New Class D amplifiers

its a compeltely different concept but the functionality of both is the same.

@orchardaudio If you ask me your amp has the greater functionality...

But the gold standard sonically remains A. Where heat/weight is not a consideration (eg., preamps) class A is the norm.

This statement is false. What is the 'gold standard' is a benign distortion product, one that allows the amplifier to be smooth, fast and detailed all at the same time. That may or may not be a class A amplifier; any amplifier is easy to mess up if feedback is poorly applied and this is often the case.

Very interesting statement; would you please elaborate on this? What kind of distortion does a class-A's output section produce that the class-D is less vulnerable to?

I cannot speak for all class D amplifiers; in our class D amp, the two sources of distortion are the encoding scheme and the use of deadtime (to prevent the output section from overheating due to finite turn-on and turn-off times of the output devices; if both are on at the same time you get 'shoot-through current' which can heat them up quite rapidly to eventually fail) in the output section. In our circuit, these non-linearities produce lower ordered harmonics rather than higher orders. Because the lower orders are benign and innocuous to the human ear, the result is a smoother sounding amp that sounds a lot like a tube amplifier- the distortion signature is really similar.

A class A amplifier, because its output section is not perfectly linear, will generate not just the lower orders but the higher orders as well. The class A operation is used to put the output device or devices in the most linear operating region, but that isn't the same as saying its actually linear and no class A output section is, so it generates distortion.

Again:

The ear is keenly sensitive to the higher orders since it uses them to sense sound pressure. This is why an amp with THD of 0.01% can still sound harsh, if the higher orders are not masked by lower ordered harmonics. The assigns tonality to all forms of distortion and higher ordered harmonics are sensed as harshness and brightness. 

going from A all the way to D strikes me as somewhat extreme for one who is used to the Pass. Why not look into a good A/B first? 

The main reason AB exists is to reduce heat and power consumption, with some of the benefit of class A operation.

Class D allows a good designer to sidestep some of the reasons for class A operation in the first place- a classic example being that of crossover distortion. Most class D amplifiers are literally incapable of this kind of distortion. In addition, the distortion of the output section isn't made in the same way that it is in a class A or AB design. The result can be that the distortion the class D amp generates can be far more benign to the ear. 

I've heard class D amps that compare more than just favorably to a Pass Labs amplifier. These days its not about class of operation, it simply about the sound.

How the measurements reflect the sound that we perceive?

I described that above.

Since the ear uses higher ordered harmonics to sense sound pressure, its keenly sensitive to them since the ear has a +120dB range. If these harmonics are not masked by the lower orders (2nd and 3rd) they will contribute to harshness and brightness, since the ear assigns tonality to all forms of distortion.

That is why you don't hear it outright as breakup in any amplifier that isn't being overloaded. You sense it as tonality instead. For example, the 'dry' quality some amplifier have is due to their distortion signature, which audiophiles call the 'sonic signature' of the amplifier.

In the interest of my own learning, would you please point out which measurements correspond to what people hear?

Sure- two measurements that have not got the attention they deserve- first, distortion vs frequency, which ideally should be a straight line or decreasing with increased frequency.

If the amp has problems with feedback due to insufficient Gain Bandwidth Product, the distortion will rise with frequency above a certain point (often between 1 and 3KHz). If this happens the amp could be perceived as bright due to how our ears perceive higher ordered harmonics and the Fletcher Munson curve.

The other measurement is the distortion spectra. Its not enough that the distortion be low. The ear has a masking principle; the way that tube amps sound nice and smooth is because they have enough 2nd and 3rd harmonic that those harmonics are able to mask higher orders. If those harmonics are insufficient, the amp may be perceived again as harsh and bright. This measurement is a bit open to interpretation as you have to know what you are looking at to be able to say how the amp is going to sound when you listen to it.

Quite often this last measurement is done with the fundamental at a low frequency - like 50 or 100Hz. If this is so and there isn't a measurement with the fundamental at 1KHz or higher then the measurement isn't all that helpful if distortion rises with frequency! IOW at a low frequency the distortion spectra might appear benign but above a certain frequency things could be considerably different.

So now you can see how an amplifier with 0.005% harmonic distortion might sound harsh and bright if those harmonics are not masked and the THD was measured at a low frequency- this can hide problems in the design. OTOH a zero feedback tube amp might have much higher distortion, but because it does not rise with frequency and the lower orders dominate, it might sound smooth and musical.

Class D offers a means to eliminate the first problem of rising distortion as well as allowing for a more benign distortion spectra.

Do I really need to explain the difference between subjective preferences and objective performance again?

There should be no difference at all! If you think there is, then you are convinced that we can't make the measurements for everything we hear (which, since about the early 1990s, we can) and also that the human perceptual rules that are shared by everyone on the planet are of no consequence, making things like deciBels irrelevant.

Of course I've found that both the objectivist and subjectivist camps don't like this heretical view. I subscribe to Daniel von Reklinghausen's take on the matter of measurement and what we hear. 

Obviously anecdote isn't reliable, but it can provide data points over time, especially if everyone's anecdote, independently of one another, describes the same thing.

Similarly, the guys that only look at measurements are notorious for their bias confirmation, and their striking ability to assemble dreadful sounding systems.

Neither side is right; both suffer bias confirmation.

I simply seeking to clarify the record by asking for any evidence of issues with Hypex or Purifi, which are much cheaper alternative.

This bit isn’t accurate. Have you priced them recently? We have. Using those modules would cause our amp to be more expensive than it is now.

