Has anyone switched from Class AB/A to Class D? Was it better? Was it worse?

@westcoastaudiophile You might try the AGD Audion, one of the Orchard Audio amps or one of ours.

WRT to the analog filter, its job is to remove the switching artifacts that occur at the switching frequency. What is left will be a sine wave called the "residual". This sine wave is at the switching frequency so might be 500KHz. Its amplitude is so low that its of no concern to any loudspeaker driver, whose inductance will block it.

The filter is designed according to normal filter theory; typically second order. Usually its designed with a target load in mind, which might be 4 or 8 Ohms. Increasing the load (speaker) impedance results in a higher Q (Quality) of the filter effect. The load will shift the frequency a bit but usually its well outside of the audio band so that’s not a problem.

The characteristics of the capacitors and inductors involved is critical, since stray inductance (which might be caused by the leads of the capacitors, which extend not just from the circuit board but well into the part itself) can cause parasitics as well as the design of the choke.

It is for this reason its usually not a good idea to replace the caps or inductor in a class D amp in an effort to get better sound since reducing RFI is part of the filter’s task. Parasitic oscillation may be introduced! We found we had to be quite careful about the inductor in our amp, which uses a ferrite core. It was so critical we had to build our own jig to wind the chokes so we could control the inductance without strays. The vendors we tried to have the part custom wound simply didn’t seem to have the ability to get it right. To make the choke work properly we use a Litz wire. 

Of course the filter introduces phase shift as any filter. However, in our case the amplifier is using quite a lot of feedback and so is able to correct phase at high frequencies.

You may dial 500mW output @ 1kHz sine on any D class and A class amp and compare waveforms to understand D class problems better!

@westcoastaudiophile Actually there are a few class D amps that have a better first Watt than a good number of class A amps. One thing about class D is for the most part (as long as they have an analog output filter) they are immune to crossover distortion. So there can be a complete lack of higher ordered harmonics at very low power levels.

That looks good on paper and sounds good too.

Put another way if you want to show what might be wrong about a class D amp compared to a class A, this particular metric isn't it.

 But tubes are not something I hear and can easily go to a class D system unless heat or space is the consideration.  A preference for tubes is in a class by itself.  The differences are easily distinguishable in a blind test.  SS v tubes, if you enjoy tubes; my advice is don’t even bother, over time you will be sorry, have spent more money, just to go back to the sound you like, tubes.

@daytrader I like the sound of tubes a lot as you might surmise. I like the sound of the class D amp in my home system better. Its just as smooth in the mids and highs and seems to delineate information in the rear of the soundstage with more clarity than the best tube amps I’ve heard. I don’t miss having to replace tubes or the heat which is a nice side benefit.

1. I don’t think we really know how long class D amps will last, how they will most likely fail, or whether they can be maintained in the same sense as a conventional amp.  They could last longer.

2. Class D amps use far more complex designs.  Many of the amplifier boards are built as modules and used by multiple amp makers.  That’s why a class D amp will often be identified as a Nilai, NCORE, Purifi, PASCAL, etc.  Some (all?) of these companies also build and sell compatible power supply boards.  Add a case, power switch, input buffer. wires & connectors, and you’re done!  I don’t think all class D amps are built this way, though.

3. The more recent introduction of GAN FETs has made for another family of class D amps.  I don’t know if these are also mostly being based on modular sub-assemblies; what makes them distinct is the use of a specific kind of transistor.

None of these are necessarily negatives, just things that occurred to me recently.  Hopefully, @atmasphere will drop by and fill in the blanks (please?)

Class D amps will last as long as any conventional solid state amp; perhaps longer since the filter caps will have less heat around them. 

Class D amps are not particularly complex except in the case of the formula to determine values in the feedback network of class D amps that are of the variety known as ’self oscillating’. In terms of components used our class D amp is simpler than most class AB solid state amps.

Not everyone uses modules from other companies- we certainly don’t; ours is designed and built by us as are power supplies we use. What we found prototyping our amp is that the power supply is critical and for best results should be designed for the application rather than bought off the shelf. That so many people do the mix and match thing is IMO why class D has had trouble getting traction in the last 20 years. 

If you know what you are doing you can build a class D amp that is in the driver’s seat rather than the back seat WRT any other kind of amplifier technology.