Which spacer thickness to use with ESL?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
arend-jan said:
The advantage of wire does not come from the orientation but the fact that you can hook each wire up with series resistors to create the electrical segmentation.

To get the same effect with the circuit board you can etch the copper to create electrically separated zones (strips). I would create vertical strips and the high frequency strip should be 3cm wide or less. If you want to do it really nice you would put the smallest strip in the middle, then flanked with two wider strips on both sides (for the middle/low frequencies) and so on.

Hi Arend-jan,

Thanks for your suggestion. Do you mean that by putting resisters series with the copper of each section would help produce different frequencies? What resister values, watts, and type of resisters can you suggest? Wow, this is a new knowledge for me.

The panel width that I will be using is 40 cm. I will probably use 2.5 cm on both left and right side for gluing mylar - with 35 cm left for perforations. With you suggestion, do you think that dividing the PCB into 3 sections of 10 cm, 3 cm, and 20 cm (2 sections of 1cm of copper will be cut out for proper spacing) will be good?

Can you show me more pictures? :)

Wachara C.
 
You've probably already figured this out, given the other responses, but I was suggesting that you keep the PCB in one piece, but electrically isolate three copper sections from each other.

To figure out the frequency response of a stator and resistor combination, use this equation:


f_c = 1/(2 pi R C)

where f_c is the cutoff frequency, R is the resistance of the resistor, and C is the capacitance of the section of the ESL panel that you're working with. If you put the resistor in series with the stator (between the transformer and the stator) then frequencies well below f_c will pass unattentuated, and frequencies above f_c will drop off at 6 dB per octave. Putting the resistor after the transformer increases the voltage it must be able to withstand but reduces the current it needs to be able to handle. You could put the resistor before the transformer but then you have to be sure to use a different value of C in the equation. The transformer changes the effective value of C quite dramatically.

If you have trouble finding resistors that can handle the high voltages you can create R by using several resistors in series. For example, if R = 5000 ohms you can use five 1000 ohm resistors in series. (I'm picking these numbers purely to make the math easy.) This has the effect of reducing the voltage across each resistor by a factor of five.

Few
 
Few said:
You've probably already figured this out, given the other responses, but I was suggesting that you keep the PCB in one piece, but electrically isolate three copper sections from each other.

To figure out the frequency response of a stator and resistor combination, use this equation:


f_c = 1/(2 pi R C)

where f_c is the cutoff frequency, R is the resistance of the resistor, and C is the capacitance of the section of the ESL panel that you're working with. If you put the resistor in series with the stator (between the transformer and the stator) then frequencies well below f_c will pass unattentuated, and frequencies above f_c will drop off at 6 dB per octave. Putting the resistor after the transformer increases the voltage it must be able to withstand but reduces the current it needs to be able to handle. You could put the resistor before the transformer but then you have to be sure to use a different value of C in the equation. The transformer changes the effective value of C quite dramatically.

If you have trouble finding resistors that can handle the high voltages you can create R by using several resistors in series. For example, if R = 5000 ohms you can use five 1000 ohm resistors in series. (I'm picking these numbers purely to make the math easy.) This has the effect of reducing the voltage across each resistor by a factor of five.

Few

Hi Few,

Thanks again for your kind advice. I will try to make a set and will report back to you soon.

Wachara C.
 
Hi,

what is an open stator? What is an closed stator?

electrically the best would be to have 100% of the diaphragm area ´covered´ with stator area. Acoustically 0% coverage would be the best. Both solution extremes would give no acoustic output at all(!) with a symmetrical ESL-Design since in the first case no sound would reach through the massive stator-plates and in the second case, there would be no driving force (no stator) at all.
So every ESL works with a partially open design and every designer tries to find a good compromise (which is for most designs between 20%-40% of openness). The openness varies depending on the materials and the design goals. Efficiency for example could -and should be- the prime goal. Here punched metal sheet stators reach the best results. Wire stators, even though they should be better by simple theory, are inferior. Wire on the other hand has advantages with larger d/s-designs and the percentage of openness of a wire stator can easily be taylored to the designers needs. Too their rounded structure is at least prone to arcing. PC-board stators could theoretically reach slightly better efficiency than wire. Their disadvantage beeing the straight shape of the hole´s inner surface when simple routing methods are used which could reduce the panel´s bandwidth to less than 20kHz!
So as usual, the best compromise depends on the aims and goals the designer defines.

jauu
Calvin
 
Calvin said:
Hi,

what is an open stator? What is an closed stator?

electrically the best would be to have 100% of the diaphragm area ´covered´ with stator area. Acoustically 0% coverage would be the best. Both solution extremes would give no acoustic output at all(!) with a symmetrical ESL-Design since in the first case no sound would reach through the massive stator-plates and in the second case, there would be no driving force (no stator) at all.
So every ESL works with a partially open design and every designer tries to find a good compromise (which is for most designs between 20%-40% of openness). The openness varies depending on the materials and the design goals. Efficiency for example could -and should be- the prime goal. Here punched metal sheet stators reach the best results. Wire stators, even though they should be better by simple theory, are inferior. Wire on the other hand has advantages with larger d/s-designs and the percentage of openness of a wire stator can easily be taylored to the designers needs. Too their rounded structure is at least prone to arcing. PC-board stators could theoretically reach slightly better efficiency than wire. Their disadvantage beeing the straight shape of the hole´s inner surface when simple routing methods are used which could reduce the panel´s bandwidth to less than 20kHz!
So as usual, the best compromise depends on the aims and goals the designer defines.

jauu
Calvin


Hello Calvin ,

Why are wired stators inferior ?

