Is D for Dry? Class D...

Ralph, don’t skew things around "to seem" to suit your next product.
I did not say "won’t handle", I said will not get "the best" out of the Alexia’s bass because of it loading.
We have the Alexia’s here in front of us and know what they like to get the best out of them in the bass, and it’s not Class-D yet. (Maybe in the future when they can come close to doubling from 8ohm to 4ohm to 2ohm wattages like big better linear amps can)

Again, according to Wilson themselves, the speaker does not suffer any performance loss strictly on account of class D. Some class D amps are not as good as others just as some class AB amps aren't as good as others.


And for the record I'm not skewing things. To be clear, if you want to get '  ' 'the best' out of the Alexia's bass', a class D amp can do that as well or better than a regular class AB amp. Here are the provisos:IF-
1) as long as it is designed with the same current in the power supply as the AB amp2) as long as it has adequate heatsinks3) as long as the output section as the same current ratings in its output devices-then it will make bass no worries, with 'the best'.

This is nothing about our upcoming product so much as its about Ohm's Law.


I was a bit astonished that in the face of advice from the speaker manufacturer in question that you were continuing to hang on to your position, which they regard as nonsense.


Now it may be that you are actually comparing some class AB amp against a particular class D amp on that speaker- and I'll totally give you that anecdote. But anecdote is not proof of your position (its a data point at best); if that is what you are experiencing then I advise you look for a different (probably newer) class D of more competent design. If you need tips on what to look for here are a few:1) avoid amps with SMPSs as most SMPSs will current limit if they are off the shelf units. Get one with a more traditional power supply.2) look for units that, if using feedback, are of the self-oscillating variety. The Hypex is a good example.2) The input circuit, if installed by the ultimate builder of the product, should use high quality opamps operating at low gain. Even though I like tubes, they are very difficult to use properly as the input circuit because the comparitor input impedance is so low. This will make it hard for the resulting amp to have good low frequency response.

From the Wilson website page for the Alexia 2:

Measurements Sensitivity 89 dB @ 1 watt @ 1 meter @ 1 kHz Nominal Impedance 4 ohms / minimum 2.54 ohms @ 85 Hz Minimum Amplifier Power 20 watts per channel Frequency Response 19 Hz – 32 kHz +/- 3 dB Room Average Response [RAR]

We can see that Wilson does not think this speaker hard to drive with a 20 watt minimum. Because the AGD has such a low output impedance I think it should be able to drive this speaker quite well, depending on the power of the amplifier and the size of the room. If you have a larger room all you would need is more power.


So I guess I’m refuting the idea that a class D can’t handle this speaker (any more or less than any other class AB amp of the same power might). I really don’t see what the problem is.

I spoke to Jerron that works at Wilson (for the last 20 years) and he is in complete agreement. He says the older Alexia was a bit harder to drive but the class D amps handled it no worries at all.
So I'm going with a 'furphy' on this one.

George goes on about this EPDR thing. But he's not clear on what the impact will be on a class D amplifier other than saying that they won't drive it.

Almost any class D amp made today has an output impedance that is often a good order of magnitude lower than traditional class AB designs. This is because the On resistance of the output devices is (especially if GaNFETs are used) literally only a few milliohms (our amp has an output impedance of 10 milliohms as an example). When you have an output impedance that low, if a speaker has an effective EPDR that causes it to be effectively slight less than 1 ohm the class D amp will still behave as a voltage source while driving it. This is for the simple fact that at 1 ohm the class D amp may well have a damping factor of over 100.


That's not the same thing as the amp falling flat on its face! However there is a chance that the class D amp will go into current limiting mode to protect the output devices. GaNFETs are tricky to heatsink and as a result the heatsinks you have to use are unlikely to move heat away from the output devices fast enough to prevent failure when driving such a load. So the prudent designer will have some sort of protection circuit even though the output devices can easily handle the current.

As I've pointed out a good number of times in the past, the limits of class D amps to drive a low impedance are the same as they are for a conventional class AB amp: current rating in the output devices, current available from the power supply and heatsinks enough to move heat away from the output devices.


Ironically its the GaNFETs that George has been so enamored of in the last couple of years where this heat problem is most likely to occur; conventional MOSFETs are often packaged in a TO220, TO240 or other package that can be easily heatsinked.


Now we should be clear about one other issue regarding EPDR. In most speakers the region where this might be a problem is often just a very narrow range of frequencies, IOW not the entire impedance curve of the speaker! If this is the case with a speaker that has such a low EPDR rating, it won't be much of an issue for a class D amp because there just isn't enough energy needed  over this narrow range to heat up the output devices. The Wilson Sasha and Sophias are good examples of speakers of this, and in practice mean almost nothing to any class D amp. At most clipping onset might occur slightly quicker, but frequency response is unaffected.

