Hi everyone,
Doin the required study before making a pair of wire stator ESL's for myself, reading up on ESL tech, how to's etc... im a bit confused about "segmentation".
I see people used electrical ladder segmentation (using the excel tool) which i understand BUT the panels themselves are divided in segments too,
to prevent ignoring the rule 1:50 /100 for the width of unsupported foil.
Wont this be mechanical segmentation? and how does it effect the electrical one?
Also in the tool the number of wires used in a "segment" is a bit "grey" it doesnt change values of the ladder resistors?
As you can read/see, i need help and clarification 🙂
Thanks all,
Jerolee
Doin the required study before making a pair of wire stator ESL's for myself, reading up on ESL tech, how to's etc... im a bit confused about "segmentation".
I see people used electrical ladder segmentation (using the excel tool) which i understand BUT the panels themselves are divided in segments too,
to prevent ignoring the rule 1:50 /100 for the width of unsupported foil.
Wont this be mechanical segmentation? and how does it effect the electrical one?
Also in the tool the number of wires used in a "segment" is a bit "grey" it doesnt change values of the ladder resistors?
As you can read/see, i need help and clarification 🙂
Thanks all,
Jerolee
Mechanical and electrical segmentation are two different things, without much affect on each other as long as the dead areas are reasonably small. The Quad ESL-63 and its successors are good examples of this. The central circle and ring sections span independent panels, but the electrically connected segments act as one acoustically. And the electrical segmentation is smaller than the mechanical segmentation, which is normal for most designs where wider dispersion is the goal.
https://www.diyaudio.com/community/threads/quad-63-988-bass-and-tweeter-panels.277486/
Of course they can also work together if your layout is amenable to that. This would typically occur in a panel that's only running something like 2 or 3 segments, instead of the ladder-style ones that typically use more.
https://www.diyaudio.com/community/threads/quad-63-988-bass-and-tweeter-panels.277486/
Of course they can also work together if your layout is amenable to that. This would typically occur in a panel that's only running something like 2 or 3 segments, instead of the ladder-style ones that typically use more.
I'm not sure I can answer your questions, but perhaps reviewing my segmented ESL panel layout will be helpful:
Referring to the sketch below, my panel has 15 physical segments but only 8 electrical segments:
The panel is 12" x 48", and the diaphragm vertically divided into three equal width sections.
Each diaphragm section is driven by 30 wires (3 X 30 = 90 wires total).
Physically, the 90 wires are divided into 15 groups of six wires.
Electrically, these 15 wire groups are driven as eight (8) discrete segments.
These eight electrical segments are arranged as (1) center wire group with (7) left/right paired wire groups on either side (7 + 1 + 7).
In the Segmented ESL Calculator Excel spreadsheet program, the "Number of Segments" input would be "8" (not 15).
Also, the Calculator "length" and "width" inputs should reflect only the area actually occupied by wires (exclude edge spacers & diaphragm supports).
For a hybrid ESL; the "LF Cutoff" frequency should not be lower than the woofer crossover. I recommend 20Hz above the crossover frequency.
Additional information can be found on my website here:
https://jazzman-esl-page.blogspot.com/
Referring to the sketch below, my panel has 15 physical segments but only 8 electrical segments:
The panel is 12" x 48", and the diaphragm vertically divided into three equal width sections.
Each diaphragm section is driven by 30 wires (3 X 30 = 90 wires total).
Physically, the 90 wires are divided into 15 groups of six wires.
Electrically, these 15 wire groups are driven as eight (8) discrete segments.
These eight electrical segments are arranged as (1) center wire group with (7) left/right paired wire groups on either side (7 + 1 + 7).
In the Segmented ESL Calculator Excel spreadsheet program, the "Number of Segments" input would be "8" (not 15).
Also, the Calculator "length" and "width" inputs should reflect only the area actually occupied by wires (exclude edge spacers & diaphragm supports).
For a hybrid ESL; the "LF Cutoff" frequency should not be lower than the woofer crossover. I recommend 20Hz above the crossover frequency.
Additional information can be found on my website here:
https://jazzman-esl-page.blogspot.com/
Hi Charly,
Now u even confuse me more 🙂
quote "Referring to the sketch below, my panel has 15 physical segments but only 8 electrical segments:" unquote
15 physical segments? i dont think that correct? 15 wire supports u mean?
Now u even confuse me more 🙂
quote "Referring to the sketch below, my panel has 15 physical segments but only 8 electrical segments:" unquote
15 physical segments? i dont think that correct? 15 wire supports u mean?
You're not alone... I too was quite confused until it was explained to me.
I will assume the Excel file you referred to is Steve Bolser's "Segmented ESL Calculator".
This Calculator provides for several optional segmentation schemes:
Number of segments:
You can choose as many segments / wire groups as you wish. Generally; more/smaller wire groups/segments gives wider/smoother-trending dispersion, but there is no advantage using wire groups smaller than 12 mm wide (wavelength of highest audible frequency).
My panel has BOTH stators segmented, and an ODD number of same-size wire groups, powered as 8 electrical segments.
****************************************************************
Back to your question:
The wire lacing sketch in the earlier post shows the wires initially laced up as a single continuous strand.
After the wires were stretched to make them straight, the wire loops were strategically cut to physically separate the 90 vertical wires into 15 separate 6-wire groups. These are the 15 physical segments.
Below is my speaker schematic showing the front and rear stators, with the diaphragm between them.
In this sketch; each 6-wire group is represented by a dotted RED line.
Note that the center wire group is numbered 1, and the groups on either side are numbered 2 thru 7.
The panel is thus symmetrically segmented on centerline, as wire groups 8 7 6 5 4 3 2 1 2 3 4 5 6 7 8.
Now the (8) Electrical Segments:
E-Segment 1 = wire group 1 (only), fed directly from the step-up transformer lead.
E-Segment 2 = LH wire group 2 + RH wire group 2, fed thru 1 set of added series resistors.
E-Segment 3 = LH wire group 3 + RH wire group 3, thru 2 sets of added series resistors.
E-Segment 4 = LH wire group 4 + RH wire group 4, fed thru 3 sets of added series resistors.
... and so on, thru E-Segment 8.
The series resistors couple with the wire group capacitances to form a series of low pass filters which tailors the dispersion by progressively chopping off the highs, from the panel centerline outward.
I will assume the Excel file you referred to is Steve Bolser's "Segmented ESL Calculator".
This Calculator provides for several optional segmentation schemes:
- Only one stator segmented
- Both stators segmented
- Odd number of wire groups with each group having the same number of wires.
- Even number of wire groups with center group having 2X the number of wires as remaining groups.
Number of segments:
You can choose as many segments / wire groups as you wish. Generally; more/smaller wire groups/segments gives wider/smoother-trending dispersion, but there is no advantage using wire groups smaller than 12 mm wide (wavelength of highest audible frequency).
My panel has BOTH stators segmented, and an ODD number of same-size wire groups, powered as 8 electrical segments.
****************************************************************
Back to your question:
The wire lacing sketch in the earlier post shows the wires initially laced up as a single continuous strand.
After the wires were stretched to make them straight, the wire loops were strategically cut to physically separate the 90 vertical wires into 15 separate 6-wire groups. These are the 15 physical segments.
Below is my speaker schematic showing the front and rear stators, with the diaphragm between them.
In this sketch; each 6-wire group is represented by a dotted RED line.
Note that the center wire group is numbered 1, and the groups on either side are numbered 2 thru 7.
The panel is thus symmetrically segmented on centerline, as wire groups 8 7 6 5 4 3 2 1 2 3 4 5 6 7 8.
Now the (8) Electrical Segments:
E-Segment 1 = wire group 1 (only), fed directly from the step-up transformer lead.
