A Segmented Wire Stator ESL with Selectable Wide/Narrow Dispersion

I was so inspired by my friend Ken Seibert's ESL project that I decided to have a go at building a pair of segmented wire stator ESL's myself-- mainly out of curiosity but it's nice to have a wider sweet spot available for when guests drop in.  

Detailed instructions with photos for building segmented welding rod stators are shown on Ken's excellent website here:  Ken Seibert Audio
My old perf-metal flat panels are still the best imaging speakers I've ever heard but their sweet spot is only wide enough for one person so they aren't much good for entertaining guests.   

It occurred to me that it might be possible to have it both ways. That is; narrow dispersion for best slam and imaging at the sweet spot and wide dispersion for entertaining guests.  So this project is a segmented wire-stator ESL with selectable wide and narrow dispersion modes. 

Each stator is a vertical array of (132) .035" diameter copper coated TIG welding rods glued onto a black plastic egg crate type florescent light diffuser. 

The welding rod conductors are physically segmented into (11) discrete groups of (12) wires and electrically segmented into (6) discrete groups; consisting of a center group with (5) paired left/right groups on either side.   

Wide dispersion is achieved the same way Peter Walker did it in the Quad ESL 63; using electrically segmented conductors receiving sequentially delayed signals via a resistor/capacitor delay network. 

But where the Quad 63 used concentric rings of conductors to emulate a point source emitting a spherical wave front, my panels use vertical wire groups to emulate a line source emitting a cylindrical wave front. 
When in narrow dispersion mode, all wire groups are directly coupled to the amps/transformers and the panels emit planar wave fronts that focus to an intense narrow sweet spot for best imaging and slam.  

When in the wide mode, the paired wire groups on either side of the center group receive sequentially delayed signals, and the panels then emit cylindrical wave fronts that spread to a wider sweet spot.  

The wire segmentation and delay networks were derived using the Segmented ESL Calculator  spread sheet developed by my friend Steve (a.k.a. "Bolserst") on the DIY Audio Forum.  The capacitance of the wire groups themselves serve as the capacitors in the network so only resistors are required to build it.  

If you're a math dummy like me, it's nice to have smart friends! 

For this build I opted for horizontal support spacers and mechanically tensioning the diaphragm, predominantly in the vertical direction only, using the vertical stretcher jig shown below.  I found it necessary to also apply some horizontal tension by hand using tape strips in order to ensure there would be no wrinkles in the film.  

First listening impressions:  

They sound perfectly electrostatic and wonderful.  Imaging in narrow mode is every bit as magical as my old perf-metal panels.  

Imaging in wide mode loses some of the magic and slam but is still excellent and doesn't put your head in a vise like narrow mode.  I would say wide mode is less "in your face"... non-fatiguing and easy to listen too.  Most of the time I'm content to leave the panels in wide mode, even for solo listening.   
Some pics from my segmented ESL build: 

Segmented ESL Schematic

Spacing the rods with 1/2-13 tpi all-thread rods

The charge ring connection

Tensioning & bonding the diaphragm
Front & back stators ready for assembly

Soviet military rotary switch (mode selector): 

Mode switch box, resistor network and ESL panel

New panel installed in the beam splitter cabinet

View from behind