I have a conventional stereo set-up with an N-core class D type amplifier and electrostatic loudspeakers. The output signal of each amplifier channel (some 40 Volts max) goes into one loudspeaker and there into a transformer that transforms it up to a few thousand Volts. Then the high voltage signal goes to the stators of the electrostat. Nothing abnormal, all as can be expected.
I am wondering if it would be possible to bring the transformer out of the loudspeaker and in to the closed loop of the class D amplifier. The primary coil of the transformer would act then as the low pass filter of the class D amplifier. The loop of the amplifier would be closed by connecting the secondary coil through the / a loop-filter to the input of the amplifier.
In a momentary lapse of reason I even throught of replacing the normal transformer by a (cheap!) fly-back transformer.
More in general: does somebody knows any high voltage class D amplifier designs?
I am wondering if it would be possible to bring the transformer out of the loudspeaker and in to the closed loop of the class D amplifier. The primary coil of the transformer would act then as the low pass filter of the class D amplifier. The loop of the amplifier would be closed by connecting the secondary coil through the / a loop-filter to the input of the amplifier.
In a momentary lapse of reason I even throught of replacing the normal transformer by a (cheap!) fly-back transformer.
More in general: does somebody knows any high voltage class D amplifier designs?
I think the big elephant in the room would be to keep everything stable with the HV xformer inside the loop. That would be hard to do, unless your name is Bruno Putzeys ;-)
Jan
Jan
should be possible. You want to generate the frequency dependent two port network for your transformer to get a linear model of it (this should be fine as I guess its only operated in the linear region). You can use a network analyser for this but you might get better results using an impedance analyser (as network analyisers have poor performance at low frequency and low impedance). Once you have a model for your amplifier you can do normal gain/phase analysis in a frequency domain simulation package after converting your class-D amplifier into a frequency domain model:
E4990A Impedance Analyzer, 20 Hz to 10/20/30/50/120 MHz | Keysight
Qucs project: Quite Universal Circuit Simulator
E4990A Impedance Analyzer, 20 Hz to 10/20/30/50/120 MHz | Keysight
Qucs project: Quite Universal Circuit Simulator
Although I know zilch about class-D design, IMO, using a winding from a transformer that is designed for audio frequencies as a high-frequency inductor to filter/smooth output from class-D amp is crazy talk... Because of core losses, I would not be surprised if you just heat up the core and cause a fire. See:
Class-D Amp LC-Filter: Coil heating