NEW GaN Amplifier : Dual ACM8816 GaN amplifier chips + AK4118AEQ [YJCO D816]

[Bleib]

The digital amp claims comes up here from time to time. Again, this is easy, so easy that even AI doesn't make any mistakes
"PWM (Pulse Width Modulation) amplifiers, commonly known as Class D amplifiers, are fundamentally analog, not digital. While they are often mistakenly called "digital amplifiers" because they operate in an On/Off switching state, their switching timing is continuous and based on an analog representation of the signal."

 

[KL....]

I agree with you there. That's how it is. All the more, I wonder how someone like you, with that attitude, ends up with not just one but two very expensive fully digital amplifiers in a row? Your words: It's an amplifier, and it sounds like any other amplifier

Thank you @Michael Kohler, perhaps the following is helpful....

• FM, DSD, and PWM are conceptually similar because they all encode information in the Time domain (varying when a transition happens or how long a state lasts) rather than the Amplitude domain (varying voltage height), although whether they can be termed Digital depends entirely on whether they are bound by a Time clock.
• Analogue PWM uses a Comparator (hence Analogue PWM) and features infinitely variable Pulse Widths.
• Digital PWM (such as with the AX5689C chip, a Digital Controler, which effectively replaces the Analogue Comparator/etc, Digitally with extremely high system precision) sets pulse widths to fixed increments determined by a Clock (hence Digital PWM).
• Note that FM is Analogue (no discrete steps or clocks) although it has always been called Digital (purely/entirely Marketing) and DSD is Digital where the Pulses are bound strictly to a rigid Clock frequency.

Nonetheless, what is delivered to the Speakers is Analogue although not all the early ClassD amps filtered/LPF (Striped away) the PWM carrier....

 

[Keith Flanker]

The lack of analog inputs is the dealbreaker for most - I2S-only ties you to source boards like this one. 300 W on a QFN48 is ambitious; thermals will decide if it holds up. Interesting, but I'd wait for measurements.

 

[Raiju]

For a "Fully Digital Amplifier", how do the speakers decode the digital signal output of the amplifier?

I think you know exactly how that happens, we have to draw the line somewhere. A cd is digital information on an analog medium, I think it is ok call a cd digital and this amp too.

digital signal => analog signal (dac) => amplification of signal => speaker

digital signal => amplification of signal => analog signal (lc filter) => speaker

Makes sense to call first one analog amp and second one digital amp

Another way to look at is when the 1's and 0's are destroyed and call it analog from that point on.

A dac destroys the 1's and 0's, you would need an a/d converter to get 1's and 0's back again and they would never be 100% the same. ( yes yes I know information cannot be destroyed, but let's not go there)

A pwm signal that comes out of the ACM8816 still has the 1's and 0's in there, since the data is not destroyed it is not analog. The amplification happened digital.

 

[Buckeye Amps]

A pwm signal that comes out of the ACM8816 still has the 1's and 0's in there, since the data is not destroyed it is not analog. The amplification happened digital.

....but what is the final output that leaves the amp?

 

[Michael Kohler]

For a "Fully Digital Amplifier", how do the speakers decode the digital signal output of the amplifier?

About a year ago, I called the technical support at NAD Germany. My question to the employee was: NAD produced so-called FDA amplifiers for several years. But the current top NAD amplifiers have analog inputs and are Class D types. Why switch back to analog amplification?
I had once listened to one of the "Direct Digital" amplifiers and I thought it sounded really good, completely "neutral," no character of its own, subjectively as clear and pure as all high-quality FDAs, which I think you can hardly distinguish.

His answer was that he didn’t know either. He said, if I remember correctly, that with analog amplifiers you can actually sell more devices: DACs.

But he explained the NAD 'Direct Digital' amplifier technology to me with a metaphor. He said you could think of a fully digital amplifier like a mobile phone display: We humans look at it and see a nice picture. But in reality, we’re seeing a huge number of pixels. Because there are so many, we believe we’re seeing an analog image, but it’s really just lots of pixels.

For all readers: Here is the next obviously stupid manufacturer who dares to call this type of device "Digital." NAD.

https://nad.de/wp-content/uploads/2017/03/DirectDigital_WhitePaper.pdf

Seriously, I'm not a technician. I don't know. I BELIEVE I KNOW that such an FDA amplifier stays digital all the way to its output filter, then there's a coil and a capacitor and *BAM* the signal is suddenly analog. How that works, I don't know, but I'd like to understand it. Maybe someone can explain it in a way that's easy to get. I'm 100% sure it wouldn't just be interesting for me.

 

[Buckeye Amps]

Maybe someone can explain it in a way that's easy to get. I'm 100% sure it wouldn't just be interesting for me.

@NTK already gave a thorough explanation that you essentially dismissed

 

[Michael Kohler]

The lack of analog inputs is the dealbreaker for most - I2S-only ties you to source boards like this one. 300 W on a QFN48 is ambitious; thermals will decide if it holds up. Interesting, but I'd wait for measurements.

Hi, what I would like to understand is: What is a QFN48?

