This design uses the Jordan JX-92 midbass driver in a transmission line enclosure. This is an overachieving little 5.25" driver. Above the Jordan driver will be a 6" x 20" electrostatic panel. My initial thoughts for this speaker is first to try to a completely passive crossover. The panel will naturally roll off at 6 dB per octave at what I think will be in the 3 KHz range. With only a single inductor, I can cross the Jordan driver over to the panel. I'm concerned that the Jordan will dominate the timber of the speaker. My second, and probably more likely to succeed, method is to use the same sort of active crossover technique that is currently working well on the DIY ESL 1.0. With the active crossover, I can also add a touch of bass lift to extend the lower end of the speaker.
Mini Mite Initial Plans (PDF 8KB)
A movie of the build-up process is available below for those of you interested. It's fairly large (5.1 meg), so be warned.
An ESL Panel Construction Slide Show (MOV-file 5.1 MB)(Available from Sheldons site only)
I've designed another high voltage supply for this speaker. It should work quite well and is cheaper to build than the switching supply I built earlier for the DIY ESL 1.0. This supply contains about $35.00 worth of parts per channel. It is capable of generating about 7500 volts max.
Simple ESL Polarizing Supply (PDF 42KB)
The bass modules are fairly easy to construct and seem to perform well. I built them up, and stuffed them using the driver impedance curve to tell me when I had stuffed the line enough. The unstuffed line looks like an odd bass reflex enclosure. Adding stuffing causes the lower impedance peak to disappear and the larger one to broaden. If the line has parallel walls, there will most likely be additional resonant peaks above the two large bass reflex type peaks. Stuffing will get rid of these peaks as well. the trick is to slowly add stuffing to the line and take impedance measurements. The line should be stuffed just until the lower peak disappears, and the upper peak becomes a smooth single peak. this can be seen in the impedance plot here. The ideal stuffing for this transmission line is 6 ounces.
After completing the Power Supplies, I was able to test the little ESL panels. I have some old output transformers from a Fisher X-100 integrated amp (EL84 tubes) which I thought would be a great match with these little panels. However The leakage inductance seems to be pretty high and that is causing a rising treble response and a peak at about 10 KHz. Pictures and Plots are shown below
Spec sheets: Jordan TL bass
The line should be stuffed just until the lower peak disappears, and the upper peak becomes a smooth single peak. This can be seen in the impedance plot here. (click on the picture to enlarge)
Jordan Impedance (click on the picture to enlarge)
Jordan Frequency response, (click on the picture to enlarge)
Jordan Polar response, (click on the picture to enlarge)