Class ‘D’ Power Output Distribution

Discussion in 'Audio Hardware' started by Khorn, Feb 7, 2021.

  1. Khorn

    Khorn Dynagrunt Thread Starter

    Are there inherent differences between the various class D modules in power output (drive) vs frequency? If so which are best in that respect?
     
  2. Agitater

    Agitater Forum Resident

    Location:
    Toronto
    There are inherent differences. In general, compared to class A and A/B, class D is more efficient at all frequencies and at all power output levels. However, within the parameters of a specific class D design it will be generally less efficient at lower power output levels, with efficiency improving dramatically as power output is increased. That seems to be true no matter what the output load (impedance) happens to be as power output increases to the speakers. That’s what it looks like from a power efficiency standpoint.

    Texas Instruments published an excellent class D technical analysis (using TI’s own class D amps of course) in 2020, but I was unable to find it in a quick search. IIRC, the piece was called “choosing the right class D amplifier” or “how to choose the right class D amp” or something very similar to that. It was a technically detailed, highly informative and well-written document. I should have bookmarked it.

    Are you interested the effects on power output of the class D amp designer’s chosen switching frequency center, or on power output vs the frequency of music being played? If it’s the latter, that might be very hard to determine because of the reality that most recorded music contains a variety of simultaneous and constantly changing frequencies at constantly changing amplitudes.
     
    Saint Johnny, Khorn and Archimago like this.
  3. Black Elk

    Black Elk Music Lover

    Location:
    Bay Area, U.S.A.
    This?

    https://www.ti.com/lit/wp/sloa290/sloa290.pdf
     
    Ingenieur and Khorn like this.
  4. Khorn

    Khorn Dynagrunt Thread Starter

    Thanks This will help a lot. I’m starting to look at new amplification after I receive and settle in my new preamp. I’m most interested in class D. I want to take my time and find a good medium priced amplifier whatever medium priced means in today’s audio market.
     
  5. Richard Austen

    Richard Austen Forum Resident

    Location:
    Hong Kong
    I like my mAmp monoblocks from Wyred4Sound - Speakers are easy to drive AN K speakers that need only 7 watts. The mAmps are 250 watts 8 ohm 430 watts 4 ohm. I have no "difficult to drive" speakers at the moment. Nicely made little amps and not a crazy sum of money.
     
    Khorn, sound chaser and macster like this.
  6. Ingenieur

    Ingenieur Going with the flow...

    Location:
    PA
    Don't all amplifiers (power devices, mech or elec) efficiency increase with output?

    eff = (output - losses)/output
    Losses are relatively constant
    Increasing a much lower rate than output

    doesn't the eff of a Class D come from the fact it is switching, not continuous, duty cycle

    in the TI paper the eff curve eff hits 70% at 1% output (based on 200 W max).
     
    The Pinhead, Mike-48 and Khorn like this.
  7. macster

    macster Forum Resident

    Location:
    San Diego, Ca. USA
    Richard, wow this was a shock. But I agree with you.


    M~
     
    Khorn likes this.
  8. Agitater

    Agitater Forum Resident

    Location:
    Toronto
    Class D generally rises to a peak power conversion efficiency of 90+%. The most efficient class A/B designs generally can’t make much past 75% at best. Additionally, class D has a significantly higher baseline efficiency than A or A/B even when comparing baselines beginning at output levels as miniscule as 0.5 mw. That’s all well-known. My own measurements have been admittedly limited, but generally bear out those efficiency numbers, the much better class D baseline efficiency, and so on.

    There are a lot of superb class D power amps and integrated amps on the market right now. As far back as five years ago, there were some great-sounding class D amps and the situation is exponentially better today. So I think @Khorn really has a lot of very satisfying auditions to look forward to.
     
    Khorn and Ingenieur like this.
  9. oldhand

    oldhand New Member

    Location:
    Inverness
    You are correct, there are some excellent Class D technologies out there. While there is a major efficiency advantage, meaning the PSU and heatsinking can be much reduced in comparison to Class A/B. There is more to the story, in that performance needs discussing.

    Class D actually falls into 2 categories.

    Analogue modulation, with a continuously variable PWM width. This operates directly on an analogue signal and needs preceding pre-amp and DAC. Various implementations are now available as modules, which give good performance (comparable with best linear technology)

    Digital modulation, with a noise shaped/quantises PWM width. This operates on a digital signal so could be considered to be an amplifying DAC, with the volume control built-in. It is difficult to achieve high performance, unless effective feedback is part of the architecture. Few have achieved this, but it is available. For the best case, performance is > 115dB SNR, THD <0.002%, DF >>1000. The key thing to remember with this topology, is that it a much larger part of the system.. Since such systems offer good jitter tolerance at the input, all digital sources can be considered perfect. The volume control is done in the digital domain, which again is perfect (It's just a multiplication), so that means you have the world's best pre-amp... It never adds distortion or noise, has perfect linearity and repeatability.

    So there is the opportunity to create an excellent system, as follows:

    Any digital source --- High performance digital amp --- Good quality speakers of your choice.

    I saw a post showing just this, using Sonos Amp and some large Tannoy speakers.
     
    head_unit and Khorn like this.
  10. Ingenieur

    Ingenieur Going with the flow...

    Location:
    PA
    Maximum theoretical efficiency:
    Class A 25%, if choke/xfmr coupled 50%
    Class B 78.5%
    Class AB between A and B, typically 60-65%
    Class D > 90%

    That is only the signal conversion losses, not the power supply, etc.

    Class D has come a long way and is in the big leagues now. Not so much for eff, but sound, cost and packaging.
     
    The Pinhead and Khorn like this.
  11. vinylbuff

    vinylbuff Forum Resident

    Location:
    North Port Florida
    I used a pair of Wyred 4 Sound SX-1000's for 9 years and enjoyed them immensely, but then all of a sudden I got tired of them and sold them. I immediately went straight to the Wyred 4 Sound website, conversed with E.J. Sarmento just long enough to order a pair of the SX-1000R's in Midnight Sparkle finish and the honeymoon continues......:love:
    They power a pair of Legacy Convergence 4 ohm 98db efficient tower speakers to ear threatening volume levels if I so desire....(which I don't) and I'm a happy camper.
     
  12. vinylbuff

    vinylbuff Forum Resident

    Location:
    North Port Florida
    I consider the W4S SX-1000r mono amps Big League. Bascom H. King designed the boards so quality of build and sound are a given. I use mine in balanced mode and the noise floor is non existent. And having 40+ amps of current and almost 1200 watts of power into 4 ohms doesn't hurt either....
     
    sound chaser, 4xoddic and Khorn like this.
  13. levimax

    levimax Forum Resident

    Location:
    California
    Currently Purfi is making some of the best Class D amp modules after Bruno Putzeys came over from Hypex and are top of class at output vs frequency consistency although not the same as a class AB .... but do you really need "continuous full output" at 20 Khz? Purifi Audio
     
    MaxBuck and Khorn like this.
  14. dcottrell6

    dcottrell6 Forum Resident

    Location:
    Eastampton, NJ
    I have the Wyred 4 Sound ST500 MKII using the ICE power modules.
    It does a great job with the Maggie .7's.
    It replaced the Parasound A23.
     
    vinylbuff, sound chaser and Khorn like this.
  15. head_unit

    head_unit Senior Member

    Location:
    Los Angeles CA USA
    Are you planning to pick an amp based on the modules? Because while yes some have achieved better performance of noise etc, it is like DAC chips-the module is just one part. The analog circuitry and especially the power supply have a huge effect, not to mention the chassis. Class D may be more efficient but still need heat sink!

    So far on our quest the ATI Class-D AT5XXNC series seem the best performance for not crazy money. Stipulation: with significantly more power than an AVR. Going to do some listening Saturday.
     
    Khorn likes this.
  16. oldhand

    oldhand New Member

    Location:
    Inverness
    Just read the datasheet for the purify amp and performance is excellent in all respects. In my view all aspects are at least as good as the best Class A or Class AB amplifiers.
    Interesting to learn if anyone thinks differently and why?

    Obviously, you don't need continuous full output power at 20kHz. But amps must be specified across full audio band to conform to FTC requirements and even without this need people would roundly criticise brands who didn't.

    As an aside, high frequency high power is easier in terms of PSU demands than low frequency high power.
     
    Khorn likes this.
  17. Ntotrar

    Ntotrar Forum Sciolist

    Location:
    Tri-Cities TN
    Very pleased with my Marantz 30.
     
    Khorn likes this.
  18. levimax

    levimax Forum Resident

    Location:
    California
    No one is doing the FTC thing anymore and I don't think it is even enforceable. I am not sure of the technical reasons why but most Class D amps, including Purfi , do not put out full continuous power at high frequency. Here is a test of an ICE module amp that puts out only 4o watts @ 20 Khz before shutting down vs over 100 watts @ 20 Hz. ICEpower 200AC & 200ASC Review (Class D Amp) All the tests of Class D that I have seen are the same to a greater or lesser degree. I don't think it matters, I think Class D amps are great and the future, but I mention it because the OP asked about it.
     
    Khorn likes this.
  19. oldhand

    oldhand New Member

    Location:
    Inverness
    Not sure about your view on FTC power rules at the moment. Here's one that certainly does obey and do full power at 20kHz.

    M32 - NAD Electronics

    Continuous power rating rules do lead to over-engineering in terms of heatsinking, as a sinewave scenario is far beyond the worst case real-life situation, which would be perhaps 1/3 in terms of mean output power.

    It would be much better to have a well thought out burst-test scenario, but I can imagine a challenge in getting agreement on a spec.
     
  20. levimax

    levimax Forum Resident

    Location:
    California
    I was old school when it came to amps and specs .... the traditional over engineered Class AB amps that were heavy and could output full power continuously from 20 Hz to 20 Khz seemed like a show of good engineering and quality. Now I am re-thinking that. Music is nothing like a continuous test signal and if an amp can be engineered to deliver higher power where needed (bass) and less where not needed (HF) and produce high peak power for transients while being lower distortion, lighter, cheaper, and more efficient even if they don't do the same continuous power tricks of the past I think that actually makes for a "better" amp.

    I agree it would be easier for the consumer if there was a "well though out" amp rating system but not sure what it would be and who would agree with it, then again maybe we are better off without one. The original FTC requirements caused manufacturers to design to pass an arbitrary test that had little to do with playing real music and cost consumers a lot of money on wasted oversize heatsinks and limited peak power. No matter what "buyer beware" has been and still is the way to approach amplifiers. Fortunately there is information available, both objective and subjective, if one wants to spend the time researching.
     
    Khorn likes this.

Share This Page

molar-endocrine