Rather than continuing to hijack James' thread, I decided to start a new one to pick all of your brains on how to configure a segmented wire-stator ESL... starting from post #148 on James' thread.
As I've said, I'm a total dummy with electronics so please indulge my ignorance.
My initial thoughts for a wire stator to fit the 10.5" x 46.5" opening in my existing speaker frame would be:
- Black egg crate plastic louvers (florescent light diffusers) to support the wires
- Wires would be .035 copper coated TIG rods (cause I just happen to have some)... they are only 36" long so I would have to butt
and solder them to get the required length.
- I would use 13-thread/inch all thread rod to space the wires across the 10.5" open width--- omitting the two wires closest to the
spacers would leave 134 vertical wires in the grid.
- After gluing the wires in place and soldering the leads, the whole assembly would be spray coated with gloss clear polyurethane.
Any thoughts on coating thickness? (bias voltage will be only 2.7KV)
- 1/16" D/S, with seven equally spaced horizontal supports, 3/8" wide. If I opt for unequally spaced supports to distribute the
diaphragm resonance, that would require adding at least one more support spacer to meet the 100 x d/s rule.
- The diaphragm would be mechanically tensioned, primarily in the vertical direction like ML's, with just enough lateral tension to pull out
the wrinkles. Even though it's a flat panel, I figure vertical tensioning would be better in this case... any thoughts on that?
- I would use my existing tandem Farnell 230V/6V toroidal transformers, 2.7KV bias supply, and digital crossover bi-amp setup.
So now, the questions are:
- Number of wire groups required (134 wires total)
- Respective widths of wire groups
- How to configure the resistor ladder.
- Resistor values.
I'm going to attempt to download the calculator and figure it out... I'm kinda lost at this point.
As I've said, I'm a total dummy with electronics so please indulge my ignorance.
My initial thoughts for a wire stator to fit the 10.5" x 46.5" opening in my existing speaker frame would be:
- Black egg crate plastic louvers (florescent light diffusers) to support the wires
- Wires would be .035 copper coated TIG rods (cause I just happen to have some)... they are only 36" long so I would have to butt
and solder them to get the required length.
- I would use 13-thread/inch all thread rod to space the wires across the 10.5" open width--- omitting the two wires closest to the
spacers would leave 134 vertical wires in the grid.
- After gluing the wires in place and soldering the leads, the whole assembly would be spray coated with gloss clear polyurethane.
Any thoughts on coating thickness? (bias voltage will be only 2.7KV)
- 1/16" D/S, with seven equally spaced horizontal supports, 3/8" wide. If I opt for unequally spaced supports to distribute the
diaphragm resonance, that would require adding at least one more support spacer to meet the 100 x d/s rule.
- The diaphragm would be mechanically tensioned, primarily in the vertical direction like ML's, with just enough lateral tension to pull out
the wrinkles. Even though it's a flat panel, I figure vertical tensioning would be better in this case... any thoughts on that?
- I would use my existing tandem Farnell 230V/6V toroidal transformers, 2.7KV bias supply, and digital crossover bi-amp setup.
So now, the questions are:
- Number of wire groups required (134 wires total)
- Respective widths of wire groups
- How to configure the resistor ladder.
- Resistor values.
I'm going to attempt to download the calculator and figure it out... I'm kinda lost at this point.
Last edited:
Hi,
Some comments :
a) Why use welding rods when PVC insulated wire is available and cheap? No need to mess with spray paint & characteristics proven, like breakdown strength and correct volumetric resistance of insulation.
b) "just enough lateral tension to pull out the wrinkles" is not a sufficient measure what the tension is. For mechanically stretched membrains stretching could be in the range of around 0.5-2%. It can be easy measured by marking some dots with a permanent pen and measuring distance before & after stretching. I suggest to make some experiments to land at correct resonance frequency. High stretching(like 2% or even more) is not recommended because stability will drift significantly over time.
c) The segmentation could be as simple as this(it's somewhat similar but smaller design). More segments are needed to make off-axis response more linear, but this is perhaps good enough. The input of wire groups mean that there are 10 wires in the center, then 14 & 18 on left & right(5 segments in total). ESL_Seg program is easy to play with and no math knowledge required.
