On Class D Amplifiers Measurements

[March Audio]

More on complex load and square wave stability tests in
https://www.audiosciencereview.com/...ad-for-power-amplifier-torture-testing.10298/

Maybe March Audio might replicate that easy test on some of his class D modules?

Im happy to but you test again isnt representative of the real world. Finding one badly designed speaker doesnt really tell us anything useful.

 

[SIY]

More on complex load and square wave stability tests in
https://www.audiosciencereview.com/...ad-for-power-amplifier-torture-testing.10298/

Maybe March Audio might replicate that easy test on some of his class D modules?

See my comment in your load thread.

 

[eliash]

To Amir:
Having worked myself through the 22 pages of comments on the subject and acknowledged also the outstanding specs of the Purifi 1ET400A amplifier,
one thing missing to make me a believer would be some "32-tone signal resembling music showing the noise and distortion" test, like for the DACs at decent power.

...actually if s.o. chops my delicate >110dB dynamic range analog signal some 500K times per second and converts level into digital pulse width, one may be suspicious...and "only" 60KHz power bandwidth was something, which (if I remember correct...my old Yamaha A960 had 50KHz, actually never had any concern about it...) was also in question in the 80s until some "safe" >150KHz were achieved by new semiconductor technology...(T+A claims 300KHz power bandwidth for my actual power amp, first version designed in the 90s...)

 

[pma]

Isn't this nice? (Wouldn't it be nice to ...)

I hope especially @restorer-john might appreciate this

(From my Kindle Stereophile edition)

 

[LTig]

Isn't this nice? (Wouldn't it be nice to ...)

I hope especially @restorer-john might appreciate this

View attachment 42212

That's the reason why I don't care about THD above 10 kHz.

 

[Spocko]

That's the reason why I don't care about THD above 10 kHz.

And more useful than just THD, specifically, is distortion mostly even order harmonics or lower order harmonics? 2 pieces of equipment may share similar THD but if most of the THD on one piece of equipment is odd higher order harmonics, it would be audibly much worse.

 

[pma]

That's the reason why I don't care about THD above 10 kHz.

Interesting view, if you'd like to learn more, you might find the article below useful

https://www.researchgate.net/public...n_of_Nonlinear_Distortion_Measurement_Methods

published by Richard Cabot, Audio Precision, Beaverton, USA

 

[ahofer]

A new allegation on Audiogon on Class D, from ridiculous cable-purveyor Teo Audio:

https://forum.audiogon.com/discussi...endgrid&utm_medium=email&utm_campaign=website

the dynamic load of the drivers and crossover will cause massive changes in the impedance of the speaker system. The drivers shift all over the map, thousands of times per second.

As high as a few hundred ohms, and as low as into almost negative numbers.

This is reacting against the Class D output filter.

This means the filter and the driver combination, is all over the map in their effectiveness of blocking or dealing with the amplifier’s output.

The trick is that the drivers are not static in their impedance value.

Which in turn, causes the effectiveness and the nature of the output filter to be all over the map.

If one measures with low values of signal and/or a sine wave, none of this real world stuff will be visible in the measurements.

I don't believe I've seen this with white noise or simulated loads. @March Audio ?

 

[SIY]

A new allegation on Audiogon on Class D, from ridiculous cable-purveyor Teo Audio:

https://forum.audiogon.com/discussi...endgrid&utm_medium=email&utm_campaign=website



I don't believe I've seen this with white noise or simulated loads. @March Audio ?

Why pay any attention to this?

 

[sq225917]

https://www.trevormarshall.com/class-d-tutorial/

Worth a read as it pertains to real world loads.

 

[boXem]

https://www.trevormarshall.com/class-d-tutorial/

Worth a read as it pertains to real world loads.

Even if the content is interesting, the title of the article is a bit misleading. It's mainly dealing with output filter issues when the filter is not part of the control loop.

 

[EB1000]

In about 10 years from now, class-d amps will operate in the several MHz switching freq range. Then it will be a real game-changer. Until then, I would not invest too much in Class-D amp, or just stick to class-AB. I'm currently using 2 ICEpower 125ASX2 BTL (runs around 600kHz). I wanted to upgrade to NC400, but the limited 460kHz was a deal-breaker for me...

 

[boXem]

In about 10 years from now, class-d amps will operate in the several MHz switching freq range. Then it will be a real game-changer. Until then, I would not invest too much in Class-D amp, or just stick to class-AB. I'm currently using 2 ICEpower 125ASX2 BTL (runs around 600kHz). I wanted to upgrade to NC400, but the limited 460kHz was a deal-breaker for me...

What would be the advantages of faster switching? What are the weaknesses of the actual switching speeds?

 

[EB1000]

What would be the advantages of faster switching? What are the weaknesses of the actual switching speeds?

The higher the switching frequency, the lower the harmonic distortion, and the higher the bandwidth (much easier to filter out higher harmonics contents than lower, which lower firm factor filters and lower order of the filter = lower losses). GaN FETs can reach 2MHz at 600V/100A, making then ideal for class-D. One of the class-d's weakest chain is the LPF filter, which needs to remove the 500kHz while keeping frequencies up to 20kHz intact. Most LPF filters are 2nd order LC, which cannot handle such a task.

 

[March Audio]

The higher the switching frequency, the lower the harmonic distortion, and the higher the bandwidth (much easier to filter out higher harmonics contents than lower, which lower firm factor filters and lower order of the filter = lower losses). GaN FETs can reach 2MHz at 600V/100A, making then ideal for class-D. One of the class-d's weakest chain is the LPF filter, which needs to remove the 500kHz while keeping frequencies up to 20kHz intact. Most LPF filters are 2nd order LC, which cannot handle such a task.

Purifi has manged exactly this and distortion is incredibly low, lower than most AB amps

 

[boXem]

The higher the switching frequency, the lower the harmonic distortion, and the higher the bandwidth (much easier to filter out higher harmonics contents than lower, which lower firm factor filters and lower order of the filter = lower losses). GaN FETs can reach 2MHz at 600V/100A, making then ideal for class-D. One of the class-d's weakest chain is the LPF filter, which needs to remove the 500kHz while keeping frequencies up to 20kHz intact. Most LPF filters are 2nd order LC, which cannot handle such a task.

1ET400A distortion is almost not measurable.
Hypex modules have been solving LPF issues since years.

 

[March Audio]

In about 10 years from now, class-d amps will operate in the several MHz switching freq range. Then it will be a real game-changer. Until then, I would not invest too much in Class-D amp, or just stick to class-AB. I'm currently using 2 ICEpower 125ASX2 BTL (runs around 600kHz). I wanted to upgrade to NC400, but the limited 460kHz was a deal-breaker for me...

So what do you hear as a problem?

I would generally comment that you shouldn't conflate any issues you think a particular brand of class d amp has (I don't like ice BTW) with there being a fundamental problem with class d.

 

[eliash]

So what do you hear as a problem?

I would generally comment that you shouldn't conflate any issues you think a particular brand of class d amp has (I don't like ice) with there being a fundamental problem with class d.

Not questioning the great looking (and probably also great sounding) capabilities of Purifi et al, but looking at power bandwidth, there is room for improvement.
In the eighties of the last millenium power bandwidth of class AB designs improved from ~50KHz (as class D has achieved now) to more than 150KHz, thanks to better semiconductors, which turned out to be some kind of iconic value until now.
I bet, in some ten years or maybe even faster, if still in business (hope so!), you will be promoting the fact that class D amps have finally achieved the same power bandwidth as class AB amps after all...

 

[March Audio]

Not questioning the great looking (and probably also great sounding) capabilities of Purifi et al, but looking at power bandwidth, there is room for improvement.
In the eighties of the last millenium power bandwidth of class AB designs improved from ~50KHz (as class D has achieved now) to more than 150KHz, thanks to better semiconductors, which turned out to be some kind of iconic value until now.
I bet, in some ten years or maybe even faster, if still in business (hope so!), you will be promoting the fact that class D amps have finally achieved the same power bandwidth as class AB amps after all...

Simple question, why?

Last time I checked most peoples hearing only goes to 20kHz. Musical content at 20kHz might well be 60dB down on 1kHz, so what is the benefit of increasing power bandwidth?

 

[A800]

Last time I checked most peoples hearing only goes to 20kHz. Musical content at 20kHz might well be 60dB down on 1kHz, so what is the benefit of increasing power bandwidth?

Highly increased airiness.