Peachtree new GaN-1 all "Digital Amplifier" the future?? (and it's not Class-D)

Status
Not open for further replies.
Jump to Latest
#245
With a normal class D amp you have to feed it an analog signal, which it converts to PWM, and then back to analog to send to the speakers. This amp directly converts a PCM signal to PWM and then back to analog to send to the speakers? Is that correct? If so, this is a distinctly different kind of device than a class D amplifier that is fed an analog signal. There is one less analog conversion step, with a digital to digital conversion step instead. Basically DSP instead of A/D conversion on the input. I think it's fair to call this a digital amplifier, or a power DAC, and needs to be differentiated from analog feed class D. Whether there's any real audible benefit to this is another question.
 
#249
ASCTim said:
You said NO!!! but then proceeded to say what I said.
Click to expand...
Oxymoron, sorry I screwed up
Yes!! we both said the same thing.
The two Peachtree amps take in digital from a digital source and keep it that way till the output stage analog to the speakers.
Normal retail Class-D's take in analog from a source d/a converter into it's input stage then convert it back to digital again it's untill the output stage back to analog to the speakers.
So in comparsion we have from a digital source direct into the Preachtree amps, 1 x less D to A conversion stage, no global feedback/just local, yet still .0004% distortion which all has to add up to be better for sound quality.

Cheers George
 
Last edited:
  • Like
Reactions: ASCTim
#250
georgehifi said:
Normal retail Class-D's take in analog from a source d/a converter into it's input stage then convert it back to digital again
Click to expand...
No it does NOT. here is where you understanding of the standard consumer Class-D amp is WRONG.

No digital conversion inside.

How hard can it be. Nothing digital in a switching amp. Period. Read up on it - but I doubt you arable to digest the tech involved or you would already know.

Here is some popular tech level information I just found from 30 seconds of googling... never heard of the company but they of course get it...

https://www.adam-audio.com/en/technology/pwm/

Read "PULSE WIDTH MODULATION" section George!

//
 
  • Like
Reactions: ASCTim
#251
So many misconceptions:

Personally I like to think of the pulse width as having discernible, quantized steps; discernible meaning it could be converted back to the source digital word as a trivial effort, such that the two representations are identical information wise except in different forms. Then, as convenient via the PW form, it becomes (how I look at it anyway) analog via the electrical action of the output filter; current going back and forth through the inductor, in developing one of those "continuous" forms of electricity across the capacitor. In my mind (which is a wild and crazy place) that's where the conversion from D to A actually takes place in these PCM to PWM translation designs.
Click to expand...
This is basically the correct description of what is going on within an amp like some TI ones, the TacT Millennium and also that Peachtree amp. But the error lies in the assumption that the digital to analog conversion takes place in the output filter. The conversion of the PWM stream into a "normal" (call it baseband) audio signal is taking place in the output filter. The digital to analog conversion is actually taking place in the PWM modulator.

It does not. It's continous. It glides continuously on the slope of a triangle wave. No numbers (0,1,2... 65535) involved. So not digital (digit=number....).
Click to expand...
In amps like the aforementioned TacT et al the PCM to PWM conversion IS actually performed in the digital domain by the use of counters. Since this would ask for exorbitantly fast counters with very high precsision just for a 44.1 ks/s 16 bit digital signal - let alone anything with higher sample rate and resolution - some tricks are used. The counters used have a resolution of only a few bits so they can be constructed with reasonable effort. In order to still achieve the desired resolution the digital input is up-sampled and noise-shaped. So amps working like this are actually power DACs.
Apart from this principle there were also attempts to use singe-bit delta sigma modulation to feed the swiching stage. The GaN FETs make this an interesting option again.


I don't doubt that the Peachtree amp is a fine piece of engineering. But I doubt that it is disruptive technology (apart from the fact that its working principle is not new at all).
If using a digital source for replay there are as many DA conversions taking place whether it is a line-level DAC feeding an ordinary class-d (or conventional) amp or the conversion is taking place in a digital modulator feeding an output stage. But the demands on switching precision and PSU stability are much higher in the second case.

Regards

Charles
 
Last edited:
#253
Yes but you stated that normal class-d amps perform ad and da conversion - they do not.

Here is your statement: "Normal retail Class-D's take in analog from a source d/a converter into it's input stage then convert it back to digital again"

It is form post #249 above.

Comment?


//
 
#254
phase_accurate said:
So many misconceptions:


... But the error lies in the assumption that the digital to analog conversion takes place in the output filter. The conversion of the PWM stream into a "normal" (call it baseband) audio signal is taking place in the output filter. The digital to analog conversion is actually taking place in the PWM modulator.
....

Charles
Click to expand...
You have it wrong too. No "digital to analog conversion is actually taking place in the PWM" - this is false - nothing going on like you describe it. There is no D/A process taking place in a basic PWM amp.

//
 
Last edited:
#255
PWM: "The comparator serves the purpose of comparing the two input signals and outputs a PWM signal. When the sawtooth triangle signal is higher than the modulation signal, the PWM signal at the output is at a HIGH. If the sawtooth triangle signal is lower than the modulation signal, then the PWM signal at the output is LOW."

This is the working of our amps if they have analog input:

"This circuit utilizes a triangle wave generator and comparator to generate a pulse-width-modulated (PWM) waveform with a duty cycle that is inversely proportional to the input voltage. An op amp and comparator generate a triangular waveform which is passed to the inverting input of a second comparator. By passing the input voltage to the non-inverting comparator input, a PWM waveform is produced. Negative feedback of the PWM waveform to an error amplifier is utilized to ensure high accuracy and linearity of the output"

from: https://www.ti.com/lit/ug/slau508/slau508.pdf?ts=1698133750960&ref_url=https%3A%2F%2Fwww.google.com%2F

//
 
#256
You have it wrong too. No "digital to analog conversion is actually taking place in the PWM" - this is false - nothing going on like you describe it. There is no D/A process taking place in a basic PWM amp.$
Click to expand...
Agreed, but I was not talking about a "basic PWM amp" but one with a digital modulator, i.e. a power DAC. And I stated that clearly. 😉
Mybe the expression "TI ones" might have been a bit misleading in this context.

Regards

Charles
 
#260
What about this take.
From wikipwdia: "In electronics, power amplifier classes are letter symbols applied to different power amplifier types."
That implies that an analog signal is involved.
If the amp do not have any analog input at any stage before the output transistors it is not an amplifier and can be called whatever class the constructor wants. It does not apply to class A, B, AB, D anyway.
Let me suggest Power PWM 😎
And this thread should not have been placed under
 
Status
Not open for further replies.
Top Bottom