About the highs... I've seen claims that class D amps have very large amounts of higher order distortion products at high frequencies, say for 10 kHz there are distortion products up in the 200 kHz range. I just cant believe those claims.... but it would be interesting to see an FFT of 10 kHz taken at a very high sample rate to see what is happening in the range up to 200 kHz or so. Can you do that?
They are nonsense. I know the game. Someone has a case to make (e.g. for the existence of supernatural beings, telepathy, etc) and they try to prove this by pointing at evidence which perchance hasn't been collected yet but which will surely vindicate their beliefs when provided. And when you say "bollocks" they'll then proceed to challenge you to disprove them, thereby suggesting that by default their theory holds. And by accepting to play their game you accept their terms. Well, before doing so, be aware of what follows: you do the hard work to do the experiment of which they claim to know the answer without bothering to try. You provide evidence which flatly refutes their claim. What happens? Instead of saying "you're right and I was wrong" they'll just invent another story for you to disprove. That tactic is called "moving the goalposts". It is one of a range of techniques nitwits use to give the impression that they are in a position to take on actual experts. I haven't the foggiest whose claims you are referring to but I'm imagining some forum discussion where someone says "oh THD is nothing, it's IMD that matters" and someone else says "go and check out the IMD plot then, it's all there" followed by the first speaker replying "yes but their plot stops at 20kHz and beyond that be dragons". It just keeps going, there's no telling what they'll come up with next.
Now, to answer your question technically. Ncore's loop gain is plotted in the white paper. Unlike linear amplifiers, open loop distortion is fully independent of frequency so the distribution of various harmonics is the same for all frequencies. You can take that as a given. You can read the distortion spectrum either from the 30W plot I posted earlier or you can glean a good indication from the IMD plot done at even higher output powers. So you look up how much lower loop gain is at various harmonics of 10kHz, you look up the amplitude of the matching harmonics from a known spectrum and from there you can work out what the HD would be at 10kHz. Since loop gain includes the output filter, be sure to scale the calculated HD back down 40dB/dec after 35kHz.
In short, the information is already there. Querulants often rely on their inability to read in order to make a point
Now, to answer your question technically. Ncore's loop gain is plotted in the white paper. Unlike linear amplifiers, open loop distortion is fully independent of frequency so the distribution of various harmonics is the same for all frequencies. You can take that as a given. You can read the distortion spectrum either from the 30W plot I posted earlier or you can glean a good indication from the IMD plot done at even higher output powers. So you look up how much lower loop gain is at various harmonics of 10kHz, you look up the amplitude of the matching harmonics from a known spectrum and from there you can work out what the HD would be at 10kHz. Since loop gain includes the output filter, be sure to scale the calculated HD back down 40dB/dec after 35kHz.
In short, the information is already there. Querulants often rely on their inability to read in order to make a point
Last edited:
people who strongly believe that WAVs sound intrinsically (as opposed to incidentally) better than FLACs do exist. they would testify to it.I have nothing against opinions of course, but when people start to spread lies, then my Viking blood starts to boil...
whether or not you bring arguments against the higher order distortion is of no use, because they can always resort to some other argument, no matter how scientifically incorrect or downright illogical. in their view it's still a valid argument. and when all arguments (valid or not) end there is the "i don't care how it measures, my ears know better" one. in the end there is the knowledgeable one wasting time and the dogmatic one still feeling superior.
is it only me who sees the futility?
Not at all - you're not alone. For practical reasons you've got to let people have their opinions but I don't have to treat those as valuable unless there's knowledge behind it. When people think they can teach me about designing amplifiers I just ask them where their design spreadsheets and their hardware are. If none are forthcoming I can do without their ideas as well.
That was what I hated about working for some Japanese companies. I was just the lowly slob who took care of the menial task of making the product work and they had the unfathomable magic genius to make it sound good. Right. Mind you, when people just have opinions about how something sounds without proffering technical advice I do listen because they might very well have an actual listening experience to report that I haven't had.
On another note and entirely off topic. In a local Italian restaurant there's a sign saying "Le opinioni sono como le palle, ognuno ha le sue". On the internet the quote is widely attributed to Clint Eastwood, except that I can only find it in Italian...