George accuses me of shilling nearly every time I explain anything about class D, so I reiterate that what I've said in this post has to do with most class D amps that are available today.  

When are your Class D amps coming out? I just sold an amp and looking for a new one.

We're doing limited shipping of Beta production right now. As with many things associated with the pandemic, we've run into supply side shortages that have caused us to design and fabricate some parts that should be 'off the shelf'.

You can argue about science but it is just thoughts in your head. What is real, is what we hear.

Expectation bias is also very real. Perhaps have doing the destatic thing, measure what it does? You'd need a CD with test signals on it.

I don’t see much point in interacting in this kind of debate if the mods are going to delete/ban people for having a different opinion.

@dougeyjones  Its not the difference of opinion that got your posts removed, it was that you violated a forum rule. As an example, you can't call someone 'stupid' or anything like that. You **can** call the comment the person made stupid. The distinction is 'attack the post, not the poster'. If you follow that rule you won't get posts removed.


I've auditioned a number of mics against my Neumann U67s and find that they are actually quite neutral. Just because something is tube does not mean that it has to be colored- so much depends on topology. For example, to get rid of the 2nd order to which so many solid state advocates object, we designed our tube amps to be fully differential from input to output. In this manner the even orders cancel in the amp, not just at the output. By doing this our primary distortion product is the 3rd harmonic. It masks the presence of the higher orders and so the amps sound very smooth, and because the even orders are mostly cancelled, also very transparent. This is BTW how a lot of solid state designers do it also!


You can make a solid state amp that makes a lot of the 2nd order, and it too will sound smoother; Sunn very famously did that with their first solid state guitar amps in the late 1960s, by making almost the entire amplifier single-ended (which is where most of the 2nd order comes from). Its more about topology than it is tube or solid state. Our class D amps have a distortion signature that looks a lot like a tube amp (as does the AGD Audion) and no surprise, these amps sound nice and smooth. 


@tweak1  To be clear, the moderators are not 'PC' so much as someone reported a post you made- that's how moderators find out about forum violations. I'm a moderator on a different site if you're wondering how I know this.

There’s a reason basically no music is mastered using tube equipment, there’d be no way to ensure a consistent experience across platforms.

This statement is false. Pick up a copy of Tape Op magazine. Tube mics, tube mic preamps, tube compressors and as pointed out, some are using tube amps for LP mastering.

I think there is significance in Alberto of AGD being Italian. He surely has special sensibilities regarding a highly attuned musicality.

From what I've seen, its more about how little distortion the amp makes.

You can’t cheap out on your amplifier. You just can’t do it. Class D is an audio company attempting to sell product that’s nearly all profit. Class A. Now that’s an amplifier. If it doesn’t heat you out of your room it’s not worth your time.

Although we've been building class A amplifiers for the last 46 years, this statement is clearly false. Its worth your time if it sounds right, not on how hot it does or does not get.

Class D, good for powered subs and car stereo amps, and that’s about it. All my opinion.

I agree that a good number of older class D amps were not all that musical. I heard some early examples that I would have thought were a joke were it not for the price.


Class A relies on the bias setting and operating point of the output devices such that you get the maximum linearity out of them. This is important to keep distortion down and for getting a nice first watt.


The advantage of a class D amplifier is that controlling distortion isn't based on the output section. This is a tremendous advantage. At this point its more about how linear the encoding system is.


In a class A circuit, the linearity is important because in most cases, the amount of feedback that can be applied to the circuit is often limited by frequency poles (like coupling capacitors; in engineering terms, this idea is called 'phase margin'), which if exceeded, will cause the amplifier to oscillate. The other problem is that the application of feedback causes distortion which manifests as IMD and higher ordered harmonics through a process called 'bifurcation'. The ear interprets this as harshness and brightness, as it assigns a tonality to all forms of distortion. The ear is also particularly sensitive to the higher ordered harmonics as it uses them to sense sound pressure.


These aspects of using feedback have been known for a long time- Norman Crowhurst wrote about them in the late 1950s. Now you can get around the distortion problem by running enough feedback, but in most class A amplifiers this isn't practical since the amp will probably go into oscillation; you need in excess of 35dB for the amp to clean up the distortion caused by feedback.

Class D amps have a means to get around this problem. You can run 35dB or more of feedback with the expectation that the amp will oscillate- in fact its encouraged. The oscillation is used as the switching frequency, resulting in a fairly simple circuit that has low distortion and lacking the higher ordered harmonics that causes harshness in so many amps.

Class D sounds dry and lifeless... thats all, carry on

The one I'm playing sounds neutral, smooth and relaxed, very transparent without any dryness at all.


Class D amps vary in sound in much the same way that tube amps do (SET vs push-pull vs OTL and so on...). To assume they all sound the same is really ridiculous. But to be clear, some (especially older designs) do sound dry; it was years before I could take class D seriously.