Class D come of age - Coherent amplification

[Mr. C]

I would agree a lot of the PWM's have a house sound, there are few that do not.

Switch frequency noise can play a large part of that issue, however most of these devices switch around 400-600Khz, and with SMPS the idea was to cancel out the switching of the amp/smps with each other.

Part of the benefits of the device I use is that you can move the switching frequency around by a large amount, which can help a great deal, is not the only issue that can be beneficial though.

[Joe]

Well, I dunno. I use two systems, one with AB amplification; the other with class D amplification. They both sound great. I've never noticed much 'signature' colouration with either, or 'switching noise' with the latter. Of course, it's entirely possible that my aged ears can't distinguish such subtle effects, and/or that I'm easily pleased.

[anthonyTD]

I have some experience of Class D amps, [going back as far as 10 years or more] some were good, with respect to musicality, and some not so good, I also agree that some of the best sounding were/are low power versions, when you compare the technology involved in three of the main topologies, I would say that Class D when done well is the closest to the best valve amps out there, they both have something in common which I feel is key, that is; they both have a measureable output impedance, valves amps having an output transformer on its output, Class D having the choke/ filter network at the end, this in turn allows the output to vary somewhat inline with the speaker's change in impedance with frequency, which all conventional cone speakers suffer from.
If you compare this to the way conventional transistor amps work, ie; they have a very low [almost immeasurable] output impedance, this is very good when the nominal speaker impedance decreases, hence the amp will put more power into the speaker load, However; as the speaker impedance rises, well,,, its a different story.
Anyway, just my ramblings.
If Tony has managed to overcome some of the failings that were a real hurdle in original class D technology, then I say great!

[Marco]

Imho the switching frequency of the module and gate driver technology are among key elements that are required to help bring a quality sound from a switching amp.

Interesting, Tony, especially given what I've just written. However, does it address the noise inherently injected (specifically its detrimental effects on the SQ of equipment sharing the same mains supply) by that type of amplifier topology?

Seriously, though...

In that respect, it would be interesting to see what would happen if you set up two systems in your demo room, both sharing the same mains supply: one using one of your Class D amps, and another using something else that doesn't switch (perhaps a good valve amp), and whilst listening to the latter, see whether the effect of switching off the Class D [taking it out of circuit] had any effect on SQ?

Any mains conditioning you were using would, of course, also need removing before the test took place.

Or maybe, since you have access to measuring equipment, and know what you're doing, measure the effect on the waveform of the mains supply when the Class D amp is connected and switching, compared with one that doesn't switch, and observe the differences?

Those are the sorts of tests I would need conducted before I'd consider buying any Class D amplifier, not just yours, simply because, as a purist, the noise issue worries me

Marco

The amplifier in room we used at Cranage was a TAD Implementation of a pwm switching amplifier their own technology again 5th harmonic bandwidth and a very good example

Thanks for that, but was it Class D, or if not, what? Sorry, I don't know what "pwm" is.

Marco.

[Mr. C]

Hi Marco

Not a problem, if a PWM (Pulse width modulation) amp. Class 'd' is purely is fed by a digital input only, things like the TACT and Devialet use internal A2D to convert the incoming analogue signal to digital then through the module and a low pass filter to take out any unwanted high frequencies artifacts.
A PWM can take analogue inputs without any A2D conversion.

PWN are not strictly Class 'D'

I have no problem in comparing these amps to traditional class a/b valves or a for that matter (with zero mains conditioning), as with everything if you pay careful attention to all of the issues within a design of this kind, especially RF propagation both external and internal you can produce some very pleasing results imho.

[Pharos]

This thread has got me all paranoidly worried now.

[Marco]

Hi Marco

Not a problem, if a PWM (Pulse width modulation) amp. Class 'd' is purely is fed by a digital input only, things like the TACT and Devialet use internal A2D to convert the incoming analogue signal to digital then through the module and a low pass filter to take out any unwanted high frequencies artifacts.
A PWM can take analogue inputs without any A2D conversion.

PWN are not strictly Class 'D'

I have no problem in comparing these amps to traditional class a/b valves or a for that matter (with zero mains conditioning), as with everything if you pay careful attention to all of the issues within a design of this kind, especially RF propagation both external and internal you can produce some very pleasing results imho.

Oh no doubt, and that also reflects Anthony's view, plus the limited times I've heard Class D

However, it would be interesting carrying out the test I suggested, particularly the measurements, as if anything awry was detected, either there or when listening, then *that* noise/distortion is what's being 'invisibly' superimposed onto the music, simply by way of how Class D works, and which one would be unaware of under normal listening conditions.

Marco.

[Clive]

Hi Marco

Not a problem, if a PWM (Pulse width modulation) amp. Class 'd' is purely is fed by a digital input only, things like the TACT and Devialet use internal A2D to convert the incoming analogue signal to digital then through the module and a low pass filter to take out any unwanted high frequencies artifacts.
A PWM can take analogue inputs without any A2D conversion.

PWN are not strictly Class 'D'

It's only words so not really of any importance but...is there a solid definition of class d? People generally think of class d as denoting a digital amplifier when they are usually PWM. We can also think of "d" the as being a random letter which people then unfortunately ascribe to meaning digital. I think of class d as being predominantly PWM (switching) amps.

[anthonyTD]

For my part Clive, I appologise, your right, many of us seem to have adopted the D for some reason, to describe all configurations of this topology.

It's only words so not really of any importance but...is there a solid definition of class d? People generally think of class d as denoting a digital amplifier when they are usually PWM. We can also think of "d" the as being a random letter which people then unfortunately ascribe to meaning digital. I think of class d as being predominantly PWM (switching) amps.

[Mr. C]

Marco

I believe you (Anthony) is referring to the high frequency artifacts that are present in the output form all PWM switching type amplifier, if you look at say a 4Khz wave form (sinusoidal) you will see a 'furry' wave form

This is the result of the switching process which leaves 'noise' on the output, however it is far outside the pass band to be audible after going through the LP filter which is usually set between 50-80Khz

That is very simplified explanation

More in depth below:-

https://www.edn.com/design/consumer/...-amps--Part-2-

They are couple of PWM amps on the market which really do not have any 'house sound' they are a close to a straight piece of wire with gain as possible, and have compared these to the DTZ mono's which are pretty much at the pinnacle of commercial available designs.

Whether an individual likes them is just that personal nothing more, if you put 10 people in a room, 5 will say brilliant, 4 will say crap worse ever sound, and the last one will not be able to give an answer until he has consulted his best friend or forum

Basically technology marches on and materials and components, become, smaller with vastly improved heat dissipation, better form factor, more controllable, better dielectric's, capacitor materials, circuit board layout, better spec's etc.

The future is coming and it is not to far away from many angles not just audio.