Is it analog or is it digital?

@dynamiclinearity 

It's unfortunate that class D amps are called class D since in audio D often stands for digital and the description of how a class D amp works(square waves) sort of sounds digital. Plus class D involves encoding albeit an analog encoding not a digital one.

On top of that, they got marketed as digital amps, and some of them take a digital input and don't require a line level d/a converter. They somehow convert PCM directly into PWM to switch the amplifier on and off, so the analog signal never emerges until the PWM signal goes through the capacitor on the amp's output. They can bypass a pre-amp too by adjusting the rail voltage on the class D amplifier to control the volume. I had a Panasonic PurePath digital receiver in the early 2000s that touted all these features. It was zero negative feedback too!

Still I agree with you, the class D topology is not working in the digital domain, most clearly when being fed an analog signal that it is converting to pulse widths that are not time bound by any clock rate. With the digital input it gets a little more fuzzy for me. In that case it seems like the pulse widths are tied to some clock rate. That seems like a disadvantage, and from the measurements I've seen, class D amps set up to take an analog line level signal perform better than the ones set up to convert the PCM signal directly to PWM.

On top of that, they got marketed as digital amps, and some of them take a digital input and don't require a line level d/a converter. They somehow convert PCM directly into PWM to switch the amplifier on and off, so the analog signal never emerges until the PWM signal goes through the capacitor on the amp's output. They can bypass a pre-amp too by adjusting the rail voltage on the class D amplifier to control the volume. I had a Panasonic PurePath digital receiver in the early 2000s that touted all these features. It was zero negative feedback too!

Still I agree with you, the class D topology is not working in the digital domain, most clearly when being fed an analog signal that it is converting to pulse widths that are not time bound by any clock rate. With the digital input it gets a little more fuzzy for me. In that case it seems like the pulse widths are tied to some clock rate. That seems like a disadvantage, and from the measurements I've seen, class D amps set up to take an analog line level signal perform better than the ones set up to convert the PCM signal directly to PWM.

Yep, so the lines on class D's "digital vs. analog-ness" get a little blurred upon closer inspection. For the purpose of THIS topic, it should be treated like a traditional class AB amp. Use a high current outlet or plug directly into the wall - do NOT plug it into a power conditioner's socket intended for low-power digital components.

Class D amplifiers have to employ HF filtering on their outputs, for a smooth response - this is at least vaguely analogous to a DAC's filtering in its conversion.

mulveling.  thanks for the clear, understandable explaination.  makes perfect sense and best way to think about it too.  

The difference between Digital or Analog on most conditioners usually has nothing to do with amplifyers.  Having said that, it’s usually better to plug amps straight into the wall on a different circuit than your power conditioner. That way the amps get all the power they need without taking away from the more sensitive bits like streamers.