Regards
Lukas
Some comments :
a) Why use welding rods when PVC insulated wire is available and cheap? No need to mess with spray paint & characteristics proven, like breakdown strength and correct volumetric resistance of insulation.
b) "just enough lateral tension to pull out the wrinkles" is not a sufficient measure what the tension is. For mechanically stretched membrains stretching could be in the range of around 0.5-2%. It can be easy measured by marking some dots with a permanent pen and measuring distance before & after stretching. I suggest to make some experiments to land at correct resonance frequency. High stretching(like 2% or even more) is not recommended because stability will drift significantly over time.
c) The segmentation could be as simple as this(it's somewhat similar but smaller design). More segments are needed to make off-axis response more linear, but this is perhaps good enough. The input of wire groups mean that there are 10 wires in the center, then 14 & 18 on left & right(5 segments in total). ESL_Seg program is easy to play with and no math knowledge required.
Regards
Lukas
Attachments
Thanks the the info Lukas,
Since I already have the TIG rods, I think I would want go with those. They are already straight and rigid which should make them easy to layout on the grid without having to tension them. And I think the clear coated bright copper wires on the gloss black egg crate would be give nice cosmetics.
Regarding tensioning the diaphragm, I was not intending to merely pull the wrinkles out. Rather, mechanically stretching the diaphragm to 1.5% elongation in the vertical direction only... and then applying just enough horizontal tension to pull out any wrinkles in the film.
I just tried to download the spreadsheet but I apparently cannot use it with my outdated, unregistered, bootleg copy of Excel. I'm hoping my GF has Excel on her laptop and will let me use it.
Last edited:
Yes, the diaphragm in any flat panel ESL could be heat shrunk, and many builders use that method. It's certainly easier than building a stretching jig; although I doubt you could achieve the same tension.
When I built the perf-metal flat panels I'm using now, which have veritcal spacers, I used a pneumatic bike tube jig to stretched the film to 1.5% elongation in all directions-- this method gives measurable and repeatable tension and you can make the tension whatever you want it to be.
For the wire stator panels I'm contemplating now, which would use horizontal support spacers, and would be driven by discrete wire groups, I figured it might be better to tension the diaphragm predominantly in the vertical direction so as to mitigate interaction between the discretely-driven areas of the diaphragm.
Hi Charlie
Ideally, the more and narrower the segments the better. Practically, I would make them all about 12 mm wide, perhaps as much as 15 mm. This will avoid the comb filter behaviour off axis that you can see in the plot.
Also, if all the segments are the same width you can use the same valued resistors between segments.
Rather than trying to join rods, just stack two panels one above the other - modular like. A single 36" panel will be OK for hybrid, and probably a bit sensitive to listener position (slightly different sitting down and standing up). Two 36" panels stacked on top of each other will be a good full range ESL.
If you can manage a little theory, I notice my paper (and other useful stuff) is posted at http://wireesl.weebly.com/uploads/2/4/9/7/24970733/segmented_esl_theory.pdf , It gives the design equations and some explanations.
best wishes
Rod
Ideally, the more and narrower the segments the better. Practically, I would make them all about 12 mm wide, perhaps as much as 15 mm. This will avoid the comb filter behaviour off axis that you can see in the plot.
Also, if all the segments are the same width you can use the same valued resistors between segments.
Rather than trying to join rods, just stack two panels one above the other - modular like. A single 36" panel will be OK for hybrid, and probably a bit sensitive to listener position (slightly different sitting down and standing up). Two 36" panels stacked on top of each other will be a good full range ESL.
