Cheap toroidal transformer for ESL panel


I'm interested in creating a new mid / tweeter esl panel together with a dynamic woofer. The advantages over a full range esl would be:

  • cheaper transformers
  • higher efficiency, lower polarizing voltage required, less aging of materials?
  • no damping screen required to tame fundamental "one note bass" resonance, which leads to better, more transparent, less muffed sound, in my personal experience (biggest reason)
  • lower step up ratio required of audio transformer resulting in extended HF range compared to full range audio transfomer

I would like to chose my cross-over frequency somewhere between 150 and 250 Hz. The esl panel will be electrically segmented.

So I've read some topics here about cheap toroidal power transformers. Unfortunately I lack some technical background, so I have some questions before I order them:

  • which primary voltage should I choose: does it make a difference and if so, which one is preferable to the other? For instance, should I pick 2x6V, 2x7V, 2x8V, 2x9V, 2x12V, 2x15V or 18V? I would think a lower voltage would result in a higher step-up ratio and lower saturation frequency limit?
  • If I'm correct, 1x230V is preferrable to 2x115V as these have a high capacitance and limit the bandwidth*
  • Would 15VA power rating suffice? If I'm correct, this would suffice for electrically segmented panels, not for unsegmented big panels*

* source:

I managed to find two transformers and I'm wondering if they would be a candidate?

230V ac, 2 x 9V ac, 50VA 2 Output

230V ac, 2 x 6V ac, 50VA 2 Output

BSAB Toroids is a source of trannies I like to use, especially with non-segmented metal sheet stators.
I never messed with dual-sec toroids -apart from the dedicated Amplimo/Plitron ESL trannies- but only used pairs of single-sec standard power toroids.
So far not a single one went gonski or went up in smoke .
A pair of 50-80VA 6V/230V for panels up o ~1nF capacitance, 80-120VA 9V/230V for panels up to ~2nF capacitance.
The transformation factor U multiplied by the efficiency figure (<1) gives a good estimate of the ´real´ transformation factor and hence a figure to calculate the transformed impedance the amplifier will see.
Don´t only think in ´voltage-terms´ when it comes to tansformation factors but also keep impedance matching in mind.
So with segmented panels it might even be advantageous to use a 2-pair configuration with doubled U

The toroidal transformers have relatively low leakage inductance, but they 've a lot of stray (parasitic) capacitance (bigger than most of panels, especially the segmented one). It provides good hi-freq. extension at unsegmented panels, but causes more severe impedance dip at higher frequencies (or must be decrease the step-up ratio).
  • Would 15VA power rating suffice? If I'm correct, this would suffice for electrically segmented panels, not for unsegmented big panels*
The power rating (230V winding's current capability) is OK for segmented panel. The smaller trafos have bigger inductances (shunt and leakage), but smaller capacitances. The 230V winding can produce bigger voltage at higher frequency (at 150-250Hz x-over freq. cca. 500-700V, which not fully proportional, because higher freq. causes rapidly increasing of core losses at near saturated operating circumstances - must be decrease the peak core induction). The desirable output voltage will determinate the number of the series connected 230V rated windings (=number of trafos). The sum of leakage inductances and combined effective winding capacitances + panel's effective capacitance (at highest frequeny will be smaller than total panel-capacitance, but there's parallel resistive component) will be definied the upper resonance frequency and Q-factor of resonance. If the resonance frequency is too low, must choice lower leaky trafos, or/and lower winding capacitance desirable.

The implementable step-up ratio can calculate from desirable voltage rated sensibility and impedance minimum.