers should be located every 8" both vertical and horizontal?
2. If the spacers practically isolate one area of the diaphragm from the neighboring ones, can I just build multiple small 8" square panels and stack them up into a big one?
Thanks for helping. At the mean time I will keep typing "multiple small panel same as one big one?" in the search 🙂
Alex
First post and my first step into the DIY ESL panel journey.
I am sorry if this has been asked before. I did a forum search already, but has not made much progress.
I have two questions:
1. Regarding spacer ration 80:1 to 100:1 is this both horizontal and vertical direction? So if I have a .080" D/S spacing and choose a 100:1 ratio then my spacers should be located every 8" both vertical and horizontal?
2. If the spacers practically isolate one area of the diaphragm from the neighboring ones, can I just build multiple small 8" square panels and stack them up into a big one?
Thanks for helping. At the mean time I will keep typing "multiple small panel same as one big one?" in the search 🙂
Alex
1. You need only apply the [recommended] 70-100 x d/s rule vertically or horizontally, but not both.
2. Yes, you could use multiple smaller panels, but since the periphery spacers need to be wider than center spacers to accommodate a periphery charge ring and/or bonding area for the diaphragm, a single larger panel would give best efficiency as its total periphery spacer area (and resulting stray capacitance) could be significantly less.
Last edited:
Thanks Jazzman, you are actually the link that started this whole endeavor.
I too was stunned by the performance of the ML ESLs, and once you go ESL you don't go back.
I am trying to combine small panels because it is more manageable for me to construct them. If I can make the wall of each panel thin enough is it almost identical to a big continuous one (given that I wire them properly)?
Thanks guys,
Alex
I too was stunned by the performance of the ML ESLs, and once you go ESL you don't go back.
I am trying to combine small panels because it is more manageable for me to construct them. If I can make the wall of each panel thin enough is it almost identical to a big continuous one (given that I wire them properly)?
Thanks guys,
Alex
Hi there, Welcome to the world of DIY ESL's!!!
The D/S is typically referenced to the Width of the diaphragm and not its Length.
This would be equal to the smallest panel Dimension and to the Stator to diaphragm spacing as a ratio.
Generally you can use a lesser value at anytime.
If you use a value of over 1:100 (IE 1:150 or 1:200)you can start to run in to stability issues with a small spacing and wide panel width.
Alot of this also depends on how much tension you are using as well.
It is just a general guideline.
Try to stay away from Square shapes if you can, as a square shape will enhance the resonate frequency of the diaphragm and mostly its harmonics upward, Greatly.
It is better to deal with one fundamental frequency than having to deal with multiples of harmonics as well.
This is why you mainly only see rectangular or trapezoidal or other irregular shapes when it comes to large Diaphragm Planar drivers.
However you can build an ESL with any shape and dimensions you choose and it will work and still most likely sound great, as it is just part of their nature!!
The width of the panel determines its horizontal dispersion quality's.
The wider they get, The more the Highest frequency's will form in to a narrow beam, and it gets even narrower as the width increases or as the frequency gets higher.
There are many discussions on this subject, Just look up Electrically Segmented ESL's or horizontal dispersion for more on that subject.
As the frequency starts out low it will have a circular pattern of dispersion.
As the frequency is raised it will transition in to a Figure 8 pattern by the time frequency's wavelength equals the width of the panel.
It will then start to focus more and more in to a narrow beam as the frequency is raised above that point.
I have shown some simulations about this here,
A Segmented Stator Desktop ESL
When it comes to how much SPL it can produce at the lower frequency's then Displacement is the determining factor as well as the front to back cancellations that are also working against you.
This is the action of Dipole cancellation.
Here also you can find some simulations I have shown of this prediction using a spreadsheet program,
A Segmented Stator Desktop ESL
And this ESL simulator is great as well,
Electrostatic Loudspeaker (ESL) Simulator
It is the lowest frequency that you want to design for a maximum SPL has the have a natural rise in response of 6db per octave or 10db per decade in the farfield, again a results caused by the effect of dipole cancellation.
So they need to be compensated for that.
The SPL at the lowest frequency will be the reference as to however loud the system as a whole will be once they are EQ'ed to be flat.
Again there is another spreadsheet in these threads that can give you an idea of how it works.
I would stick with a Hybrid design for a first build as you will have the most success right off of the bat than trying to go fullrange using a woofer to cover everything below starting about 300Hz to 1000Hz.
Many are still trying to take the Full Range route as it is a difficult one to realize good performance, although quite possible if one understands all of the aspects of their operation.
Not to mention the high cost of an interface that can preform well down to 40Hz or 20Hz range.
At that a electrically segmented large panel still should be considered for such a design.
I personally have been building smaller panels as they can be made to perform very good as well as they are cheaper to build, easier to handle, and, won't break the bank in materials should one have to start over.
It is just a suggestion until you get the hang of the process of building them and what is involved.
By the same token you can get super result in a longer one just the same as long as you follow the same points of our guidelines.
I have seen many First time builders have great success here in the last few years.
Not having a thick enough stator coating is the most common issue of the First time builder.
Other than that, I can't recall anyone whom has failed!!! 😉
Welcome and Cheers to you !!!!
jer 🙂
The D/S is typically referenced to the Width of the diaphragm and not its Length.
This would be equal to the smallest panel Dimension and to the Stator to diaphragm spacing as a ratio.
Generally you can use a lesser value at anytime.
If you use a value of over 1:100 (IE 1:150 or 1:200)you can start to run in to stability issues with a small spacing and wide panel width.
Alot of this also depends on how much tension you are using as well.
It is just a general guideline.
Try to stay away from Square shapes if you can, as a square shape will enhance the resonate frequency of the diaphragm and mostly its harmonics upward, Greatly.
It is better to deal with one fundamental frequency than having to deal with multiples of harmonics as well.
This is why you mainly only see rectangular or trapezoidal or other irregular shapes when it comes to large Diaphragm Planar drivers.
However you can build an ESL with any shape and dimensions you choose and it will work and still most likely sound great, as it is just part of their nature!!
The width of the panel determines its horizontal dispersion quality's.
The wider they get, The more the Highest frequency's will form in to a narrow beam, and it gets even narrower as the width increases or as the frequency gets higher.
There are many discussions on this subject, Just look up Electrically Segmented ESL's or horizontal dispersion for more on that subject.
As the frequency starts out low it will have a circular pattern of dispersion.
As the frequency is raised it will transition in to a Figure 8 pattern by the time frequency's wavelength equals the width of the panel.
It will then start to focus more and more in to a narrow beam as the frequency is raised above that point.
I have shown some simulations about this here,
A Segmented Stator Desktop ESL
When it comes to how much SPL it can produce at the lower frequency's then Displacement is the determining factor as well as the front to back cancellations that are also working against you.
This is the action of Dipole cancellation.
Here also you can find some simulations I have shown of this prediction using a spreadsheet program,
A Segmented Stator Desktop ESL
And this ESL simulator is great as well,
Electrostatic Loudspeaker (ESL) Simulator
It is the lowest frequency that you want to design for a maximum SPL has the have a natural rise in response of 6db per octave or 10db per decade in the farfield, again a results caused by the effect of dipole cancellation.
So they need to be compensated for that.
The SPL at the lowest frequency will be the reference as to however loud the system as a whole will be once they are EQ'ed to be flat.
Again there is another spreadsheet in these threads that can give you an idea of how it works.
I would stick with a Hybrid design for a first build as you will have the most success right off of the bat than trying to go fullrange using a woofer to cover everything below starting about 300Hz to 1000Hz.
Many are still trying to take the Full Range route as it is a difficult one to realize good performance, although quite possible if one understands all of the aspects of their operation.
Not to mention the high cost of an interface that can preform well down to 40Hz or 20Hz range.
At that a electrically segmented large panel still should be considered for such a design.
I personally have been building smaller panels as they can be made to perform very good as well as they are cheaper to build, easier to handle, and, won't break the bank in materials should one have to start over.
It is just a suggestion until you get the hang of the process of building them and what is involved.
By the same token you can get super result in a longer one just the same as long as you follow the same points of our guidelines.
I have seen many First time builders have great success here in the last few years.
Not having a thick enough stator coating is the most common issue of the First time builder.
Other than that, I can't recall anyone whom has failed!!! 😉
Welcome and Cheers to you !!!!
jer 🙂
Wow thanks Jer
That's a lot to digest and I will take my time studying it. Just want to say thank you first.
I have access to hard anodize so I probably going to hard anodize my stators.
Regarding linking multiple panels together I figure it's just one big conducting plane so a well soldered wire should link all the panels together right? or am I overlooking something.
Once I get all the concept cleared up i will post some drawings.
Thanks again.
That's a lot to digest and I will take my time studying it. Just want to say thank you first.
I have access to hard anodize so I probably going to hard anodize my stators.
Regarding linking multiple panels together I figure it's just one big conducting plane so a well soldered wire should link all the panels together right? or am I overlooking something.
Once I get all the concept cleared up i will post some drawings.
Thanks again.
Anodizing is good but you will still need more coating on them as it is only good up to a certain amount of voltage.
I have been wanting experiment with anodizing as it is easy to do but last time I had gone out to the garage to try it, I had found that a critter had dumped over what sulfuric acid (Battery acid) I had left!!! 🙁
Yes, you can tie multiple panels together by just paralleling them.
Also you can run multiple panels or even the left and right channels from the same bias supply as well.
But you need to feed each diaphragm through its own feed resistor as well of about 10 meg to 30meg is good.
I explained all of that again and why in a very recent thread.
I will post the link for you if I can run across it again.
jer 🙂
P.S I found the original description of why you need to use resistors here,
http://www.diyaudio.com/forums/plan...tor-insulation-mylar-coating.html#post2791419
http://www.diyaudio.com/forums/plan...easure-bias-voltage-acoustat.html#post3800214
This also true to cut down the crosstalk between the two systems (L & R)if driven from the same bias supply as well.
There is another post on that too. 😉
I have been wanting experiment with anodizing as it is easy to do but last time I had gone out to the garage to try it, I had found that a critter had dumped over what sulfuric acid (Battery acid) I had left!!! 🙁
Yes, you can tie multiple panels together by just paralleling them.
Also you can run multiple panels or even the left and right channels from the same bias supply as well.
But you need to feed each diaphragm through its own feed resistor as well of about 10 meg to 30meg is good.
I explained all of that again and why in a very recent thread.
I will post the link for you if I can run across it again.
jer 🙂
P.S I found the original description of why you need to use resistors here,
http://www.diyaudio.com/forums/plan...tor-insulation-mylar-coating.html#post2791419
http://www.diyaudio.com/forums/plan...easure-bias-voltage-acoustat.html#post3800214
This also true to cut down the crosstalk between the two systems (L & R)if driven from the same bias supply as well.
There is another post on that too. 😉
Last edited:
- Status
- Not open for further replies.