1. Things you need to know about the new ‘Conversations’ PM system:

    a) DO NOT REPLY TO THE NOTIFICATION EMAIL! I get them, not the intended recipient. I get a lot of them and I do not want them! It is just a notification, log into the site and reply from there.

    b) To delete old conversations use the ‘Leave conversation’ option. This is just delete by another name.
    Dismiss Notice

Anyone compared high end Class D power amps?

Discussion in 'audio' started by Thetiminator, Jan 7, 2020.

  1. martin clark

    martin clark pinko bodger

    Maybe so - but if stuff is sold/shilled on metrics and heavens know that's a default position of some tedious people & some tedious fora, let's be damn sure that the metrics are at least directly-comparable.

    (Yeah, I know, there are lies, damn lies, and statistics; and the next Circle down is 'amplifier performance measurements...')
  2. Emlin

    Emlin MQA Hater!

    Is it a religious site, because lots believe in that sort of nonsense too.
  3. Shadders

    Shadders pfm Member

    For those not aware, when the THD is measured for class D amplifier, then the filter applied to limit the harmonics is set at 22kHz. For class A/B it is usually set to 80kHz.

    The reason for this, is that there is so much "mush" on the output of the class D amplifier, that it affects the measuring equipment. A key manufacturer of measuring equipment is Audio Precision, and they implement a pre analyser filter for measuring class D amplifiers. Not required for class A/B.

    See slide 12 :

    So, for class D, the filter is set to 22kHz because it has a lot of noise (mush) on its output, and for class A/B the filter is set to 80kHz.

    The higher bandwidth filter lets more noise through. So when you see measurements for class A/B with a measurement bandwidth of 80kHz, the amplifier is at a severe disadvantage to class D since class D is using a measurement bandwidth of 22kHz.

    Dozey likes this.
  4. Alan Brown

    Alan Brown Registered LUser

  5. Cereal Killer

    Cereal Killer 432 Point5

    Here's some measured results of the Audionet DNA, the first iteration of the DNA 2.0.

    Distortion factor (THD+N, 10 watts @ 4 ohms): 0.0062 %
    IM distortions SMPTE (5 watts @ 4 ohms): 0.0075 %
    IM distortions CCIF (5 watts @ 4 ohms): 0.0008%

    S/N ratios:

    Unweighted SNR (- 20 kHz): -85.4 dB
    Unweighted SNR (- 250 kHz): -71.5 dB
    S/N ratio (A-weighted): -87.9 dB

    Upper limit frequency (-3dB / 10 W @ 4 ohms): > 185 kHz

    Channel deviation:
    0.034 dB
    Input impedance: 45.5 kohms
    DC output offset: < 0.5 mV

    Idea: lets make the nerdiest ever set of top-trump cards - HiFi specs :D:D:D
  6. Tony L

    Tony L Administrator

    Wrong, it just requires an electronic circuit where things such as resistance, capacitance and inductance have some influence, i.e. most audio circuits. If someone really can’t hear the difference between say NAC A5 and Kimber 8TC, which clearly measure vastly differently to one another, I’d argue either their system or levels of perception are so poor they shouldn’t really be commenting on audio at all!
    Dozey, Darren L, CJ14 and 2 others like this.
  7. Shadders

    Shadders pfm Member

    I disagree. If you complete a null test on a cable, what goes in, comes out the end exactly the same.

    For an amplifier, the null test will show the distortion, albeit very small for a good performing amplifier.

    For the cables you mention, there are differences in parameters, the Kimber 8TC seems to be predominantly capacitive, and the NAC from statement on PFM, is predominantly inductive (as a substitute for the missing output inductor in the amplifier). All these will do is modify the frequency response, and for the Kimber it may be causing some near instability in the amplifier, as the pF/m is rather high.

    I suppose if a cable is purposefully and unnecessarily designed to be extreme, it may cause problems which are perceived as differences in cable sound, when in fact it is causing a problem with the electronics. I suppose the amplifier manufacturers don't specify the range of cable parameters that provide the expected performance.

    From the tests i have seen in Hifi mags, such as the group tests of cables, so many are very similar in parameters (no extremes) then i fail to see how they can sound different. I have tried twin core + earth, QED 79strand, QED 42strand, and Van Damme Blue 2.5mm2, and no difference.

    My statement was for the many cables i have seen reviewed, with similar parameters, not the extremist types.

  8. Cereal Killer

    Cereal Killer 432 Point5

    Is that all the cables you have tried!?
  9. Shadders

    Shadders pfm Member

    Yes - why would i try a speaker cable that has been manufactured to an extreme specification which has no basis in performing as required ?

    As i said above, some cables may cause mild/occasional/spurious instability - and that is not what i want in my equipment.

  10. Cereal Killer

    Cereal Killer 432 Point5

    Thing you've answered your own question....

    IF VD-blue is the best cable you've tried then you clearly don't have any merit to criticise other cables
  11. Shadders

    Shadders pfm Member

    No, i have not. What i have tried are cables that are solid core, stranded 42 and 79, and the Van Damme (not QED), and not heard any difference between any of them.

    What this shows me is that in general, cables make no difference. Of course, my system may not be resolving enough, or my hearing is sub optimal.

  12. hifinutt

    hifinutt hifinutt

    anyway keeping to the subject of class d amps ............;)
    Cereal Killer likes this.
  13. Arkless Electronics

    Arkless Electronics Trade: Amp design and repairs.

    That may be your opinion (and an opinion is all it is), one on which many would disagree. The capacitance and inductance you mention have no effect at audio frequencies and at the levels found in typical cables. Now if these cables were being used as RF feeder to carry 100MHz radio signals to an aerial then the measurements would be crucial, but not at audio frequencies, where transmission line effects do not occur. Resistance will have an effect, which is why speaker cable needs to be thick. Make it suitably thick and no more problem. I'd argue that people suffering from such severe auditory hallucinations as to hear non existent differences shouldn't really be commenting on audio at all:D:)
  14. Cereal Killer

    Cereal Killer 432 Point5

    And do you absolutely know whats its affect is on the audio frequencies themselves?
  15. Arkless Electronics

    Arkless Electronics Trade: Amp design and repairs.

    Yes. It's very basic stuff.
  16. Cereal Killer

    Cereal Killer 432 Point5

    Is that your opinion or do you measurements to prove it?
  17. Cereal Killer

    Cereal Killer 432 Point5

    Agreed, here's some more spec regarding the DNA 2.0

    2 x 165 Watt into 8 Ohm
    2 x 280 Watt into 4 Ohm
    2 x 440 Watt into 2 Ohm

    Typ. 1000 at 100 Hz

    1 – 500,000 Hz


    SNR:> 103dB @ 1kHz
  18. Cereal Killer

    Cereal Killer 432 Point5

    Black by Bel Canto ACI 600 Manual
    24bit Data to 192ks/s: AES, SPDIF, TOSLINK
    24bit Data to 192ks/s, MQA, and DSD64: 10/100 Ethernet
    24bits to 384ks/s, MQA, and DSD64/128 (DoP): USB2 Audio
    Low Level Outputs:
    Line Level Analog: 4.5Vrms with Bass Management
    Headphone: 4.5Vrms maximum, 32ohm minimum load
    MM/MC Input:
    MM: 2.5mV to 5mV; 47K ohms
    MC: 0.25mV to 0.5mV; 50, 100, 500, 1k
    RIAA Accuracy: +/- 0.25dB, 50Hz-15kHz
    THD+N: <0.01% 1kHz A-Weighted
    SNR: >70dB A-Weighted
    Line Inputs:
    Maximum Input: 2.2 Vrms RCA
    Input Impedance: 10k ohms RCA
    THD+N: 0.003%, 1kHz
    Dynamic Range: 110dB, A-weighted 20Hz-20kHz
    Loudspeaker Output:
    Maximum Power Output: 600W-4ohm, 300W-8ohm
    Minimum Load: 2 ohms
    Peak Output Current: 27 amperes
    Frequency Response: -3 dB 0.5Hz-50kHz, all loads
  19. John Phillips

    John Phillips pfm Member

    A paper was published in JAES in 1991 (link - the PDF can be viewed) showing the degree of non-flatness in the amplifier - cable - speaker interface for various cables (Kimber types included), amplifiers and loudspeakers.

    Loudspeaker cables are among the least understood yet mandatory components of an audio system. How cables work and interact with loudspeaker and amplifier is often based more on presumption and speculation than on fact. The literature on loudspeaker cable behavior and effects is minimal. Measurements were made with 12 cables covering a variety of geometries, gauges, and types. The measured data indicate distinct differences among the cables as frequency-dependent impedance, subtle response variations with loudspeakers, and reactance interactions between amplifier, cable, and loudspeaker. In some cases the effects of the amplifier overwhelm the cable's effects. Mathematical models that provide insight into the interaction mechanisms were constructed and compared to the measured data.

    A disagreement followed in the pages of Audio magazine between the author (Fred E. Davis) and Edgar Villchur (founder of Acoustic Research) about whether the effects were audible or not. Glossing over the details, Davis thought they could reach to audible levels but Villchur disagreed.
  20. Shadders

    Shadders pfm Member

    Whatever the specification stated, you have to remember that the THD for class D is specified with a 22kHz measurement bandwidth. As per the link of the slide show earlier, a class D amplifier can cause issues with the measuring equipment above the 22kHz due to the mush (noise) it puts out.

    When you see a class A/B amplifier measurement, it is in general measured with a bandwidth of 80kHz.

    You are not being conned with the class D THD specification, but class A/B is getting a raw deal since its measurement for THD is harder. Class D can not compete.

    The point of my initial posting in this thread, is to highlight what class D is in simplistic terms, and to contrast this with differences heard in cables.

    If someone can hear differences in cables, where those differences (for the non-extremist version) are so very subtle in frequency response, and yet, knowing that class D output stage is a rectangular wave which goes through a coil, how can the information on class D not get you to mull over whether your hearing is as astute as you believe it is ??


Share This Page


  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice