hypex ncore

[serengetiplains]

Chris, I'm not going to sidetrack into digital-analogue semantics, but I will say this, and just once to keep from going OT. The 1-bit pulse train (call it what you will) is discrete, which is all I mean by "digital." Because it's discrete, it is a mathematical representation of the music signal and, as such, can be mathematically manipulated at source in a manipulation that includes the entire amplifier---or the entire stereo, if one is adventurous. In such a manipulation (call this feedback if you want), there by definition exists no time-problems applying the manipulation. It's pure. It gets to the very source of the signal itself.

[Julf]

Originally Posted by serengetiplains

Chris, I'm not going to sidetrack into digital-analogue semantics, but I will say this, and just once to keep from going OT. The 1-bit pulse train (call it what you will) is discrete, which is all I mean by "digital." Because it's discrete, it is a mathematical representation of the music signal and, as such, can be mathematically manipulated at source in a manipulation that includes the entire amplifier---or the entire stereo, if one is adventurous. In such a manipulation (call this feedback if you want), there by definition exists no time-problems applying the manipulation. It's pure. It gets to the very source of the signal itself.

I am not sure that made any sense. Sounds like what you are talking about is calculating some sort of global, static inverse of the transfer function of the amp/system, and applying that to the signal. That has nothing to do with feedback.

[ChrisPa]

This is already ot
If you use a digital feedback (or feedforward) system it's no longer class d

To use feedforward you need to predictably know about the system characteristics under all operating circumstances and component tolerances in order to provide predictive compensation for distortion

And therefore you use feedback
And you can choose to use digital feedback if you wish
Except it probably won't actually be digital - there will be some form of adc element to provide fine control

It's not pure it's just different

[serengetiplains]

Originally Posted by Julf

I am not sure that made any sense. Sounds like what you are talking about is calculating some sort of global, static inverse of the transfer function of the amp/system, and applying that to the signal. That has nothing to do with feedback.

Semantics aside, that's it, Julf. A Zetex designer captures what I was thinking in the Zetex thread. He uses almost the same language I used, which confirms I was at least on some realistic track:

The digital circuits then predistort the input to the PWM modulator which drives the actual power output stage to compensate for the timing and amplitude errors in it, such that it has the same performance as the reference DAC and the error signal tends to zero.

The end result is an amplifier with performance at the speaker output pretty much as good as the best DACs available have at the DAC output -- not my opinion here, this has been demonstrated by both measurements and listening tests. In fact, Zetex are also applying just the modulator chip as an 8-channel DAC, with performance up there with the best on the market.

I'm not guessing here about how this works, I designed the low-jitter (picoseconds) clock and PWM output stages that produce this reference PWM DAC output :-)

[waltzingbear]

Originally Posted by serengetiplains

Semantics aside, that's it, Julf. A Zetex designer captures what I was thinking in the Zetex thread. He uses almost the same language I used, which confirms I was at least on some realistic track:

that sounds like what has been called predistortion, this is a technique that has been used for many years. applying it to other systems (ie digital amps) is nothing new.

later I add:
I know nothing of the Zetex amps, just the description provided, which appears to not convey to me the correct topology. According to Bruno, see below, it is not predistortion.

Alan

[Julf]

Originally Posted by waltzingbear

that sounds like what has been called predistortion, this is a technique that has been used for many years. applying it to other systems (ie digital amps) is nothing new.

And predistortion only works when the non-linearity is predictable and non-dynamic.

[Bruno Putzeys]

We've had this discussion in this thread before. An approximate diagram of the Zetex amp is given in http://www.hypex.nl/docs/papers/AES124BP.pdf, page 51.

As you can see there is no predistortion, unless you want to call the signal internal to the feedback loop "predistorted", in which case ALL feedback amplifiers use predistortion.

The end result is an amplifier with performance at the speaker output pretty much as good as the best DACs available have at the DAC output -- not my opinion here, this has been demonstrated by both measurements and listening tests.

I would be interested to see those measurements. The ones I made are seriously at variance with that claim.

In fact, Zetex are also applying just the modulator chip as an 8-channel DAC, with performance up there with the best on the market.

That, however, is true.

[serengetiplains]

Originally Posted by Julf

And predistortion only works when the non-linearity is predictable and non-dynamic.

Yes, so one would have to come up with some way of tracking real-time, some sort of dynamic, adaptive algorithm, perhaps one that learns on the fly, but in any event one that receives then recalibrates the comparator pulse output to negate distortion products.

Thank you for digging out that reference, Bruno.

Enough speculation for now.

[slowlearner]

Thank all of you for your answers to my questions so far. I have one more question.

Can the Mogami shield, Nampon wire and the XLR 1Q all be connected to the same chassis screw or should they be connected to different points near each other?

Sorry for the simplistic question, but I am very close to finishing and I want to fire this thing up.

[waltzingbear]

Originally Posted by serengetiplains

Yes, so one would have to come up with some way of tracking real-time, some sort of dynamic, adaptive algorithm, perhaps one that learns on the fly, but in any event one that receives then recalibrates the comparator pulse output to negate distortion products.

its called negative feedback

Alan Garren