Class D amplifiers. What's the future look like?

As to the future, is it just a matter of time until all module makers switch to using ganfet?

I don't think so. In the decade since GaNFETs first showed up, MOSFET producers have improved their craft and are almost as fast- certainly so for any class D circuit. In the meantime, there are other semiconductor types that are already in production and on the horizon.

IBM was also the source of the parity bit...

So my thought is that calling them digital amps is a reasonable slang and everybody knows what it is referring to.

The distinction here is that digital has specific on and off states that form words (which might be 16, 32 or 64 bits) that represent a value of a voltage. While class D is a switching technology, there are no 'words'; the on and off states are an analog of the input voltage so is considered an analog technology- Pulse Width Modulation goes back a really long way back to the tube era.

@erik_squires Sure- that would work :)

Class D is switching, but there’s no analog to digital/digital to analog process involved. It’s all done with feedback. :)

@erik_squires You can build a class D that is zero feedback as well.

Given the filters and phase shift, does class d design lend itself to mathematical modeling?

Yes. Like anything else though it depends on the quality of your models.

As well as other new applications of technology such as Westminster Labs sliding bias voltage Class A cool running higher power solid state amps?   They may well function superior to Class D amps.  

The issue has been for the last 70 years and remains to this day Gain Bandwidth Product; if you plan to run feedback you need a lot of it. Easy to do with class D, very hard with conventional A or AB circuits.

A crappy or over hyped review shouldn’t change anyone’s desire to audition or dump a component. A true test of a component is how it sounds (gels) within your system and to your ears.

@lalitk You got that right. If the manufacturer’s, distributor’s, dealer’s or reviewer’s lips are moving, they are lying- we’ve lived with that fact for so many decades, most of us grew up with it. So the only way to know what works, even for the jaundiced ’measurement only’ guys is to take the device home and play it in their system.

I think he might need to work on his business model 😁

@soix FWIW Eric uses a set of DeVore Fidelity speakers in his system; the same that he used for the M-60 review. IIRC the o/96s.

when i just look at his D with the chassis open, it ain’t looking like much to me.

Class D, done properly, is never going to look like much. Its very different from a tube amp in that regard!! The biggest issue class D has to contend with is noise generation; to that end the module has to be as small as possible to minimize parasitics caused by stray inductance (and at the frequencies involved, a lead of a part easily causes problems...). Because we are using GaNFETs, which don't require large heatsinks, the heatsinks (which are machined to 1 mil tolerances) are beneath the module and serve also as its mounting. So ya, not looking like much 😉

Speaking of bass and guitar players, with tubes you can tune and fine tune the sound by tube rolling. And with SS, any class ?

@inna Many guitar and bass players rely on their effects 'pedals' to sculpt their sound. So a lot of the time they just need an amp that sounds clean and is easy to play loud. Its less common to rely on the sound of the amp when it overdriven (clipping); those that do that though are more likely to have a tube amp. Just saying that since that latter practice is becoming less common (because effects pedals provide the distortion/fuzz instead), tubes will be on the wane in the musical instrument world. It makes a big difference at 3:AM if you have to move equipment out of a gig and the amp weighs 85 pounds or 15 pounds...

I wish my ears were more forgiving. I think I could make some Class D work for my ears with tube line stages and dacs, but, i really don't want to deal with the confoundedness/pestilence/mainteance crap with tubes anymore...

@deep_333  @inna I get the whole thing about dealing with tubes. Imagine what that looks like when you are a provider of tube equipment! Now comes the tricky bit: imagine that as a manufacturer, if you produced a class D amp after 45 years of class A tube products, that if you make a substandard amp you will damage your market because people will think you've gone nuts, can't hear what you're doing, cashed in, stuff like that. Add to that the simple fact that every OTL manufacturer that ever moved from their OTL mainstays to solid state promptly went out of business.

So you can see we didn't go into this blindly. We knew the class D R&D would be a bust if it didn't sound every bit as good or better than our OTLs (which, in the tube world, rule the roost when it comes to transparency and bandwidth...).

As I've pointed out on numerous occasions, the sound of class D amps varies from worst to best over a wider range than tube amps and for that matter conventional solid state. What this means in plain English is if you've heard one, there's no way you've heard them all. Some are grossly incompetent IMO, some are soso, some are excellent. Most are somewhere in between.

So I've no surprise when someone says class D isn't for them. The truth of the matter is if they heard the right amp they likely wouldn't say that, but they have no way of knowing.

No-one will remember this class D in twenty years, and no- one really needs it except a few fashionistas. But we do need a completely new amplification device to be invented, both tubes and transistors virtually exhausted their potential. This is one more commercial BS, this class D, similar to another cd reissue, slightly better or slightly worse than the previous one.

