When did class D start to not suck?

[sngreen]

Class D = an amplifier that switches the output devices really fast between positive and negative rail. The analog signal, a bit like DSD in that sense, is obtained using an LC low pass filter. Very efficient in power.

Class G = non switching amplifier (usually class A or AB) with several (more than 1) different power supply rails. The sole purpose of this technique is increasing efficiency, heat. Carver PA amps worked that way decades ago already. Small rack sized amps with huge power and little cooling needed.

Class H is well.. kind-off a class G variant. The difference with class G is that while class G has more than 1 power rails (so in 'steps') the class-H amp has a (switching) power supply that creates an internal voltage rail (actually 2, 1 positive and 1 negative) that does not switch (like class G) nor is constant (A, AB) but is modulated (when voltages above a certain voltage) are needed and thus always is just a few volt higher than the required output voltage.
The advantage is even higher efficiency than class-G. Almost on par with class-D.
Normal class AB (or even class A) amps can be created with these techniques while wasting of heat is reduced (higher efficiency)

class G and H are not related to class-D but the rail voltage modulation is switching so class-D-ish.

Would you name a few brands that do use Glass G technique? I see only Arcam when google for it, but I am sure there are more. Thanks.

 

[solderdude]

That would probably be mostly used in older PA amp designs where lots of power needs to be available in a small rack with relatively small weight.
Nowadays this will all be class-D which has an even higher efficiency, is more light-weight and can be made small.

There might well be some hifi amps that used this technique.

A subset of this was NAD power envelope. This allowed for high voltage peaks but only for short moments. Great for peaks in music.

 

[DonH56]

Would you name a few brands that do use Glass G technique? I see only Arcam when google for it, but I am sure there are more. Thanks.

Benchmark and Emotiva come to mind. I imagine there are others. Both mention class H but operate as class G per their technical docs and engineers' comments. I read recently that class H is defined differently by European standards; I use the USA definitions so class G switches the power rails and class H power rails track the signal.

 

[voodooless]

Gene is late to the party:

 

[syn08]

That's a whopping 0.07 mm difference. Moving your head has more influence, or rather: trying to keep your head still has more influence

Maybe they used a head jig for the listening sessions. I think I’ve seen one in a museum of Medieval History.

 

[sngreen]

Gene is late to the party:

I take Pass Labs any time over NAD, regardless of the presented measurements.

Played M23 over the weekend, promptly returned. Really loved the others, though.

Musical Fidelity M8 500s sounds far nicer to my ears.

And FirtsWatt (50W Class-A) even more so. One of the most pleasant amps to listen to.

 

[dadregga]

I take Pass Labs any time over NAD, regardless of the presented measurements.

Played M23 over the weekend, promptly returned. Really loved the others, though.

View attachment 231002

Musical Fidelity M8 500s sounds far nicer to my ears.

View attachment 231003

And FirtsWatt (50W Class-A) even more so. One of the most pleasant amps to listen to.

It's fine to prefer amps that aren't provably entirely neutral, it's just a rather expensive way to apply EQ

 

[antcollinet]

Gene is late to the party:

Better late than never.

And is promoting an *absolutely* correct message.

(Also - from reading another of his articles - correctly believes that the term "wife acceptance factor" should have been left in a previous century - where it didn't belong even then)

 

[Doodski]

Better late than never.

And is promoting an *absolutely* correct message.

(Also - from reading another of his articles - correctly believes that the term "wife acceptance factor" should have been left in a previous century - where it didn't belong even then)

I've been buying SMPS PWM power supplies for my desktop PC for a long time and no issues at all after the first cheap one burst a few caps and went thermally intermittent. I also repaired thousands of high end SMPS car amps and most if not near all of the faults where customer use and abuse conditions. So his being snarky with SMPS stuff is not warranted. The NAD does measure well.

 

[antcollinet]

I've been buying SMPS PWM power supplies for my desktop PC for a long time and no issues at all after the first cheap one burst a few caps and went thermally intermittent. I also repaired thousands of high end SMPS car amps and most if not near all of the faults where customer use and abuse conditions. So his being snarky with SMPS stuff is not warranted. The NAD does measure well.

True - I'd forgotten that bit.

 

[Vacceo]

Gene posted a video on Audioholics about class D amps just some hours ago. If you wanted a second opionion based on measured performance beyond what Amir has been doing for years, there you have it.

 

[sngreen]

It's fine to prefer amps that aren't provably entirely neutral, it's just a rather expensive way to apply EQ

Actually the voices through Pass sound the most natural and realistic to me. The thing with NAD is that it gives very short decay on the notes - maybe that is why it measures so clean - which leaves the sound stage plane flat and almost without any kind of layering behind it. Good for electronic music, but not so for jazz and forget the big arrangements classical. It's just not how the instruments sound.

 

[sngreen]

Gene posted a video on Audioholics about class D amps just some hours ago. If you wanted a second opionion based on measured performance beyond what Amir has been doing for years, there you have it.

Unless the amplifiers are designed strictly for measurements, not for the listening, I would leave the word performance out of it.

 

[sngreen]

Better late than never.

And is promoting an *absolutely* correct message.

(Also - from reading another of his articles - correctly believes that the term "wife acceptance factor" should have been left in a previous century - where it didn't belong even then)

There are more cars on the streets than there are high performance amplifiers in people's houses. And every one of them has a radio. So yes, the message is *absolutely* correct, the future is now.

 

[Urvile]

Just put a deposit on a Buckeye amp, (Purifi 1ET400A Amplifier, 3-channel) and I still don't have a pre-amp to drive it!

 

[Phorize]

Class D never stopped sucking.

LMAO! Obviously you don't need me to point out the folly of your ways though

 

[Phorize]

And what is your opinion on the unnecessarily heated discussion over semantics?

I think we need to strive for a design that runs cooler.

 

[Vacceo]

Actually the voices through Pass sound the most natural and realistic to me. The thing with NAD is that it gives very short decay on the notes - maybe that is why it measures so clean - which leaves the sound stage plane flat and almost without any kind of layering behind it. Good for electronic music, but not so for jazz and forget the big arrangements classical. It's just not how the instruments sound.

You will never get the sound of the instruments because all you play is a record, not the instrument.

For distortion of the source, there is EQ

 

[sngreen]

You will never get the sound of the instruments because all you play is a record, not the instrument.

For distortion of the source, there is EQ

Not debating it. Except that the first sounds realistic and the other digital. Whichever way it is achieved.

Flat and compressed sound stage (or maybe it was just not burnt in) is not the type of the presentations I am after. Neither would I recommend it.

 

[Vacceo]

Blame the record, not the reproduction gear. Digital has no sound, we cannot listen to digital data, we listen to mechanical waves.