But if we had it all made overseas it would be cheaper. If we didn’t have a dealer network to support the customer the same. If we didn’t care about the product being rugged enough to survive shipping abuse things would be different.

I’ve bought cheaper stuff that was supposed to be better. I hate to do it, because so often it simply doesn’t work out.

https://www.youtube.com/watch?v=SVDEE30XB5A

We care about these things- so the amps are made here in Minnesota, they have dealer support, they hold together and have a lasting finish- and they meet emissions directives for real.

FWIW in the last 40 years we’ve been less expensive than our competition operating in the same performance bracket, often by several times. In the case of our class D, its less expensive than the amps it competes against, in some cases by mulitiple times.

FWIW if you bought an amplifier and put it in storage for 10 years, it would likely need to be refurbished at the end of that time since the filter caps would be shot at the very least.

Fortunately our class D isn’t one of the ’heavy amps and flashy cases’ you are talking about...

Its very obvious you have a personal agenda that has nothing to do with amplifiers.

I haven’t heard or read of any issues with customer service from the many US and European sellers of class d amps using Hypex or Purifi modules. Do you have actual evidence of poor customer service you can share from any of the main stream amp providers?

This is a strawman; I didn't mention anything about the brands above.

I didn’t bring up your product here, you entered the discussion after I commented that there is no need to spend $5000 on a class d amp when there are better performing amps available for a fraction of that. I didn’t mention your amp by name. You jumped in and offered a justification for the high price.

And no, I have not singled you out, I made a general comment. I would respond the same way to any product here that has a performance not scaled to the price. I appreciate substance over form.

Calling BS. If you are going to take that approach, you need to up your game.

FWIW we did not make our class D with an idea of making it to a certain price point. Instead we wanted good quality in all the parts with good reliability and priced to the same formula we've used for the last 40 years.

As I said, there is no reason to spend $3000, $4000, $5000, or more on a class d amp to get state of the art performance. When you spend that sort of money, you are paying for cosmetics rather than performance. If that’s what is important to you, have at it.

So this is what you think Technics is up to?

I have chosen to troll this thread? No, I made a general statement that you decided to challenge. I simply responded.

But, ok, I have failed to identify the market you are attempting to serve. Let’s say then it must be that segment of the market where high prices are a feature, not a bug, and where "value" isn’t a consideration. That’s fine and honest. No need to justify your price here, I get it.

Again, calling BS. And FWIW, still don't seem to understand the market. Your strawman here demonstrates that obviously enough.

Is it really a binary choice? What I am saying is making an amp easy to service and update should be part of it’s design and doesn’t need to be an expensive operation to perform. How hard is it to remove a few screws, unattach a few wires, and swap a module?

@kuribo 

😀 Apparently you've not been on the service side of the industry to see how things are done. I agree, what you say above is true, but sadly often is not the case in practice. Customers are also wanting to know that someone stands behind their purchase, and that sadly too is lacking with a lot of inexpensive options.

Huh? I don’t recall saying an amp doesn’t need a power supply- I simply mentioned that linear supplies in this day and age are rather archaic. I specifically said that a module swap was all most class d amps need to be repaired or upgrades, nothing innovative there.

Nor did I say that you had... I was just pointing out that all amps need a supply- that's not going to change. WRT SMPSs, we found that if you really want to get the most out of any module, the supply should be designed for the application. When you do that, all of a sudden the use of a conventional supply is easier and less expensive.

Again, is it really a binary choice?

No, nor did I say it was. What you don't seem to get, despite this being told to you a number of times when you've chosen to troll threads like this, is you've failed to identify the market. For example, you don't seem to understand who our competition is. From the way you consistently single us out, its pretty clear that this has nothing to do with price since there are so many more class D products that are more expensive.  For some reason, this seems personal for you for reasons unclear to me.

Does anyone actually think anyone will be using the same class d amp in 20-30 years? With the way the technology is still evolving, I find that bordering on pure fantasy. I see no reason why having something that is easily repaired (what can be easier than swapping out a module?) or updated should add significantly to the price.

@kuribo The flip side of what you're saying is that no care should be taken to make the amp easy to service and update??

Since you are not a manufacturer you won't be familiar with the idea that a customer might contact you looking for the 'last amplifier I will ever have to buy'. I've experienced that many times and I'm sure other manufacturers have too. Believe it or not, a certain portion of the market appreciates a quality build that doesn't look like a denizen only suitable in a mancave. Also, we have found that the ability to update and reactivate the warranty of any of our older products has the effect of increasing their resale value on the used market so it natural to have that same ethos apply to any newer products we make.

Amplifiers might change, but they will always need a power supply. So being able to update the amplifier with an improved module might not make sense to you, but it makes plenty of sense to our customers. 

The advantage that GaNFETs have is they have very low parasitic inductance. What this means is its easier to build a module that is low noise- and by noise, I mean the kind that radiates from the module over the air and thru the AC power. Its important to keep switching noise and parasitic noise low because it can mess with other gear and affect its performance. As an example, our class D amp makes less noise on the AC line than most tube amps. In terms of noise that it makes in the loudspeaker itself, 90% of that comes from the input buffer circuit, but even then is obviously quieter than tube amplifiers (I use horns at home FWIW).

We've been docked a bit on account of the cost of our class D (which is cheaper than most of our tube products). We built it to be rugged (so it will ship without damage), to last decades, to be easily repaired and updated and to that end it has a fairly expensive chassis and power supply. It would actually be more expensive if we used someone else's module.