If you can manage a little theory, I notice my paper (and other useful stuff) is posted at http://wireesl.weebly.com/uploads/2/4/9/7/24970733/segmented_esl_theory.pdf , It gives the design equations and some explanations.
best wishes
Rod
Thanks Rod,
I've got some boning up to do before starting this project and the paper you referenced looks like a great place to start.
I'm afraid two 36" panels stacked would not work for me, since the proposed wire-panels would be replacing the 12"x48" perf-metal panels in my current speakers (see link below)-- the perf-metal panels sound great but they do put your head in a vise!
Charlie
I've got some boning up to do before starting this project and the paper you referenced looks like a great place to start.
I'm afraid two 36" panels stacked would not work for me, since the proposed wire-panels would be replacing the 12"x48" perf-metal panels in my current speakers (see link below)-- the perf-metal panels sound great but they do put your head in a vise!
Charlie
Last edited:
I don't know, as I have no experience with wire stator ESL's. I opted for .035 because I happen to have some on hand. I would imagine the greater mass and stiffness of .060 wire would allow greater spacing between wire supports. However, wire on a roll might be difficult to straighten; whereas, TIG rods are already straight.
Hi Charlie,
Very long time lurker here! I'm in the final information gathering phase of my ESL build, and although I felt I had not gathered all my thoughts enough to ask cogent questions, this seems as good as time as any to jump in! But first I want to complement posters like you, SyBorg, Ger, and others for taking the time to actually photograph and share your triumphs, and even some mistakes with us all. I can't tell you how much seeing actual photos of finished working ESL's means along with the photos of the construction steps, much more than vague illustrations, graphs, schematics, or math proofs ever could! In this way, the last few years especially have been a renaissance on this board!
But your new thread here hits upon not only the construction method I had settled upon, the TIG type wire construction over light diffuser, (poster SyBorg on his website has an excellent example of this method along with a source for thicker acrylic egg crate), you also outline some characteristics that I am interested in...
-clear coating the TIG wire to preserve the cool copper look.
-the possibility of easy segmentation of the stator by method of the TIG wire method of construction.
I must admit I've never heard a segmented electrostatic, and I've heard the original very first ML Monolith that a friend owned before ML even had a dealer network, the ML CLS, the ML Statement E2, the Sander's Innersound, and at Roger Sander's house back when he lived in California, I actually went to his house in 1982 and heard his original wire ESL that's in the photos in Audio Amateur 1977 and Speaker Builder 1980. And yeah, he's just as nice and sharing as everyone has said! So, in the last few months as I've been reading here, all the talk here and there in passing about segmentation has me wondering what it would sound like, and especially why many commercial electrostats are not segmented. Also, many posters on threads here talk about segmentation with phrases like "I've settled on segmentation for my next build" or "could someone help me with resistor values for my segmented stator". Not really helpful at all to a long term audiophile that's interested in state of the art sound. What I'm looking for is "I've built both a segmented and a non-segmented stator, to my ears here are the differences, the pros and the cons". Now that would be helpful! And Charlie, I can see that in this thread I can get some of these questions finally answered...and some photos! You seem to be asking the same questions. To my ears the unsegmented ESL (in the best commercial examples anyway) has a focus, a striking resolution on the sound that's just incredible. Charlie, I believe it was your friend that called them "remote headphones". An apt description! What could have better resolution than giant electrostatic headphones? When I listen and am very involved in the music I'm not really interested in moving around, I call this my "audio seance mode". Would the segmented stator offer me advantages? And the disadvantages? I've never heard them discussed. Is there really a totally free lunch? I must admit to a strong bias, after being blown away from my experience listening to Roger Sander's own speakers, and reading his articles where he makes a case for only directional speakers over the years, I'm skeptical. Skeptical but open.
As for the clear coat, perhaps posters SyBorg (who's done quite a lot of experimenting with panel construction) and Ger can share if it's possible to get a reliable working ESL with good longevity at 2.7kv using only clear coats? One of the issues of living in California is that finishes that contain "hot" solvents are not really available, either not at all, or only in small quantities via rattle can. This is one of the reasons I settled on building the TIG wire stators, as SyBorg reports at the time of his writing they seem to be behaving using I believe California sourced materials.
