If that is the case, then @restorer-john these manufacturers published spec doesn't carry too much weight.
100% agree, even at the worse measured test frequency input, it's still less than .05%. But if that is the current technical limitations of Class D, then this gives a lot more reasons why Class AB is not antiquated and its not the technology of yesterday. AND. . .that Class A/B still is deserving of your desires today.
www.audiosciencereview.com
Agree that it's inaudible, just like 100db SINAD is way beyond our human hearing threshold, yet. . .here we are, still measuring and buying based on these measurements.This is all FAQ #7.
Class D FAQ
This topic is dedicated to Frequently Asked Questions about class D amplifiers. If you have been directed here the reason is that it is easier to concentrate the frequent answers here instead of repeating them in several topics. Q1: Class D power ratings cannot be true. How can these small and...
Dear @restorer-john
My guess is that the AP AUX filters are not rated for precision measurements above 40 kHz. Amir reports his wide bandwidth THD+N to 45 kHz, which covers HD3 up to 15 kHz.
We can debate the merit of measuring up to less than 20,000 Hz, or even a little higher, but I don't think anyone would claim, or even suspect they can hear 30,000 Hz and above. So the time and effort to measure SINAD at 20 kHz or higher should better be spent on additonal test/measurements that are more useful, potentially. Even at 20 kHz, as you know the lowest frequency of the harmonics is 40 kHz. Above that, even dogs, including the OCD ones, are not going to complain:Agree that it's inaudible, just like 100db SINAD is way beyond our human hearing threshold, yet. . .here we are, still measuring and buying based on these measurements.
I can't tell you why others do it, but I do it because I appreciate the engineering marvel, I have OCD and it gives me a peace of mind and let's me sleep better at night.
The statement from @restorer-john was: "Peak or continuous power at 1kHz has always been a cop-out for amplifiers incapable of delivering their rated continuous power across the audible bandwidth. Because amplifiers back in the day (and plenty now) could not produce their rated power at frequency extremes. It's always been the "realm" (your word) of manufacturers who attempt to game the system with smoke and mirrors. It was one of the main reasons the Amplifier Rule was created and enforced in the first place!"
The statement from @restorer-john was: "Peak or continuous power at 1kHz has always been a cop-out for amplifiers incapable of delivering their rated continuous power across the audible bandwidth. Because amplifiers back in the day (and plenty now) could not produce their rated power at frequency extremes. It's always been the "realm" (your word) of manufacturers who attempt to game the system with smoke and mirrors. It was one of the main reasons the Amplifier Rule was created and enforced in the first place!"
Smoke and mirrors?
Certainly, I will proceed with measurements at the elevated frequencies as requested. Let me know if there are any additional specifics you'd like addressed.
I think that Apollon deserves credit for continuing to provide patient, thoughtful replies while basically being attacked for delivering high quality products and service at great prices.The statement from @restorer-john was: "Peak or continuous power at 1kHz has always been a cop-out for amplifiers incapable of delivering their rated continuous power across the audible bandwidth. Because amplifiers back in the day (and plenty now) could not produce their rated power at frequency extremes. It's always been the "realm" (your word) of manufacturers who attempt to game the system with smoke and mirrors. It was one of the main reasons the Amplifier Rule was created and enforced in the first place!"
Smoke and mirrors?
Certainly, I will proceed with measurements at the elevated frequencies as requested. Let me know if there are any additional specifics you'd like addressed.
Too be fair to @restorer-john, he is pointing out a legitimate reason why he wanted 20kHz. And he also pointed out the reason why (most) manufacturers don't do it.The statement from @restorer-john was: "Peak or continuous power at 1kHz has always been a cop-out for amplifiers incapable of delivering their rated continuous power across the audible bandwidth. Because amplifiers back in the day (and plenty now) could not produce their rated power at frequency extremes. It's always been the "realm" (your word) of manufacturers who attempt to game the system with smoke and mirrors. It was one of the main reasons the Amplifier Rule was created and enforced in the first place!"
Smoke and mirrors?
Certainly, I will proceed with measurements at the elevated frequencies as requested. Let me know if there are any additional specifics you'd like addressed.
???So, I assume CleanSound wants to do the following one but with test frequency 20 kHz, Amir stopped at 15 kHz. He even admitted it is just for "OCD" so if you are kind enough to do it, I think people just want to see it once, no need to even do it again..
Thanks, now I remember.. seems like 15 kHz is a better limit especially for class D amps. I actually think 10 kHz is enough, and it wasn't me who suggested the 20 kHz test.???
Actually, until this review, I ran the test up to 20 kHz. I do not show it because the measurement is misleading as it under-represents the level of distortion due to limited bandwidth of 45 kHz. Here is the sweep from the amp I tested before this one with 20 kHz now highlighted in red:
View attachment 311540
Notice how it is much lower value than 10 and 15 kHz (blue and green) which obviously is not correct. 45 kHz only allows one harmonic to be included in the computation of THD whereas we have 2 to 3 in 10 and 5 kHz respectively.
You might say that measurement bandwidth should be expanded to include more harmonics. We had this argument at length when I first introduced this test. If I increase the bandwidth, it will suck in the noise shaping that occurs in class D amp. That would cause people to then think there is distortion when in reality it is ultrasonic noise. You can see this in the spectrum of Purifi amplifier:
Notice how noise level increases as we approach 100 kHz. Even at 45 kHz I am probably picking up some of that and with it, disadvantaging class D amps a bit.
Net, net, this is a difficult test to get right and we are the only outfit that is running it against every amp right now. It gives us insight into the transfer function of amplifier at multiple frequencies so the criticism that we only focus on 1 kHz is flat wrong.
Note that the wiggles you see in class D amps like this are more highlighted due to their very low noise. If noise were higher, it would mask them and graphs would look much smoother.
He is not fair. He has read the datasheet so he should very well know that these class D amps come with level of documentation and measurements that no traditional audio company provides. Here is the very thing he is asking about from Hypex NC500 datasheet:
Here you go:Only because you ask, I would add the one requested by @CleanSound on Amir. I suspectd Amir has already returned the DUT to it's owner, so there is no chance for him to do the 20 kHz test frequency graph even if he has time to do it.
So, I assume CleanSound wants to do the following one but with test frequency 20 kHz, Amir stopped at 15 kHz. He even admitted it is just for "OCD" so if you are kind enough to do it, I think people just want to see it once, no need to even do it again..
You probably would increase the bandwidth to the next step above 45 kHz, I assume..
The curve will be expectedly uglier, but I think no one will complain if your graph emphasizes the ugliness is due to noise and distortions in the ultrasonic range.
And for the FFT, it would be like the following it, but test freq 10 kHz, or at least 5 kHz, if possible:
Thanks again.
You are very welcome. We offer global shipping. Kindly be advised that the measurements provided pertain to the NCx500 DM dual mono version, which I have readily available in my studio. At this moment, I do not have the smaller NCx500ST version with one power supply on hand. The distinguishing feature of the dual mono version is that each NCx500 module is paired with its own dedicated Hypex SMPS1200 power supply.Many thanks! This amp is super clean, do you ship to Canada? Wonder what the shipping and import duty would cost, in addition to our HST, harmonize sales tax.
I hope you will include such test data on your website, that will for sure keep the OCD type happier.