E-Segment 2 = LH wire group 2 + RH wire group 2, fed thru 1 set of added series resistors.
E-Segment 3 = LH wire group 3 + RH wire group 3, thru 2 sets of added series resistors.
E-Segment 4 = LH wire group 4 + RH wire group 4, fed thru 3 sets of added series resistors.
... and so on, thru E-Segment 8.
The series resistors couple with the wire group capacitances to form a series of low pass filters which tailors the dispersion by progressively chopping off the highs, from the panel centerline outward.
Last edited:
BTW; I will be happy to give you my drawing and parts list but it only works for a 12 x 40 panel size an UL-1061 20awg single-strand wire.
Jerolee,
After re-reading your post; it seems I forgot to address one of your questions...
"Also in the tool the number of wires used in a "segment" is a bit "grey" it doesnt change values of the ladder resistors?"
The Excel Calculator does not use the number of wires to calculate the resistance or capacitance. Rather, it assumes the panel is constructed of two parallel metal plates, and applies the standard parallel plate capacitance formula, using the plate area and gap between plates.
The capacitance is calculated from the Inputs for LF Cutoff, D/S (diaphragm-to-stator gap), Length and Width (i.e. panel area). The panel isn't two solid plates so the calculation will not be exactly correct, but close enough for our purpose.
The resistance is then calculated as the capacitance divided by the number of electrical segments.
*********************
Most panels have diaphragm supports within the active panel area which divides the diaphragm into two or more sections. When the diaphragm sections are also referred to as "segments" it can get really confusion as to which "segments"we are talking about.
In the case of my panel;
There are two vertical diaphragm supports which divide the diaphragm into three vertical sections.
Each section is driven by 30 wires (five 6-wire groups, actually)
The three diaphragm sections are collectively driven by 90 wires (15 physical six-wire groups)
The (15) physical wire groups are electrically powered as eight (8) segments (see previous post).
We see how this can be very confusing, depending on how we define "segments"
Perhaps the most confusing inputs on the Excel spreadsheet are "Width" and "No. Sections".
- Width must include be only the areas actually occupied by wires, not the total panel width.
(the wires in my 12" wide panel collectively occupy only 8.2" width, so the correct Width input is "8.2")
- My panel electrically divides the 15 wire groups into 8 driven segments, so the No. Segments input is "8"
After re-reading your post; it seems I forgot to address one of your questions...
"Also in the tool the number of wires used in a "segment" is a bit "grey" it doesnt change values of the ladder resistors?"
The Excel Calculator does not use the number of wires to calculate the resistance or capacitance. Rather, it assumes the panel is constructed of two parallel metal plates, and applies the standard parallel plate capacitance formula, using the plate area and gap between plates.
The capacitance is calculated from the Inputs for LF Cutoff, D/S (diaphragm-to-stator gap), Length and Width (i.e. panel area). The panel isn't two solid plates so the calculation will not be exactly correct, but close enough for our purpose.
The resistance is then calculated as the capacitance divided by the number of electrical segments.
*********************
Most panels have diaphragm supports within the active panel area which divides the diaphragm into two or more sections. When the diaphragm sections are also referred to as "segments" it can get really confusion as to which "segments"we are talking about.
In the case of my panel;
There are two vertical diaphragm supports which divide the diaphragm into three vertical sections.
Each section is driven by 30 wires (five 6-wire groups, actually)
The three diaphragm sections are collectively driven by 90 wires (15 physical six-wire groups)
The (15) physical wire groups are electrically powered as eight (8) segments (see previous post).
We see how this can be very confusing, depending on how we define "segments"
Perhaps the most confusing inputs on the Excel spreadsheet are "Width" and "No. Sections".
- Width must include be only the areas actually occupied by wires, not the total panel width.
(the wires in my 12" wide panel collectively occupy only 8.2" width, so the correct Width input is "8.2")
- My panel electrically divides the 15 wire groups into 8 driven segments, so the No. Segments input is "8"
Last edited: