• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

On Class D Amplifiers Measurements

March Audio

Master Contributor
Audio Company
Joined
Mar 1, 2016
Messages
6,378
Likes
9,317
Location
Albany Western Australia
There are also DAC's that may produce US crap all the way up there when music is playing. Think of those non filtered NOS DAC's that have no filtering whatsover (because filters are baaaaaadd) that may produce crap all the way up there that may fold back.
Then again ... such individuals deserve to be be punished :D

Anyways... when a swicthing amp works on a high enough frequency and has proper input filtering on most situations I don't think it will lead to any practical issues.

Absolutely, I say let them suffer for their ignorance ;) haha :)

I have said right from the start this is a possible problem, the mechanism is there, but that doesn't mean it happens in normal circumstances.

I will post some other tests tomorrow if I get time, I have various dacs, Blu Ray players etc.

Another thing to consider is that you need to be using the amp with a dac directly to create the situation. A pre amp may be unlikely to pass it through
 
Last edited:
OP
pma

pma

Major Contributor
Joined
Feb 23, 2019
Messages
4,591
Likes
10,727
Location
Prague
Yep. Now plug it into an amp and demonstrate it causes problems.

So again you post a plot which contradicts your assertion that 10mV is a representative and realistic noise level. You have levels at the switching frequency of about -90dBV which is 31 uV.

Have to ask, that hardly seems a quality player. Whats the 60kHz and harmonics all about?

I have no NC400 here, you do, so I cannot test what happens. My amps are designed with respect to HF interference at 1st place. That's why they have very high BW and input LPF can be made very high above audio band.

"So again you post a plot which contradicts your assertion that 10mV"

mV level is enough to trigger NC400 aliases and has to be near two times switching frequency.
The most sensitive frequency is at 2*fsw stands to reason since there are 2 sampling points per switching period. Have you checked all the players? You know there will never be a problem?

60kHz is most probably from SMPS. Many players have this issue. Yes the player was cheap, like $35, similar price as for some junk that Amir is testing here.
You will find everything at the output of this player, because of fast edges of the spikes.
1574859486258.png


You do not care, right? Well I do not care wasting my time with ignorants or marketing promoters.
 

March Audio

Master Contributor
Audio Company
Joined
Mar 1, 2016
Messages
6,378
Likes
9,317
Location
Albany Western Australia
I have no NC400 here, .

"So again you post a plot which contradicts your assertion that 10mV is a representative and realistic noise level"

mV level is enough
to trigger NC400 aliases and has to be near two times switching frequency.
The most sensitive frequency is at 2*fsw stands to reason since there are 2 sampling points per switching period. Have you checked all the players? You know there will never be a problem?

60kHz is most probably from SMPS. Many players have this issue. Yes the player was cheap, like $35, similar price as for some junk that Amir is testing here.
You will find everything at the output of this player, because of fast edges of the spikes.
View attachment 40490

You do not care, right? Well I do not care wasting my time with ignorants or marketing promoters.


Yep and and so far everything you have posted is in the low uV range, so what's your point?

You are contradicting yourself with your own data.

I do care which is why I am correcting your erroneous assertions. I have not said there can never be a problem, but you are clearly grossly exaggerating the possibility and trying to spread FUD.

So again you resort to being rude and offensive instead of technically justifying your claims. That's disappointing.
 
Last edited:

tmtomh

Major Contributor
Forum Donor
Joined
Aug 14, 2018
Messages
2,635
Likes
7,486
I lack the technical knowledge of @restorer-john , @March Audio , and @pma . But as a layperson reading through this discussion, particularly the disagreements between March Audio and Pavel/pma, I see pretty clearly that pma is able to produce signals and tests that can reveal certain design characteristics of Class D amps. But - and I think this is part of March Audio's point - I have not yet seen any decisive evidence that these design characteristics are design flaws.

So for example, if I use a Class AB amp and I drive it hard enough to put it into Class B mode, it will have switching distortion, yes? So should the possibility of switching distortion on an amp running in Class B mode be more or less of a concern to me than the possibility of a Class D amp exhibiting some phase nonlinearity in frequencies above, say, 15kHz?

My answer is, I don't know. I don't know because I don't know how to compare those two different kinds of nonlinearities (aka deviations from perfect fidelity) - but also because I don't know how likely I am to run into either one. Do I really listen to my music loud enough to trigger my amp into Class B mode, and if so, can I even hear any switching distortion? Conversely, are any of my source components even able to produce the phase issues pma displays? And if so, if those phase nonlinearities occur above 15kHz, can I even hear them? (My 50 year-old ears can still hear up to slightly above 17kHz last I tested them with a frequency generator, but at my age I can only hear above 15kHz if I turn my head so one of my ears is more or less directly facing the speaker's tweeter. So I very much doubt I am hearing much above 15kHz when listening to actual music from my normal listening position.)

I'm not saying we shouldn't pursue and discuss these issues. I'm just saying that all topologies have certain characteristics, and I think those characteristics do not correlate in any simple or obvious way with audible problems or design flaws.
 

miero

Active Member
Joined
Aug 1, 2018
Messages
241
Likes
289
@tmtomh pma says, that nCore if fed with a signal with frequencies in the range of medium-wave AM radio, it can generate sound intermodulations that is in the hearable range. He knows it, because he tested my nCore NC400 and he discovered this issue.

It is a design "feature" of one of the best class-D amplifier Hypex nCore by Bruno Putzeys. AFAIK it was not a problem of his older Hypex UcD design and it should be fixed in his newest Purify Eigentakt amplifiers.

In most cases this HF noise sensitivy of nCore is not a problem, but the workaround (for all cases) is simple - use additional HF filter on input. Nevertheless input impedance of nCore isn't stellar and additional buffer might help.

EDIT: @March Audio corrected me bellow - Thanks!
 
Last edited:
OP
pma

pma

Major Contributor
Joined
Feb 23, 2019
Messages
4,591
Likes
10,727
Location
Prague
It is a design "feature" of one of the best class-D amplifier Hypex nCore by Bruno Putzeys. AFAIK it was not a problem of his older Hypex UcD design and it should be fixed in his newest Purify Eigentakt amplifiers.

In most cases this HF noise sensitivy of nCore is not a problem, but the workaround (for all cases) is simple - use additional HF filter on input. Nevertheless input impedance of nCore isn't stellar and additional buffer might help.

That's absolutely correct.
 

March Audio

Master Contributor
Audio Company
Joined
Mar 1, 2016
Messages
6,378
Likes
9,317
Location
Albany Western Australia
In most cases this HF noise sensitivy of nCore is not a problem, but the workaround (for all cases) is simple - use additional HF filter on input. Nevertheless input impedance of nCore isn't stellar and additional buffer might help.

?

Amps with integrated input buffers:

NcxxxMP modules input Z is 47kohms
Nc1200 94k
Nc400 104k

The other standalone ncore modules are supplied without buffers. These are Intended for use with buffer supplied by the vendor. Input Z is set by that buffer design.
 
Last edited:

March Audio

Master Contributor
Audio Company
Joined
Mar 1, 2016
Messages
6,378
Likes
9,317
Location
Albany Western Australia
I
I'm not saying we shouldn't pursue and discuss these issues. I'm just saying that all topologies have certain characteristics, and I think those characteristics do not correlate in any simple or obvious way with audible problems or design flaws.

Indeed, we do need to look at this, but I'm not quite sure that's whats been happening here.

I think some are forgetting that A or A/B amps are also sensitive to RF and should have input filters too.
 

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,579
Likes
38,273
Location
Gold Coast, Queensland, Australia
I think some are forgetting that A or A/B amps are also sensitive to RF and should have input filters too.

They do, all of them. Never seen one without multiple networks. The NFB loops also have 100% feedback at HF (and near DC) frequencies, determined by the designer. Sometimes (often) they are tailored to give certain amplifiers in the range a better or worse performance that the models above and below. A common trick to differentiate the offerings.

What HF filtering have you incorporated into the front end of your buffer stage BTW?
 

March Audio

Master Contributor
Audio Company
Joined
Mar 1, 2016
Messages
6,378
Likes
9,317
Location
Albany Western Australia
They do, all of them. Never seen one without multiple networks. The NFB loops also have 100% feedback at HF (and near DC) frequencies, determined by the designer. Sometimes (often) they are tailored to give certain amplifiers in the range a better or worse performance that the models above and below. A common trick to differentiate the offerings.

What HF filtering have you incorporated into the front end of your buffer stage BTW?
Secret sauce John ;)
 

March Audio

Master Contributor
Audio Company
Joined
Mar 1, 2016
Messages
6,378
Likes
9,317
Location
Albany Western Australia

stereo coffee

Active Member
Audio Company
Joined
Jan 2, 2019
Messages
143
Likes
44
You are forgetting the Long Wave broadcast band which operates from 150kHz. Stations tend to to have BIG power output up to 2Mw

https://en.wikipedia.org/wiki/Longwave

Noting there are very few remaining transmissions on LW, and the use of shielded cabling techniques, which is good practice to be used in amplifiers ,circumvents reception of such frequencies.
 

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,579
Likes
38,273
Location
Gold Coast, Queensland, Australia
You are forgetting the Long Wave broadcast band which operates from 150kHz. Stations tend to to have BIG power output up to 2Mw

Are there many of them left? I read the other month that the BBC longwave services have been retired. I know we don't do Radio Australia on LW anymore (I think it 3LO had a monster LW transmitter back in the day)

I just turned on my Sangean ATS-803 (AM continuous 150KHz-30MHz) and at 150KHz through to 250KHz, there's a absolute ton of noise/garbage pinning the signal strength meter. I'd have to start shutting stuff off to find the issues (it's not the PC or the router). It only goes quiet at 260KHz and I've got a beacon (morse) at 280KHz a single station (auto voice beacon) at 358KHz. Up to 2MHz, there's random crap.

Honestly, I reckon a few Class D amplifiers in the house would be nothing compared to the crap across my dial on this highly sensitive receiver.
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,198
Likes
16,981
Location
Riverview FL
Honestly, I reckon a few Class D amplifiers in the house would be nothing compared to the crap across my dial on this highly sensitive receiver.


You need to put an RF filter on that antenna input. It will cut a lot of that down.
 

March Audio

Master Contributor
Audio Company
Joined
Mar 1, 2016
Messages
6,378
Likes
9,317
Location
Albany Western Australia
Are there many of them left? I read the other month that the BBC longwave services have been retired. I know we don't do Radio Australia on LW anymore (I think it 3LO had a monster LW transmitter back in the day)

I just turned on my Sangean ATS-803 (AM continuous 150KHz-30MHz) and at 150KHz through to 250KHz, there's a absolute ton of noise/garbage pinning the signal strength meter. I'd have to start shutting stuff off to find the issues (it's not the PC or the router). It only goes quiet at 260KHz and I've got a beacon (morse) at 280KHz a single station (auto voice beacon) at 358KHz. Up to 2MHz, there's random crap.

Honestly, I reckon a few Class D amplifiers in the house would be nothing compared to the crap across my dial on this highly sensitive receiver.

BBC R4 LW is still going, they have been talking about shutting it down for years but it hasnt happened yet. Last I read it was going to be when the last of the TX valves pegs out.

https://www.bbc.co.uk/programmes/articles/98FthRzhxJ4z0fXYJnsvlM/about-radio-4

The shipping forecast is still used by the marine sector and the LW coverage area is very beneficial for this purpose.


Isnt it odd that Australia doesnt use LW for broadcast considering the distances involved.

Nothing of note in the band here.

1574906509584.png
 
Last edited:

March Audio

Master Contributor
Audio Company
Joined
Mar 1, 2016
Messages
6,378
Likes
9,317
Location
Albany Western Australia
Noting there are very few remaining transmissions on LW, and the use of shielded cabling techniques, which is good practice to be used in amplifiers ,circumvents reception of such frequencies.


Its alive and kicking in Europe.

Same (shielding / filtering) applies to MW.
 
Last edited:

maty

Major Contributor
Joined
Dec 12, 2017
Messages
4,596
Likes
3,160
Location
Tarragona (Spain)

maty

Major Contributor
Joined
Dec 12, 2017
Messages
4,596
Likes
3,160
Location
Tarragona (Spain)
Yesterday, https://www.stereophile.com/content/devialet-expert-140-pro-integrated-amplifier-measurements
As I always do with class-D amplifiers, I inserted an Audio Precision auxiliary AUX-0025 passive low-pass filter, which eliminates noise above 200kHz, between the test load. This eliminates the risk of driving the analyzer's input stage into slew-rate limiting, which would interfere with the distortion measurements. Without the Audio Precision filter, there was 120mV of ultrasonic noise present at the amplifier's outputs...

and the measurements... pfff but $6490. But it is fashion.

[Polish] https://audio.com.pl/testy/stereo/wzmacniacze-stereo/3055-devialet-expert-140-pro
Signal / noise ratio (A-weighted filter, relative to 1W) [dB] 71
 
Last edited:
Top Bottom