Wire is insulated , why would it be more prone to arching ?

Yes efficiency would be lower , but the advantages out way the lower efficiency .

Have you had a chance to compare both ?

Regards,

A.Wayne
 
Hi,

wire stators are inferior to punched metal sheet stators with regard to efficiency.
Excuse me, but english is not my mother tongue. It should have been: "...the round structure is the least prone to arcing".
The advantage of wire is that it is ready insulated available at ridiculous cost and can easily be handled and worked with.
But again...the best is what suits the demands of the designer best. And I prefer metal sheet stators, since they suit my demands best, as there are: optimized efficiency and great dynamic range, lowest distortion values, optics.
To reach these aims the sheets I have the metal sheets treated and insulated in several delicate, time and money consuming steps (which no other manufacturer does at the moment). The results are exceptional resolution and a dynamic range comparable to a large horn at distortion values lower than I´ve seen any other before!

If I´d design a panel with increased d/s, I´d surely opt for a wire stator. But a ESL shines only, when it doesn´t need to provide for large excursions, hence small d/s. So I leave the frequency range below ~200Hz where excursion is needed to dynamic speakers and combine the best of both worlds.

jauu
Calvin
 
Calvin said:
Hi,

Efficiency for example could -and should be- the prime goal. Here punched metal sheet stators reach the best results. Wire stators, even though they should be better by simple theory, are inferior.
jauu
Calvin


Why should a wire stator show lower efficiency? Because of their thicker insulation?
Audiostatic (large spacing) as well as Capaciti modules (smaller spacing) are simply good, with wires! No problems with low efficiency as far as I know
Open area of a famous dutch DIY design (ESL175) was around 60%! With wires of course.
 
Calvin said:
Hi,

If I´d design a panel with increased d/s, I´d surely opt for a wire stator.

jauu
Calvin

Hello Calvin,

Currently I am building a couple of wire ESLs. It will have open area of ~27%, radiating area 0.35 m^2. I use PVC insulated single stranded wire(O/d 2 mm). Why do you say that wire panels are good only with larger d/s spacing ? I plan to use 1.5 mm.

Regards,
Lukas.
 
Hi,

I think I have to improve on my english a lot :rolleyes:
I wrote: "inferior with regard to efficiency".
I didn´t write:"...vastly inferior with regard to efficiency".
The lower efficiency has been tested and prooved in praxis.
I assume two reasons for this:
1) With same d/s a metal sheet stator typically has a less thick insulation layer, simply because it´s not easy to manufacture thick coatings with reasonable effort and cost. So with same epsilon-values of the insulator material a thicker insulation layer means higher signal loss, hence less efficiency.
2) the mean d/s distance of the copper surface is greater with round wires than with the rather flat metal sheet.
Already fractions of a mm make a non negligable difference in the efficiency values.

I wrote: "If I´d design a panel with increased d/s, I´d surely opt for a wire stator."
I didn´t write: "wire panels are good only with larger d/s spacing"
When You analyze and simulate different stator-structures You will find, that in very close proximity to the stator surface the electrical field distribution is not very homogenous. Field strength is higher between adjacent metal surfaces of the two stators than it is between the holes. It needs a small distance to the stator surface for the field to become homogenous. This distance is smaller with metal sheet stators featuring small hole diameters and small distances between the holes (btw. this leads to lower distortion values too). In other words...To reach the same level of homogenity of the field I have to distance wire stators further apart than with a punched metal sheet stator. So if I design for small d/s distances (<1.5mm), the metal sheet is my first choice. With larger d/s (>1.5mm) the handling and cost advantages of wire and the possibility of electrical segmentation would make it my prime choice. With greater d/s You can also use thicker wire, up to standard H07VU, like Audiostatic does. You won´t need as much wires per panel and less effort to straighten out the wires and the wire is mechanically quite strong. It doesn´t need many supporting frame ribs. With small d/s You´d need thin and less strong wire, many wires per panel and more supports. Metal sheets, especially when curved, exhibit high mechanical strength and rigidity even though the panel builds thinner than a wire panel. You don´t need supporting structures other than the mounting frame.

jauu
Calvin
 
Hi Calvin,

Your comment about the use of PCB as stator would not allow bandwidth to be more than 20kHz confuses me. What do you mean when you say "Their disadvantage beeing the straight shape of the hole´s inner surface when simple routing methods are used"? Do you mean that the holes being punched out are not the same as the holes being routed out? As far as I can see they are equally round. I am really confused.