Last edited:
Hi,
Given that everyone should have the freedom to express themselves, however, the reader should distinguish the "balla". Certainly, not everyone has the experience to understand a few tricks. I agree with Bruno about the lack of documentation or new SMPS circuits to overcome the current difficulties on the web.
But this can bring out the real capabilities of a person. fill pages of theories and formulaic is simple (I did not this time). Then, the theories do not suggest the circuitry to get the result.
Regarding the sound, I spent much more time to adjust some parameters (staying in the range of measures sound good). not have sense, super excellent measure if the sound is recognized to Class D. (just only an example).
I know the real difficulties, I know every little reaction from thousands of combinations Attitude in D class. so do not believe a word of those who say (I'm referring to marketing in general). .. this amp sound good, nice sound, or those that limit the parameters (because it would be far too complex to extend them) and convince the others .. are not limited, but sounds better. or that it is useless to extend them. I think that with exactly the reality of 5 instruments playing, it is not enough even a class AB amp to 29kg. from € 20M.
Everyone grows their own backyard!
Regards
Roberto P.
Given that everyone should have the freedom to express themselves, however, the reader should distinguish the "balla". Certainly, not everyone has the experience to understand a few tricks. I agree with Bruno about the lack of documentation or new SMPS circuits to overcome the current difficulties on the web.
But this can bring out the real capabilities of a person. fill pages of theories and formulaic is simple (I did not this time). Then, the theories do not suggest the circuitry to get the result.
Regarding the sound, I spent much more time to adjust some parameters (staying in the range of measures sound good). not have sense, super excellent measure if the sound is recognized to Class D. (just only an example).
I know the real difficulties, I know every little reaction from thousands of combinations Attitude in D class. so do not believe a word of those who say (I'm referring to marketing in general). .. this amp sound good, nice sound, or those that limit the parameters (because it would be far too complex to extend them) and convince the others .. are not limited, but sounds better. or that it is useless to extend them. I think that with exactly the reality of 5 instruments playing, it is not enough even a class AB amp to 29kg. from € 20M.
Everyone grows their own backyard!
Regards
Roberto P.
Since I can't contribute much in the technical discussion:
Harry Callahan in The Dead Pool: "Well, opinions are like ********. Everybody has one."
Above the stars would be a word in two parts starting with the double-s ending US English version of the UK English word "bottom" followed by "holes".
Last edited:
What made me a bit grumpy was the fact that some guy said all class-D manufacturers deliberately manipulated their TDH measurements to make them look good. He claimed to have measured large amounts of 20th and 22nd order distortion products. That is not a subjective "my ears know what they hear" thing. Of course he could not provide any proof for this, and I could provide proof that is was not true.....
20th and 22nd order distortion?? The guy who is measuring this maybe better of writing a book called "How to create a problem"
Regards Gerhard
I guess they either don't know or they deliberatley claim that things like the AES 17 filter are there to make the measurements look nicer. They are indeed there because some analysers might have problems with the carrier residual.
Apart from that: Who would actually care about the 21st harmonic of say a 5 kHz signal. O.k. it might disturb some bats to some degree.
Regards
Charles
Ah, there are people that claim that harmonics up to 200 kHz is absolutely necessary to repropduce the overtones of musical instruments, hence the need for DXD, DSD and SACD. Some also use this as an argument to support their purchase of up-sampled 24/96 files from HDtracks.com... (which is rubbish by the way). The same people have probably never seen the frequency response of a Telefunken U-47 recording microphone.....
Hi,
I agree that 200Khz are not necessary.
But I disagree that you can limit the amplifier to 20Khz or do you think has the microphone as 22Khz bandwidth, does not contain the sounds captured information to reproduce the harmonic spectrum? musical instruments do not emit pure periodic sinusoids. especially when playing a group of musical instruments.
if you see the track, I'm sure you realize, how much work has to do the amplifier and ..loudspeaker .. very imperfect for this work.
Regards
I agree that 200Khz are not necessary.
But I disagree that you can limit the amplifier to 20Khz or do you think has the microphone as 22Khz bandwidth, does not contain the sounds captured information to reproduce the harmonic spectrum? musical instruments do not emit pure periodic sinusoids. especially when playing a group of musical instruments.
if you see the track, I'm sure you realize, how much work has to do the amplifier and ..loudspeaker .. very imperfect for this work.
Regards
- Status
- Not open for further replies.