And I understand Charlie, that you would like to build a panel that would fit in your existing opening of your (very striking) perforated metal ESL. That would be especially helpful for us readers too, as we can read about an apples to apples comparison as you share your listening experiences. Will the fully segmented ESL lose some of it's resolving power compared to your original?
I read that you are considering soldering the shorter 36" lengths of TIG together so you can build the right sized panel to fit your existing speakers. The soldering aspect makes me wonder though if this area of the wire might be more sensitive to arcing? Since I want to build a full panel height, I've been working through in my mind the same issue of the shorter TIG rods. Wouldn't it be better just to order the longer TIG on a roll, fight thorough the additional work of slightly coiled misbehaving wire with additional pieces of strip magnets and threaded rods to hold things down while gluing? SyBorg has some good photos of the magnet strips and threaded rods in action, both are super cheap even if you had to buy three times as many.
This is the roll TIG wire I settled on...
http://www.amazon.com/gp/product/B00KGDHFTA/ref=ox_sc_sfl_title_7?ie=UTF8&psc=1&smid=A2ZMBCWCT70IMM
I look forward to everyone's thoughts, and how this project comes along.
Take good care, Neil
Very long time lurker here! I'm in the final information gathering phase of my ESL build, and although I felt I had not gathered all my thoughts enough to ask cogent questions, this seems as good as time as any to jump in! But first I want to complement posters like you, SyBorg, Ger, and others for taking the time to actually photograph and share your triumphs, and even some mistakes with us all. I can't tell you how much seeing actual photos of finished working ESL's means along with the photos of the construction steps, much more than vague illustrations, graphs, schematics, or math proofs ever could! In this way, the last few years especially have been a renaissance on this board!
But your new thread here hits upon not only the construction method I had settled upon, the TIG type wire construction over light diffuser, (poster SyBorg on his website has an excellent example of this method along with a source for thicker acrylic egg crate), you also outline some characteristics that I am interested in...
-clear coating the TIG wire to preserve the cool copper look.
-the possibility of easy segmentation of the stator by method of the TIG wire method of construction.
I must admit I've never heard a segmented electrostatic, and I've heard the original very first ML Monolith that a friend owned before ML even had a dealer network, the ML CLS, the ML Statement E2, the Sander's Innersound, and at Roger Sander's house back when he lived in California, I actually went to his house in 1982 and heard his original wire ESL that's in the photos in Audio Amateur 1977 and Speaker Builder 1980. And yeah, he's just as nice and sharing as everyone has said! So, in the last few months as I've been reading here, all the talk here and there in passing about segmentation has me wondering what it would sound like, and especially why many commercial electrostats are not segmented. Also, many posters on threads here talk about segmentation with phrases like "I've settled on segmentation for my next build" or "could someone help me with resistor values for my segmented stator". Not really helpful at all to a long term audiophile that's interested in state of the art sound. What I'm looking for is "I've built both a segmented and a non-segmented stator, to my ears here are the differences, the pros and the cons". Now that would be helpful! And Charlie, I can see that in this thread I can get some of these questions finally answered...and some photos! You seem to be asking the same questions. To my ears the unsegmented ESL (in the best commercial examples anyway) has a focus, a striking resolution on the sound that's just incredible. Charlie, I believe it was your friend that called them "remote headphones". An apt description! What could have better resolution than giant electrostatic headphones? When I listen and am very involved in the music I'm not really interested in moving around, I call this my "audio seance mode". Would the segmented stator offer me advantages? And the disadvantages? I've never heard them discussed. Is there really a totally free lunch? I must admit to a strong bias, after being blown away from my experience listening to Roger Sander's own speakers, and reading his articles where he makes a case for only directional speakers over the years, I'm skeptical. Skeptical but open.
As for the clear coat, perhaps posters SyBorg (who's done quite a lot of experimenting with panel construction) and Ger can share if it's possible to get a reliable working ESL with good longevity at 2.7kv using only clear coats? One of the issues of living in California is that finishes that contain "hot" solvents are not really available, either not at all, or only in small quantities via rattle can. This is one of the reasons I settled on building the TIG wire stators, as SyBorg reports at the time of his writing they seem to be behaving using I believe California sourced materials.
And I understand Charlie, that you would like to build a panel that would fit in your existing opening of your (very striking) perforated metal ESL. That would be especially helpful for us readers too, as we can read about an apples to apples comparison as you share your listening experiences. Will the fully segmented ESL lose some of it's resolving power compared to your original?
I read that you are considering soldering the shorter 36" lengths of TIG together so you can build the right sized panel to fit your existing speakers. The soldering aspect makes me wonder though if this area of the wire might be more sensitive to arcing? Since I want to build a full panel height, I've been working through in my mind the same issue of the shorter TIG rods. Wouldn't it be better just to order the longer TIG on a roll, fight thorough the additional work of slightly coiled misbehaving wire with additional pieces of strip magnets and threaded rods to hold things down while gluing? SyBorg has some good photos of the magnet strips and threaded rods in action, both are super cheap even if you had to buy three times as many.
This is the roll TIG wire I settled on...
http://www.amazon.com/gp/product/B00KGDHFTA/ref=ox_sc_sfl_title_7?ie=UTF8&psc=1&smid=A2ZMBCWCT70IMM
I look forward to everyone's thoughts, and how this project comes along.
Take good care, Neil
Hi,
the differences of segmented to non-segmented panels, whether curved or flat, have been discribed in some threads before (probabely a few years old by now).
Therefore just in short:
- non-segmented metal sheet stators typically achieve the highest efficiency and highest dynamic range.
- they require high quality transformers with very low leakage inductance, but mostly lower transformation factors.
- The lower requirement in drive voltage and transformation factor is positive regarding ageing, insulation and safety.
- The requirement for high quality lowloss transformers, combined with the high capacitance towards higher frequencies presents a more difficult load for the driving amplifier.
- curving a panel widens the distribution character at the top end of the frequency range (sweet spot) and increases the panel's mechanical stability/stiffness.
Its highly recommended for this kind of panel
- curved panels are rather oly useable for hybrid ESLs.
- segmented panels, especially wire stators are easy to build and cost can be very low.
- using wires the otherwise critical Q of how to insulate the stator may already be answered.
- segmentation allows to widen the distribution character with a flat panel.
At the same it allows to a degree a amplitude equalization.
- The transformation factor is typically higher, but the transformer quality can be chosen lower, thereby cheaper transfomers and, more important, reducing the amplifier's load.
- Due to the reduced capacitance towards the top end it's easier to achieve higher bandwidth limits.
- flat panels allow the design of fullrange ESLs.
- in case of thinking about direct drive amplifiers, a segmented panel is rather a must.
I built both styles, flat wire stators and curved metal sheet stators.
Both have their merits and both principles are vastly superior to dynamic speakers.
Sonic differences are not as large between both styles as to dynamic speakers.
So choose after Your manual capabilities, optical taste, budget, or sourceability of parts, or whatever.
As a first time or beginners project I'd recommend a flat wire stators though.
jauu
Calvin
jauu
Calvin
the differences of segmented to non-segmented panels, whether curved or flat, have been discribed in some threads before (probabely a few years old by now).
Therefore just in short:
- non-segmented metal sheet stators typically achieve the highest efficiency and highest dynamic range.
- they require high quality transformers with very low leakage inductance, but mostly lower transformation factors.
- The lower requirement in drive voltage and transformation factor is positive regarding ageing, insulation and safety.
- The requirement for high quality lowloss transformers, combined with the high capacitance towards higher frequencies presents a more difficult load for the driving amplifier.
- curving a panel widens the distribution character at the top end of the frequency range (sweet spot) and increases the panel's mechanical stability/stiffness.
Its highly recommended for this kind of panel
- curved panels are rather oly useable for hybrid ESLs.
- segmented panels, especially wire stators are easy to build and cost can be very low.
- using wires the otherwise critical Q of how to insulate the stator may already be answered.
- segmentation allows to widen the distribution character with a flat panel.
At the same it allows to a degree a amplitude equalization.
- The transformation factor is typically higher, but the transformer quality can be chosen lower, thereby cheaper transfomers and, more important, reducing the amplifier's load.
- Due to the reduced capacitance towards the top end it's easier to achieve higher bandwidth limits.
- flat panels allow the design of fullrange ESLs.
- in case of thinking about direct drive amplifiers, a segmented panel is rather a must.
I built both styles, flat wire stators and curved metal sheet stators.
Both have their merits and both principles are vastly superior to dynamic speakers.
Sonic differences are not as large between both styles as to dynamic speakers.
So choose after Your manual capabilities, optical taste, budget, or sourceability of parts, or whatever.
As a first time or beginners project I'd recommend a flat wire stators though.
jauu
Calvin
jauu
Calvin
Hi Charlie,
Though I haven't tried it, but I am sure you can use heat shrink tubes with the weld rods for good insulation.
Wachara C.
Though I haven't tried it, but I am sure you can use heat shrink tubes with the weld rods for good insulation.
An externally hosted image should be here but it was not working when we last tested it.
Wachara C.
Sure, but pvc or nylon heatshrink is what you need. Relatively expensive though.
Ordinary one is cross linked polyolefine and has really low epsilon, low conductivity and low adhesion.
EDIT
HSTTP
Last edited:
The plan is to use black plastic louvers to contrast with the bright copper coated TIG rods and then spray paint with gloss clear polyrethane. The only black grids I could find are only .375 thick, which maybe too flexible for wider panels (mine will be 12"): BLACK EGG CRATE LOUVERS
I was inspired by Ken Seibert's project shown HERE:
The bitch is gonna be building 48" tall panels using 36" TIG rods. I will have to butt them at one of the horizontal spacer locations and either solder them together there or feed them from both ends.
I think Calvin's post here covered the advantages and disadvantages better that I could explain it.
I'm interested to see you have any problems getting the rolled wire straightened. With a roll of continuous wire, you may be able to stack two louvers face-to-face with a threaded rod at each end to set spacing, and them wrap them--- doing both stators at once would save a lot of time!
Hi Charlie, i haven't read the whole thread yet.
But little info for your First post.
I think .035 TIG would be great!!!
The only draw back about PCB is that you have to be quick and proficient or else to much heat will warp and /or destroy the material.
Some Plain ole .025" Copper or Brass Hobby Sheet plate my be a better alternative for making the individual sections and then mount the to your frame.
Whatever you choose for the base material to solder to, it helps greatly to tin the ends of each rod before you attempt permanently solder them.
Using this material you could even make slide buss bar switch to short the resistor sections as well.
I used PCB material to solder the ends to and it works very nicely.
I used some feeler gauges to properly (re) position the rods as I went along.
All of the pictures I had of this are now gone at ESLDIY !!
Go through the calculator before you decide on exactly how many segments you plan on having.
I used the doubling of the radiating area width per octave method as this reduces the number of sections.
I started with a center section width of 3/8" to 1/2" for 20Khz and then doubled the width of section on down to 2.5Khz as that was all of the width I had going for on my design.
The drawback to this is having many different values of resistors to contend with.
The second method would to use the same amount of rods per section and the same value of resistors for each section .
I think that it has been said that this method may give a smother transition of dispersion through out the range.
The First few sets of resistors starting from the center will have the most voltage across them in the string at the highest frequency's.
For these you may want to use a few in series to get the value you need for two reason's,
#1 Power handling the first few sets will be getting most all of the power you put in to the panel at the highest frequency's.
#2 The resistors Voltage Coefficient and/or max voltage rating, the voltage across the sections will be divided equally among the resistors as per frequency.
So this means the for the lowest frequency the voltage is equally divided between all of the sections.
So lets say you have 10 sections and 2500v of driving voltage at the lowest frequency each section will only have 250v difference between them.
But as the frequency rises then the number of sections gets reduced and there is more voltage between each section.
This also starts to require more resistors as well if you are using some common types that are only rated for 250v to 500v.
It may be cheaper in the long run, but then you have to have 4 or more resistors per the frist few sections, and this can become another construction issue as it is with my little panel design.
Anyhow if you can get some 2kv to 5kv resistors that have the value you need then all is good, but not for 134 sections at $2 to $4 a pop!!!
I guess this is not to much of a deal for average use, but it can stress the resistors greatly if you plan on running some high power test sweeps or maybe even if you listen to a lot of organ or electronica type of music at loud levels.
This was the other reason I choose the method I used.
My little panel is predicted to have a flat response with a slight rise at the the high end.
I chose this method to eliminate the first resistor and just use the treble control on my amp to Taylor the high end as I needed.
This is done with only 7 panel sections (4 sections electrically) with total a width of 3.5" end rod-to-end rod.
Therefore I would only need to add two more end sections of 1.75" each to bring it to a full 7" wide.
This fits well within the 1:100 D/S ratio as I do have it set for 1:50 for more excursion room being that is just a small panel.
The above formula will be my next built once I get some thinner TIG as well.
Even though I'll have to contend with many values of resistor's there won't be as many in overall quantity.
I still have my 11"x 36" 1/16 TIG rod stator's that I can either cut in half and use them for the extra sections or I can just finish them as is and use them for some bass only panels.
As far as Coating, I have as measured 10.5mil on my .0625 rods at .0835' total diameter.
As tested it withstands +13.8kv with the ground wire touching it.
This is using the same coating technique I had wrote about in the past using common Spray Acrylic Enamel.
It has been used in a recent build as well and I have yet to hear of any breakdown issues.
I know you already know about coating but I get at least a 1500v (guaranteed safe limit) to 1900v (max.) per mil with this stuff and I have not had the time to test any of the poly's yet.
According to any of the spec sheets I have read, poly's are only good for 450V to 900V per mil depending on which one you are using.
This is a big deal when you have to calculate your center to center rod spacing when you mount them and in the calculator spreadsheet.
Yes, the rods do ring as expected.
But, Once they are mounted and they get coated with 10mil of material it is not an issue anymore.
Especially if you have extra support sections throughout their length.
I would suggest to use some .045" Rod if you have it, I have given this some thought as well.
But if you only have .035 by all means use it I would use it, adding supports for stability at every 6" to 9" should be good.
Make sure you have the rods coated before (and tested) before you mount them to the Egg Crate material.
I glued my First screen's to the stuff before I coated them and this also may have contributed to a poor result do to micro cracking.
My black screens were coated before they were mounted and you know how those came out !!
It still pisses me off that I had torched them, But Hey, They cost next to nothing to build!!
Once you get a design going let me know and will Model it in LTSpice or Circuitmaker2000 for your resistors rating spec's.
Cheers!!!
jer
But little info for your First post.
I think .035 TIG would be great!!!
The only draw back about PCB is that you have to be quick and proficient or else to much heat will warp and /or destroy the material.
Some Plain ole .025" Copper or Brass Hobby Sheet plate my be a better alternative for making the individual sections and then mount the to your frame.
Whatever you choose for the base material to solder to, it helps greatly to tin the ends of each rod before you attempt permanently solder them.
Using this material you could even make slide buss bar switch to short the resistor sections as well.
I used PCB material to solder the ends to and it works very nicely.
I used some feeler gauges to properly (re) position the rods as I went along.
All of the pictures I had of this are now gone at ESLDIY !!
Go through the calculator before you decide on exactly how many segments you plan on having.
I used the doubling of the radiating area width per octave method as this reduces the number of sections.
I started with a center section width of 3/8" to 1/2" for 20Khz and then doubled the width of section on down to 2.5Khz as that was all of the width I had going for on my design.
The drawback to this is having many different values of resistors to contend with.
The second method would to use the same amount of rods per section and the same value of resistors for each section .
I think that it has been said that this method may give a smother transition of dispersion through out the range.
The First few sets of resistors starting from the center will have the most voltage across them in the string at the highest frequency's.
For these you may want to use a few in series to get the value you need for two reason's,
#1 Power handling the first few sets will be getting most all of the power you put in to the panel at the highest frequency's.
#2 The resistors Voltage Coefficient and/or max voltage rating, the voltage across the sections will be divided equally among the resistors as per frequency.
So this means the for the lowest frequency the voltage is equally divided between all of the sections.
So lets say you have 10 sections and 2500v of driving voltage at the lowest frequency each section will only have 250v difference between them.
But as the frequency rises then the number of sections gets reduced and there is more voltage between each section.
This also starts to require more resistors as well if you are using some common types that are only rated for 250v to 500v.
It may be cheaper in the long run, but then you have to have 4 or more resistors per the frist few sections, and this can become another construction issue as it is with my little panel design.
Anyhow if you can get some 2kv to 5kv resistors that have the value you need then all is good, but not for 134 sections at $2 to $4 a pop!!!
I guess this is not to much of a deal for average use, but it can stress the resistors greatly if you plan on running some high power test sweeps or maybe even if you listen to a lot of organ or electronica type of music at loud levels.
This was the other reason I choose the method I used.
My little panel is predicted to have a flat response with a slight rise at the the high end.
I chose this method to eliminate the first resistor and just use the treble control on my amp to Taylor the high end as I needed.
This is done with only 7 panel sections (4 sections electrically) with total a width of 3.5" end rod-to-end rod.
Therefore I would only need to add two more end sections of 1.75" each to bring it to a full 7" wide.
This fits well within the 1:100 D/S ratio as I do have it set for 1:50 for more excursion room being that is just a small panel.
The above formula will be my next built once I get some thinner TIG as well.
Even though I'll have to contend with many values of resistor's there won't be as many in overall quantity.
I still have my 11"x 36" 1/16 TIG rod stator's that I can either cut in half and use them for the extra sections or I can just finish them as is and use them for some bass only panels.
As far as Coating, I have as measured 10.5mil on my .0625 rods at .0835' total diameter.
As tested it withstands +13.8kv with the ground wire touching it.
This is using the same coating technique I had wrote about in the past using common Spray Acrylic Enamel.
It has been used in a recent build as well and I have yet to hear of any breakdown issues.
I know you already know about coating but I get at least a 1500v (guaranteed safe limit) to 1900v (max.) per mil with this stuff and I have not had the time to test any of the poly's yet.
According to any of the spec sheets I have read, poly's are only good for 450V to 900V per mil depending on which one you are using.
This is a big deal when you have to calculate your center to center rod spacing when you mount them and in the calculator spreadsheet.
Yes, the rods do ring as expected.
But, Once they are mounted and they get coated with 10mil of material it is not an issue anymore.
Especially if you have extra support sections throughout their length.
I would suggest to use some .045" Rod if you have it, I have given this some thought as well.
But if you only have .035 by all means use it I would use it, adding supports for stability at every 6" to 9" should be good.
Make sure you have the rods coated before (and tested) before you mount them to the Egg Crate material.
I glued my First screen's to the stuff before I coated them and this also may have contributed to a poor result do to micro cracking.
My black screens were coated before they were mounted and you know how those came out !!
It still pisses me off that I had torched them, But Hey, They cost next to nothing to build!!
Once you get a design going let me know and will Model it in LTSpice or Circuitmaker2000 for your resistors rating spec's.
Cheers!!!
jer
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.