Yes you could ask, but that wouldn´t make any sence, at this is a twin loop design, where Purifi and Hypex nCore are single loop.
And as you can see, the gain is very low in both nCore and Purify, which calls for additinal amplification, and then most of the impressive noise levels are lost.
It is actually two very different ways to approach the class D challenge.
Would still be interesting to know the overall figure, though. How much are you applying in each loop?
Well it is considerably lower.
But one should remember the the purposes of the loop in class D amps.
First of all they are needed to make the amplifier selfoscilate.
That done, the loop technology can do a lot of distortion reduction.
I.e some will go for very tight dead time to reduce distortion, where others find sophisticated loop technology more attractive and lasting.
Anyways I don´t think you can use loop gain as a benchmark for performance.
I´d rather look at the outcome.
But one should remember the the purposes of the loop in class D amps.
First of all they are needed to make the amplifier selfoscilate.
That done, the loop technology can do a lot of distortion reduction.
I.e some will go for very tight dead time to reduce distortion, where others find sophisticated loop technology more attractive and lasting.
Anyways I don´t think you can use loop gain as a benchmark for performance.
I´d rather look at the outcome.
This will be available somewhat later, but it is a lot of work.
Measurements are already done, but only with lab supplies, and that is not really what you want to see.
But most important is the fact, that the group delay throughout the audioband is constant 9 µ sec, which to my knowledge, is not seen in any other designs.
Non linear group delay is distortion just as well as harmonic distortion is, but rarely considered.
Measurements are already done, but only with lab supplies, and that is not really what you want to see.
But most important is the fact, that the group delay throughout the audioband is constant 9 µ sec, which to my knowledge, is not seen in any other designs.
Non linear group delay is distortion just as well as harmonic distortion is, but rarely considered.
Out of plain curiosity, why did you choose the nCores as reference? From your earlier posts I would have thought the 1200AS would be your reference.
This would be evident through any nonlinearity of the phase plots (vs frequency).
Hi Stein
As the person who started the thread I am OK with Søren's posts so far.
In fact I have a question for him.
If the second loop is nested inside an outer loop then it makes perfect sense to discuss the feedback applied.
You can plot either the inside of the nested loop, or the outer loop with the inner loop considered as a "black box".
So what do the loop plots look like?
Best wishes
David
In ICE Power there is an inside and an outside loop.
This is to understand, that the inner loop is a pre-filter loop.
The outer loop is a post-filter loop.
This makes the amplifier both very stable and the filter can to some extend be considered imune to impact from the load.
Also Zout will be lower this way.
In our design ther are no such things as inner and outer loops, and that is one of the big differences between this next gen technology and ICE Power.
With an inner and outer loop, you put stress on the components, as they are to be very fast, because the signal contains high frequencies.
The next gen doesn´t really have any need for speed, but it still has twin loops, creating very stable operation.
This is to understand, that the inner loop is a pre-filter loop.
The outer loop is a post-filter loop.
This makes the amplifier both very stable and the filter can to some extend be considered imune to impact from the load.
Also Zout will be lower this way.
In our design ther are no such things as inner and outer loops, and that is one of the big differences between this next gen technology and ICE Power.
With an inner and outer loop, you put stress on the components, as they are to be very fast, because the signal contains high frequencies.
The next gen doesn´t really have any need for speed, but it still has twin loops, creating very stable operation.
The 1200 AS is my favourite, but regarding measurements the nCore is better.
That´s what I meant.
Sorry for the confusion.
Ok, it's still possible to analyse the feedback, Bode showed how way back in 1943.
He used the term "Return Ratio" to indicate it was not linked to any specific loop.
Mason later showed a nice visual way to represent this.
What does your loop structure look like?
Best wishes
David
Thank you. Glad to see you value what your ears tell you above measurements!
David
Hopefully you understand that Søren will never show anything when it comes to his "design" He and his partners is only interested in bragging and marketing their "design".
For them It's a good idea to to take over this thread and have their amp mentioned as often as possible when SOTA is discussed.
All the best
Stein
All music is produced for A and AB class amplifiers and will never sound as good as it should on a class D. Without some sort of signal processing or an equaliser a class D will lack warmth but have higher definition. This is not a flaw in class D amplification, it is just physics. I use Audirvana on a Mac with a free 64bit equaliser and just adjusting the levels by only tiny amounts it changes the sound drastically, I can add warmth without losing any clarity/definition. For those who feel it is important that the signal is bit perfect I would say, why don’t all DACs sound the same? Each DAC does its own thing to the bits, they don’t all convert the same way. It is important to use a good equaliser, I tried a few that were terrible and in the end I went with one by Tokyo Dawn Records that when almost flat I could not tell it was active. This is just my opinion, but the point I am trying to make is that music has not been produced for class D and music levels are a slightly wrong for it. I have only experimented using an equaliser for about week but have been amazed what a subtle adjustment could do.
I think he is talking about harmonics or dumping factor, not frequency itself. Mimicking lower damping factor amp with bass EQ would not be a terrible idea depends on situation / preference.
Controlling H2 /H3 has been popular on Pass forum, and I add some low frequency H2 / H3 for Class-D amp, which actually sounds pretty good to my ears for casual listening. We can find some free VST plugins adding harmonics, so I would recommend to try one if the music source is something like Audivana.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.