Wachara C.
 
Hi,

if a hole is just punched or routed or lasered) straight through the sheet such that the ´walls´ take the form of a cylinder, each of the small cylindrical holes will act as a tiny Helmholtz-resonator, ie like a bassreflex casing.
The bassreflex is a box-tuning with bandpass character. Dependant of the relations of hole number count, dimensions of the holes, and the enclosed air-volume between the stators the upper bandwidth limit could be as low as 10kHz. A narrowing or kind of double flare of the crossection of the hole´s shape reaches a much higher upper bandwidth. This is automatically the case with wires as well as thickly coated punched metal, but not the case with thin, or not at all coated PC-board and just punched (or routed or lasered) sheet metal stators. There are formulas to calculate the bandwidth limit, but to my experience these formuas are notv very precise. But at least they give You an idea if Your stator design has flaws regarding this point.

jauu
Calvin
 
A wired stator is less efficient due to having less field area vs a metal stator , but it suffers from less refraction and hence tend to subjectively sound better than punched metal stators ...

Wired stators with large openings tend to sound and measure better than their metal counterparts , but as Calvin has suggested will be less efficient and more difficult to build compared to a punched hole metal stator..

The efficiency can be addressed by the amount of loops ( turns ) and the field gap. I would suggest listening to both types before making your decision as both can be made to play to reasonable loud levels...

For me the open type ( grid ) ESL is the best .............................

Hers is a pic of the newest quad , just to show the grid ( note this is not the stator, I'm showing the grid so most can see what we are talking about ) then picture it with a wired loop, of course you would use a much shallower grid pattern , what was typically used was about 1/4" or slightly less , of course this is dependent on wire size....
 

Attachments

  • quad.jpg
    quad.jpg
    35.5 KB · Views: 218
Hi,

the distribution character of a panel is solely dependant on the physical dimensions. Refraction at the stator holes doesn´t play a role to this as far as I know. Stators with equal opening percentage tend to be sonically equal too, apart from the one with higher efficiency tending to sound more ´dynamic´.
Stators with high opening percentage surely don´t measure better, especially under higher power conditions, because of the greater inhomogenity of the electrical field. A smaller percentage of openness could even help to reduce the problem of the high Q of the base resonance a bit, which you have to tackle with additional mechanical or electrical damping.

Wayne, the way You use ´grid´ lets me assume, that we are talking about different things. The grid, that can be seen in the Quad-picture just has the function to mechanically stabilize the stator plates, which in this case are kind of PCB-style. The openness of a stator has basically nothing to do with this ´supporting structure´. The panel could even work completely without it.

jauu
Calvin
 
chinsettawong said:
Hi a.wayne,

From your picture, I can't see very well what kind of stator they are using. The outside grids look to be just a supporting structure of the stator.

Wachara C.


Hello ,
If you read my post again you will see where i had said it was just a support structure for the speaker grille.

The picture was posted to show the egg crate look,of an open stator, as this is what i was alluding to when i had mentioned an open wired stator ... the one in the pic is not the stator .
 
Hi,

@a.wayne:
Still don´t have a clue what You mean by ´open stator´.
Every stator has to have a certain degree of openness! Otherwise there would be no functionality (sound) at all. Since the optimum percentage of ´openeness´ lies within a relatively small target range, metal sheet-, PCB- and wire-stators are equally well suited to reach that range. metal sheets having the advantage of not necessarily needing mechanically supporting structures, while PCB- and wire stators do need those structures. Wire stators could easily be constructed for a larger percentage of openeness, but that would be far off of the optimum value, so that any discussion about it is useless. So what do You mean with open (wire) stator?

@chinsettawong
Quad doesn´t use wire stators: see post #34
The smart thing about PCB-material is, that it can easily be designed with CAD-programs and could be premanufactured with good precision. The electrical characteristics of FR4-base material are very competitive to other insulative materials and the stators would be quite safe. Electrical segmentation is possible.
The best solution would be to have the copper layer encapsulated between two base-layers as it is written in R.Sanders patent (btw. could anyone tell me what is new with this patent? The described techniques are sota for at least 30years!!). You won´t need any more coating but solely for optical reasons. So it´d be good to have the hole´s shapes fabricated with taperings or roundings. For example routed with a special tool of appropiate form.
Quad used just single sided base material and coated it with insulative laquer. Here a functional and thick coating is necessary.
The bandwidth problem depends on several parameters:
- enclosed air volume of the stators
- number of holes
- diameter of holes, or dimensions of the openings if its not a circular hole
- length of the hole, i.e. thickness of the stator sheet
- crossectional shape of the holes

The trick is, that the air within the hole shouldn´t behave like a small piston. A non-cylindre-crossection is good, like trapezoid or flared, double-flared or rounded. If you can´t provide for such a shape the ´length´ of the cylindre, i.e. the thicknes of the sheet, should be smaller than the hole´s diameter. So larger holes allow for thicker sheets, while smaller holes ask for thinner sheets.

jauu
Calvin
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.