Wings, Wetzlar 2005
Wings, Wetzlar 2005

In 2005 the 2nd HiFi-Music-World, a DIY-show took place in Wetzlar.

Driven by the wish to promote the electrostatic loudspeaker principle and the good personal contact to the organizers of the show, the idea rose to have a booth where a non-commercial hobby DIYer could present his speakers and concepts and at the same share his KnowHow with the visitors in form of seminars. The idea set off and the non-commercial hobbyists semianrs have become a fix part of the show.

At the show the Wings played as satellites together with a dipolar Mono subwoofer driven via an active crossover.

At the stipulated crossover frequency of 100Hz the acoustic phase cancellation is already very effective with a 10" wide panel.

If one desn´t want to correct too much for the acoustic phase cancellation one can only widen the baffle of teh speaker, which offsets the effect towards lower frequencies.

So i added simple plain 6" wide ´wings´, made of high density board and tilted slightly backwards by 15°.

The obsolete Bass casing was replaced by a small casing below the panel, containing the electronics.

Regarding the audio trannies I experimented with simple power transformers. Till then I used ´special´ audio transformers manufactured by Sombetzki (rem: a german ESL manufacturer) in an array of 4.

The fact that the transformers somehow looked very similar to 10VA EI-core power transformers, the array connection and not the least the fact that most of the trannies had died by internal flashovers, made me search for alternatives. A test with a sixpack of 10VA-EI-core power transformers (1x230V/1x6V) in a series-parallel array connection lead to nearly identical results apart from a slightly reduced transformation factor. Amplitude response and sound was the same at just 1/4 the cost. And so far not one of these transformers has died of internal flashovers.

The only drawback of these transormers is there low primary inductance value which discards them for fullrange useage. They can only be used from the lower midrange on.

The series-parallel array connection improves the behaviour considerably, because it easily allows for more windings and increased core area crossection. This goes with higher possible power ratings and lower low frequency bandwidth limit. The value of the strayinductance sinks at the same. The strayinductance forms together with the panels capacitance a resonating circuit at the upper bandwidth limit. Above this limit the amplitude response drops with -12dB/oct. You can see this resonance as a minimum in the electrical impedance plot. With good audio trannies, the resonance point lies above the audible range. Some commercial ESLs using crappy transformers showed the resonance within the audible range (MartinLogan Sequel 12kHz, Audiostatics <15kHz).

The highfrequency resonance can be tamed with a series resistor, to achieve a more linear amplitude response. At the same this resistor increases the low impedance value at this point to ´friendlier´ values for the driving amplifier. This is a smart way to change a bug into a feature.

No question that these crappy cheap parts are marketed as special and expensive devices. Without any doubt can better results without relying on tricks be achieved with better audio transformers. Lower distortion figures and higher upper bandwidth limit are pronounced effects.

Good transformers do cost more rightfully!