Inna, I think you are mistaken about this. Class D has a longer history then it appears you know of from this post. It dates back to the 1950s and the first home applications were sold in the 1960s. Sony and Yamaha made class D amps in the 1980s. Like any technology its evolved. No-one takes germanium transistors seriously in this age but they were de rigor for solid state amps in the 1960s; only the very advanced designs used silicon. These days silicon has gone the way in favor of MOSFETs and IGBTs.

The big change that seemed to really push class D amps was the advent of self-oscillating circuits, which appeared about 20-25 years ago. This allowed for more reliable, stable and simple designs. 

Tubes have not seen much evolution in the last 40 years by comparison. The big advances in the tube world have come from improved materials for construction (better capacitors and resistors for example) and not to put too much of a point on it, but innovations from designers. I'm one of those designers having several patents in the OTL world; another is David Berning with his excellent zero hysteresis radio frequency coupling system (usually called 'ZOTL' a meaningless acronym created by Harvey Rosenburg; they are not OTLs...) and perhaps Jack Elliano, who to my knowledge is the only one to push SETs further, with his patented class A3 technology and Ultra Path design.

I realize I'm not going to convince you of anything- listening is the only way to do that!  But there is something you'll want to know: Over time, class D has been making inroads into the musical instrument market (bass and guitar amps). The first ones I saw, as in high end audio, were hard to take seriously. But that's changed. Guitar and bass players alike appreciate the reduction of weight while at the same time being able to get their 'sound' about which they are surprisingly picky- not unlike audiophiles. As class D invades that market, tubes will get pushed out. The musical instrument market is the bread and butter of tube producers. It appears that market will look very different in 10 years. So class D is something that has to be confronted and figured out. Whether that's already been done is a different story 😉  

@deep_333 I've seen that list of quotes before. They got debunked due to their age. 

A lot of those quotes you listed are from nearly 20 years ago. At that time I was of the same opinion as seen in them. One way you can date the quotes: look at Thorsten Loesch's comment about a 300KHz switching speed. No-one has made class D amps with that low a switching speed in a very long time 😁

So these comments can be discounted as simply out of date.

Your comment about GaNFETs seems a bit uninformed to me. Its not that GaNFETs are somehow 'the answer' so much as when they started turning up was also about the time that class D got a lot better figured out (and that was ten years ago...).

There's long been a tubes vs solid state debate on the internet, older that the internet itself. All technologies improve in time so one can safely conclude that sooner or later solid state would get good enough that tubes would simply be eclipsed. To tube aficionados like myself this also means that that new technology will eclipse solid state A and AB amplifiers as well.

I cannot speak for other manufacturers, only myself and that should be taken with a grain of salt since I am associated with a manufacturer. I've made no secret that I replaced my triode class A OTL amplifiers with a set of class D amps about two years ago and I don't hear any tradeoffs whatsoever. The class D amps are every bit as good and better in some ways then the OTLs. FWIW, the OTLs have been getting rave reviews and awards in the high end press since sometime in the 1990s.

As a result I'm of the opinion that class D is something to be reckoned with and isn't at all as you described; it dominates what we in the high end audio world call 'mid fi'- stuff you get at Best Buy and the like. Its been making inroads into high end the last 20 years and at this point, seems evolved enough that any manufacturer of amplifiers will be going out on a limb if they don't get class D figured out. Its that simple.

Why is there this "push" to Class-D in the first place? What do you gain? In what way would they ever be sonically better than a good Class A or AB amp?

@moonwatcher Class D offers something that is very hard to achieve with A and AB amplifiers: a very high value of Gain Bandwidth Product. Most solid state amps use feedback, but as you might know, feedback has gotten a bad reputation in high end audio, not because it doesn’t work, but because it usually gets poorly applied- and so causes distortion of its own, adding many higher ordered harmonics (to which the ear is keenly sensitive and interprets as harshness and brightness).

So the answer is pretty technical. One reason feedback has this bad rap is because when the GBP limit is reached, feedback decreases on a 6dB slope and perhaps faster with succeeding octaves. This can and does happen as low as 1KHz, so distortion will begin to rise- putting higher ordered harmonics in the most sensitive region of human hearing. What I’m talking about here is easily measured: distortion vs frequency.

When you have enough GBP, distortion vs frequency can be a ruler flat line across the audio band. This allows the amplifier to be smoother in its presentation, with greater detail at higher frequencies (since distortion obscures detail) without harshness. Smoother and more detailed at the same time is a good thing IME.

Class D makes an enormous value of GBP available to the designer. So you can run very high amounts of feedback without getting into trouble; we’re running 10x more than conventional A or AB amplifiers. This makes possible an amp that is smoother and more detailed than conventional A or AB designs (even in tube embodiments), eliminating the benefit that class A used to offer.

The reduced heat, reduced size, weight and cost are all nice side benefits.

They will go the way of cd players, in ten years no one will need them. Ralph and others perhaps have some time to cash in. He does not prefer his class D to his and many others tube amps. He won't fool me.

@inna I don't think if I was trying to fool people that I would last very long- things like that have a way of being found out. There's a very simple way to tell if I've been straight up 😉