 

[Michael Kohler]

The digital amp claims comes up here from time to time. Again, this is easy, so easy that even AI doesn't make any mistakes
"PWM (Pulse Width Modulation) amplifiers, commonly known as Class D amplifiers, are fundamentally analog, not digital. While they are often mistakenly called "digital amplifiers" because they operate in an On/Off switching state, their switching timing is continuous and based on an analog representation of the signal."

In the amplifier type 'FDA', the amplification is in the digital domain. An FDA is not a Class D amplifier. The AI explanation you posted describes a Class D amplifier.

Out of the NAD Direct Digital Whitepaper:

"DirectDigital is the next step forward, eliminating all analogue circuits completely. With NAD DirectDigital the conversion from digital data stream to electrical waveform occurs right at the speaker output jack of the amplifier; up to that point the signal remains in the digital domain."

 

[Buckeye Amps]

In the amplifier type 'FDA', the amplification is in the digital domain. An FDA is not a Class D amplifier. The AI explanation you posted describes a Class D amplifier.

Out of the NAD Direct Digital Whitepaper:

"DirectDigital is the next step forward, eliminating all analogue circuits completely. With NAD DirectDigital the conversion from digital data stream to electrical waveform occurs right at the speaker output jack of the amplifier; up to that point the signal remains in the digital domain."

In the most simple terms, PWM is still occurring within the NAD DirectDigital amplifiers.

You can split hairs all you want, but in one way or another a digital to analog conversion still takes place even in a "Fully Digital Amplifier" before the signal leaves the amp to go to the speakers.

 

[Raiju]

In the most simple terms, PWM is still occurring within the NAD DirectDigital amplifiers.

You can split hairs all you want, but in one way or another a digital to analog conversion still takes place even in a "Fully Digital Amplifier" before the signal leaves the amp to go to the speakers.

Well they are different enough to deserve their own class. And honestly just calling them digital amps is the easiest way. What do you propose? calling them digital to analog amps and the rest analog to analog amps?

 

[Sokel]

All these amps are just as analog and just as digital as a DAC;s RCA outputs.
At the end of their case there must be voltage (small or large) to drive the next stage.

And that's as analog as it gets.

 

[Buckeye Amps]

Well they are different enough to deserve their own class. And honestly just calling them digital amps is the easiest way. What do you propose? calling them digital to analog amps and the rest analog to analog amps?

You can call them whatever you want really but it won't stop people from properly pointing out how they work.

 

[pma]

Well they are different enough to deserve their own class. And honestly just calling them digital amps is the easiest way. What do you propose? calling them digital to analog amps and the rest analog to analog amps?

Either you understand the difference or not. The difference is in the presence or absence of quantization. PWM transfers input analog voltage to output analog frequency, linearly, without quantization steps. Digital signal is charecterized by quantization with the resolution depending on number of bits used. No, PWM is not digital.

 

[Raiju]

Either you understand the difference or not. The difference is in the presence or absence of quantization. PWM transfers input analog voltage to output analog frequency, linearly, without quantization steps. Digital signal is charecterized by quantization with the resolution depending on number of bits used. No, PWM is not digital.

You cannot say that, it can be both. It depends if the quantization can still be extracted, if yes => digital, if not => analog.

 

[Mnyb]

Regarding this products it plays on the actual strengths of GaN , thermal management ! GaN power supplies seems common these days for the same reason .
Small and cool .

Inside for example a Bluetooth speakers it makes sense to use this kind of chip , no room inside and nowhere for heat to go .

For absolute fidelity neither Purify or Hypex have done anything yet for well stated reasons . maybe we see something in the PA realm eventually at really high power ?

 

[Sir Sanders Zingmore]

Sometimes I think the point of this “new” tech is to draw people into arguments/discussions about it. That creates noise and attention and distracts from the fact that as long as an amp is linear enough and powerful enough it could be called class banana for all I care.

 

[jasonhanjk]

My digital input amp have a built in DAC. Which IC claims not to have a DAC?

 

[Michael Kohler]

In the most simple terms, PWM is still occurring within the NAD DirectDigital amplifiers.

You can split hairs all you want, but in one way or another a digital to analog conversion still takes place even in a "Fully Digital Amplifier" before the signal leaves the amp to go to the speakers.

I'm not splitting hairs. I'm comparing amplifiers that boost music in the digital domain (the FDA, Full Digital Amplifiers) with amplifiers that do it in the analog domain (the Class D types). I've understood that by now.

Is it a disadvantage to stay in the digital domain as long as possible? No, first of all, signals in the digital domain are less prone to interference. So from a technical point of view, it's actually more likely better than worse, purely in terms of signal transmission.

I'm talking here also about the wiring of the devices, RCA analog stereo or XLR Balanced analog stereo versus digital RCA or optical or USB. The digital connections are better and usually cheaper. In any case, digital connections are not a disadvantage.

But also from a financial perspective. I don't need a DAC. The full digital amp IS the DAC at the same time.

 

[Buckeye Amps]

I've understood that by now.

:thumbs up: