Thanks for this explanation, Paul! I've had several different class D amps in the past and have finally returned to a conventional class AB bipolar transistor amp... the detail is much better overall than with any of the class D amps that I've had... and one of them was actually pretty good, but the detail in some instruments never came through like it does with a conventional AB amp. Thanks again for the great video! Cheers!
letter D used to designate an amplifier class is simply the next letter after C and, although occasionally used as such, does not stand for digital. Class-D and class-E amplifiers are sometimes mistakenly described as "digital" because the output waveform superficially resembles a pulse-train of digital symbols, but a class-D amplifier merely converts an input waveform into a continuously pulse-width modulated analog signal. (A digital waveform would be pulse-code modulated.)
It seems that Class D amps and Direct Digital Amps are interesting technical innovations with the objective of lowering the cost of building power amps. Their challenge is to perfect a very complicated circuit to sound as good as the old fashioned class A or AB amps while making it small, cheap and efficient. I am happy that this may make great audio accessible to a wider audience, use less power, and fit into ever shrinking personal spaces. OTOH, I am happy to use giant monoblocs with vacuum tubes and lots of big nasty toroidal transformers that fit in my big detached home. The 350W per channel heat dissipation and the red glow of the tubes is like sitting by a fireplace while listening to great music. And I will drink a glass of fine wine made by century old vintners and not synthesized by AI. I hope that after I am gone someone recycles the copper and enjoys living in one of the tiny condos they build where my house used to be. Just say I feel lucky to have been alive while all this was possible.
To me a digital modulator looks like an other extra DA converter in the audio chain or an I wrong? You doing a great job in explaining all that stuff ;)
I have a pair of Danish 4 ways active speakers with digital amps (also referered to PowerDACs) inside and they sound amazing. Amazing value for the buck. The entire electronics block incl. the WISAmodule inside the speaker is just 2.5 by 2.5”. I spoke to the designer and he said that what is needed when the speaker and the amp is designed for each other.
Im not an engineer but i am an audio enthusiast, Im trying to get my head round this, so, does the signal carry the sound or is the signal the sound? Great explanation thank you
Does it mean that you go from a digital source directly to the amplification? So without a DAC? I mean are the DAC and the amplifier integrated? Or is an analogue signal digitally converted to steer the amplification? Or is there a A to D converter for the analogue signals?
Yeah… I think this is where Paul was going at the end of the video. The current situation when using a DAC feeding a separate class D amp is like: digital source → DAC → analog signal → ADC conversion to digital PCM in class D amp → amplification DAC in class D amp → analog → speakers. By combining the DAC and class D amplification, it becomes possible to eliminate one analog → digital → analog round trip in the signal path.
@@CraigArnolduk Yeah… I guess it is arguable if the last stage of class D amplification is considered or not a digital to analog conversion. (Confusingly I also typed PCM above when I meant PWM… sorry.) My point was, though, the traditional way of connecting a DAC’s analog output to a class D amplifier results in an unnecessary ADC operation in the signal path. With the assumption that for highest fidelity, converting to and from digital should be minimized, this hypothetical digital class D amplifier described by Paul would minimize the number of these analog/digital domain transitions. Since the first stage of class D amplification is an analog to digital PWM conversion, this could easily be avoided if the class D amplifier was simply fed a digital signal rather than an analog one, like: digital (PCM) → amp → math converting PCM to PWM → amplified PWM analog → speakers, or even simpler when source material is DSD, like: digital (DSD/PWM) → amp → amplified PWM analog → speakers. (as described, strikingly similar to the PS Audio DirectStream DACs). Is anyone making such a digital input class D amplifier currently?
@@jerry6789 IMHO he usually got confused when using it. These are off the cuff discussions and he does better when he doesn't document the thought process.
As far as I understand the Technics SU-R1000 is such a direct digital amp. Sure it doesn't sound bad at all, but I heard it at my dealer's several times with different speakers. The sound is too dry and analytical for my taste, kinda "papery" if that makes sense. I much prefer classic A/B or A amps, for now...
Thank you Paul, this was a great explanation of a technical subtlety I wasn’t aware of. Purely by chance I came across an in-depth explanation of some of these principles and how they are used in trains in this video: ua-cam.com/video/IRJIJPTUXXE/v-deo.htmlsi=RBI9XRfXN5F7O58s This is for trains’ AC motors, but the principle is the same, just using a rotary motor rather than the linear motor in a speaker. At 13:45, you can hear an ÖBB 1016 making a musical scale as it pulls away, seems the Siemens engineers have something of a sense of humour.
Class D, (PWM) a continuous average voltage value based the on and off pulse widths. If the off pulse is 10% of the cycle, and the on pulse is 90% then the average value of these two is 90% of the amplifier’s full voltage capability. A 50% “0” and a 50% “1” has an average of 0V with a Class A amplifier. However, let’s be clear, Class D was invented for one reason, thermal / power efficiency. It was not invented as a low distortion, low IMD, low anything else. It has undergone a huge amount of work in order to make it acceptable for HiFi use, but the main reason behind it is the high thermal efficiency, it only requires a small heatsink, smaller PCB, smaller chassis, cheaper PSU, or put another way, its cheap.
Yes, it’s about more power in a smaller size with less losses at a lower cost. In the late 90s I was at B&O where we worked on class D with the aim of best possible sound quality and called it IcePower. I still have a pair of Beolab 1 prototype active speakers working well now for more than 25 years with class-D in my home theatre setup. I wouldn’t use them as audiophile speakers, but my KEF LS60 in my living room and KEF LSX II in my office room are really not bad for class-D. For my listening room I use class-A.
Is it a misnomer to call it "direct digital" when the last actual digital stages is before the PWM stage and integrator for the analog stage? Seems an actual direct digital would be the output of the DAC through lowpass filtering.
What you're saying is the case, "direct digital" describes a power-DAC. There's no active stage in it that isn't "digital," including the output/PDM stage, which is just a high power transistor doing the exact same switching that a 1-bit DAC would do. The integrator in a power-DAC is a digital integrator, unlike the analog "integrator" (output filter) in a class D.
My response to that question and your answer is, "how important is it really, to understand all that in order to just enjoy the music sound that is coming out of our speakers?" For me, not very.
I now know more than I did... And also realize how little i know!
Love the channel and the way you explain things!
Cheers from Nashville!
Yep, I'm an analog guy in a digital world.
I never heard of this one before
Thanks for this explanation, Paul! I've had several different class D amps in the past and have finally returned to a conventional class AB bipolar transistor amp... the detail is much better overall than with any of the class D amps that I've had... and one of them was actually pretty good, but the detail in some instruments never came through like it does with a conventional AB amp. Thanks again for the great video! Cheers!
letter D used to designate an amplifier class is simply the next letter after C and, although occasionally used as such, does not stand for digital. Class-D and class-E amplifiers are sometimes mistakenly described as "digital" because the output waveform superficially resembles a pulse-train of digital symbols, but a class-D amplifier merely converts an input waveform into a continuously pulse-width modulated analog signal. (A digital waveform would be pulse-code modulated.)
This explanation makes sense and now I know how something works that I did not know before watching this video.
It seems that Class D amps and Direct Digital Amps are interesting technical innovations with the objective of lowering the cost of building power amps. Their challenge is to perfect a very complicated circuit to sound as good as the old fashioned class A or AB amps while making it small, cheap and efficient. I am happy that this may make great audio accessible to a wider audience, use less power, and fit into ever shrinking personal spaces.
OTOH, I am happy to use giant monoblocs with vacuum tubes and lots of big nasty toroidal transformers that fit in my big detached home. The 350W per channel heat dissipation and the red glow of the tubes is like sitting by a fireplace while listening to great music. And I will drink a glass of fine wine made by century old vintners and not synthesized by AI.
I hope that after I am gone someone recycles the copper and enjoys living in one of the tiny condos they build where my house used to be. Just say I feel lucky to have been alive while all this was possible.
You have definitely educated me! Thanks for the answer, Paul. Rohit.
Very well explained, Paul. And thanks to my friend Rohit to ask this interesting question.
Paul, I finally get Pulse Width Modulation, thank you so much.
Wow, didn’t know. Would be interesting to see DSD’s relation to GAN.
Paul sir always try to explain things very well.
Great explanation
To me a digital modulator looks like an other extra DA converter in the audio chain or an I wrong? You doing a great job in explaining all that stuff ;)
And finally no matter what, a continuous analog signal is sent to the speakers.
I have a pair of Danish 4 ways active speakers with digital amps (also referered to PowerDACs) inside and they sound amazing. Amazing value for the buck. The entire electronics block incl. the WISAmodule inside the speaker is just 2.5 by 2.5”. I spoke to the designer and he said that what is needed when the speaker and the amp is designed for each other.
Im not an engineer but i am an audio enthusiast, Im trying to get my head round this, so, does the signal carry the sound or is the signal the sound? Great explanation thank you
Does it mean that you go from a digital source directly to the amplification? So without a DAC? I mean are the DAC and the amplifier integrated? Or is an analogue signal digitally converted to steer the amplification? Or is there a A to D converter for the analogue signals?
I like the idea of having the DAC and amplifier integrated. But it makes it impossible to match gear…
Yeah… I think this is where Paul was going at the end of the video. The current situation when using a DAC feeding a separate class D amp is like: digital source → DAC → analog signal → ADC conversion to digital PCM in class D amp → amplification DAC in class D amp → analog → speakers. By combining the DAC and class D amplification, it becomes possible to eliminate one analog → digital → analog round trip in the signal path.
Digital amplifiers don't have a DAC. That's part of the point.
@@CraigArnolduk Yeah… I guess it is arguable if the last stage of class D amplification is considered or not a digital to analog conversion. (Confusingly I also typed PCM above when I meant PWM… sorry.)
My point was, though, the traditional way of connecting a DAC’s analog output to a class D amplifier results in an unnecessary ADC operation in the signal path. With the assumption that for highest fidelity, converting to and from digital should be minimized, this hypothetical digital class D amplifier described by Paul would minimize the number of these analog/digital domain transitions. Since the first stage of class D amplification is an analog to digital PWM conversion, this could easily be avoided if the class D amplifier was simply fed a digital signal rather than an analog one, like: digital (PCM) → amp → math converting PCM to PWM → amplified PWM analog → speakers, or even simpler when source material is DSD, like: digital (DSD/PWM) → amp → amplified PWM analog → speakers. (as described, strikingly similar to the PS Audio DirectStream DACs).
Is anyone making such a digital input class D amplifier currently?
Someone buy that man a whiteboard. For anyone confused, Wikipedia does a good job explaining it.
He used to use one. But people complained that he got too technical when using it. Lol.
@@jerry6789 IMHO he usually got confused when using it. These are off the cuff discussions and he does better when he doesn't document the thought process.
thank you, we understand it now
Will PS Audio eventually switch over to mostly class D? That’s what I’m wondering.
As far as I understand the Technics SU-R1000 is such a direct digital amp.
Sure it doesn't sound bad at all, but I heard it at my dealer's several times with different speakers.
The sound is too dry and analytical for my taste, kinda "papery" if that makes sense. I much prefer classic A/B or A amps, for now...
Thank you Paul, this was a great explanation of a technical subtlety I wasn’t aware of. Purely by chance I came across an in-depth explanation of some of these principles and how they are used in trains in this video: ua-cam.com/video/IRJIJPTUXXE/v-deo.htmlsi=RBI9XRfXN5F7O58s
This is for trains’ AC motors, but the principle is the same, just using a rotary motor rather than the linear motor in a speaker. At 13:45, you can hear an ÖBB 1016 making a musical scale as it pulls away, seems the Siemens engineers have something of a sense of humour.
Class D, (PWM) a continuous average voltage value based the on and off pulse widths. If the off pulse is 10% of the cycle, and the on pulse is 90% then the average value of these two is 90% of the amplifier’s full voltage capability. A 50% “0” and a 50% “1” has an average of 0V with a Class A amplifier. However, let’s be clear, Class D was invented for one reason, thermal / power efficiency. It was not invented as a low distortion, low IMD, low anything else. It has undergone a huge amount of work in order to make it acceptable for HiFi use, but the main reason behind it is the high thermal efficiency, it only requires a small heatsink, smaller PCB, smaller chassis, cheaper PSU, or put another way, its cheap.
Yes, it’s about more power in a smaller size with less losses at a lower cost. In the late 90s I was at B&O where we worked on class D with the aim of best possible sound quality and called it IcePower. I still have a pair of Beolab 1 prototype active speakers working well now for more than 25 years with class-D in my home theatre setup. I wouldn’t use them as audiophile speakers, but my KEF LS60 in my living room and KEF LSX II in my office room are really not bad for class-D. For my listening room I use class-A.
You left out amperage. PWM sends amperage with voltage. Ohm's Law.
Class D is Analog. Why on earth doesn't he start with that. 😢
Who currently makes this type of amplification ?
There are many companies, the amplifiers are very small and do not heat up, but they sound like a summer garden.....
Lyndorf is one of the companies.
Lyngdorf TDAI-1120 is such an amplifier.
Panasonic if I’m not wrong
All new Technics amps are using it, also Lyngdorf
Is it a misnomer to call it "direct digital" when the last actual digital stages is before the PWM stage and integrator for the analog stage? Seems an actual direct digital would be the output of the DAC through lowpass filtering.
What you're saying is the case, "direct digital" describes a power-DAC. There's no active stage in it that isn't "digital," including the output/PDM stage, which is just a high power transistor doing the exact same switching that a 1-bit DAC would do. The integrator in a power-DAC is a digital integrator, unlike the analog "integrator" (output filter) in a class D.
My response to that question and your answer is, "how important is it really, to understand all that in order to just enjoy the music sound that is coming out of our speakers?" For me, not very.
Perhaps toss Class D completely and use some DSD approach. Though there might be issues with a 1KW @ 11.2Mz transmitter in your living room.
Combine that with a pure DSD DAC, and a real good filter, and the result will be criminal 👌😊
Just take a look at Lyngdorf
You must use a white board!!☹
Another case for a white board…
Direct digital is better? Is that what your ears tell you?