Class D step-up possible?

[Conrad Hoffman]

I want to step up the output voltage of a class D amp. Did some searches and didn't like the answers that said the output transformer needs to pass the lowest audio frequency. Can somebody explain this in more detail? If the output transformer passes a suitable HF PWM signal it seems possible. No doubt I'm missing something- never worked with class D. I thought isolation amps that want down to DC used a method like this. Been wrong before though.

Thanks,
CH

 

[nigelwright7557]

You would need to take the output after the class d inductor so would need a normal audio transformer. Of course stepping up the voltage also steps down the current so it wont be any louder.

 

[Conrad Hoffman]

Why can't I step up the PWM signal, then filter it after the transformer? This isn't for a dynamic speaker application. Think more like electrostatic, but not quite so high a voltage. I want to get up to a kV or so without having to start with a HV supply or switch that much voltage.

Thanks,
Conrad

 

[Pafi]

PWM signal contains audio freq components also. This is the reason why it can be restored by a simple filtering.

 

[theAnonymous1]

I'll admit I'm a complete idiot with severe comprehension problems, but I feel like taking a guess at this one.

Is it that the inductor would have to be large enough to pass the lowest audio frequencies, which in turns means the inductance would be unsuitably high?

 

[Eva]

The PWM signal contains all the audio frequencies plus the carrier HF. Feeding the unfiltered PWM signal to a big transformer is not a good idea due to possible EMI (leakage inductance would do the output inductor job).

The transformer needs a big enough core and a high enough amount of primary turns to prevent saturation at low frequencies. Exactly as every other audio step up/down transformer.

 

[nigelwright7557]

One way to do it is to take the pwm out and feed it into more power transistors fed with a higher voltage.
This might need an interface because of the change in voltage.

 

[Conrad Hoffman]

Well, this just doesn't sound promising. Back to the drawing board- maybe just go where I want to go directly with a tube amp.

 

[Eva]

Driving electrostatic panels?

 

[Conrad Hoffman]

No, non-audio instrumentation application, but very similar to electrostatic drive.

CH

 

[phase_accurate]

I know someone who used multiple floating class-d amps connected in series in order to achieve voltages up to 1000 Volts (to drive a piezo actuator) with success. The disadvantage is the complexity that already starts at the PSU level (multiple foting supply voltages).

Regards

Charles