SONY CDP-101 : CX20017 dac improvement

Discussion in 'Digital Sources' started by cizeta, Jan 18, 2018.

  1. cizeta

    cizeta Member

    Messages:
    78
    Location:
    Milan - ITALY
    1. This post is to spread the possibility of upgrading the glorious CDP-101 replacing the old DAC CX20017 with a double converter produced by Philips and that works with the standard Sony signal bus (not I2S). The new DAC is the
      TDA1311A and has only 8 pins. A simple base has been created that adapts the feet of the new dac to the preexisting CX20017.
      The attached schematic translates the pins of the old dac with the correspondents of the new one.
      It is a continuous calibration dac (CC-DAC) and it has already voltage output instead of current like the old one. After replacement, it is therefore necessary to remove IC508 and IC509 and bridge the pins 6 and 1 of the empty holes of IC508 and a bridge between pins 2 and 7 of the empty holes of IC509. Keep in mind that the output has the absolute phase reversed, but this is not a problem (many readers do). If desired, simply invert the signal by resetting the two ic 508 and 509 and modifying the circuit to eliminate this inversion. The integrated IC512 must be removed and the resistance R541 must be modified with 15 Kohm to make it equal to that of the left channel. With this double DAC the 11,34 us group delay (audible by expert ear or in MONO play) disappears in the fixtures with a single converter and the sound is updated bringing it closer to the much appreciated one of the Philips DAC. No more audio with harshness ... the sound becomes warm and velvety, even if the upper part of the spectrum is always present naturally.[​IMG][​IMG][​IMG][​IMG][​IMG][​IMG]

    2. If you want to try, you can also remove the anti-aliasing filters F501 and F502 (removing them) by connecting directly pin 18 dac (out L) to pin 6 of F501 (and the same for the right channel pin 17 dac to pin 6 of F502) to have the unfiltered NOS version of Sony CDP-101, but I do not recommend it, because the unfiltered signals can be dangerous for the amplifier connected to the output.[​IMG]
     
    Last edited: Jan 19, 2018
    FileFixer and eizner23 like this.

     

    Please register to disable this ad.

  2. cizeta

    cizeta Member

    Messages:
    78
    Location:
    Milan - ITALY
  3. restorer-john

    restorer-john Addicted Member

    Messages:
    5,372
    Location:
    Australia
    Interesting thing to do but I cannot understand why on earth anyone would replace the CX20017 with a cheap, nasty DAC with really poor specifications. And pulling out the LPFs? Seriously this butchering of perfectly good machines has to stop.

    The TDA1311A's THD@0dB is 0.04%, @ -60dB it is 36%! That's not a misprint.

    The standard CDP-101 (I have six of them) measures 0.002% @ 0dBFS 1KHz. Why on earth mess with a classic and deliberately make it worse? This is not an 'improvement' in any way.

    Heitaro Nakajima (R.I.P.) would not be impressed. :(
     
    Last edited: Jan 18, 2018
    roger2 likes this.
  4. restorer-john

    restorer-john Addicted Member

    Messages:
    5,372
    Location:
    Australia
    I tell you what, post an oscillogram (scope shot) of a single impulse (1 sample 0dB), a few square waves and a few IMD high frequency traces of your 'improved' CDP-101 and I'll do the same with a stock CDP-101.

    Do you have a test CD? If you don't, PM me and I can email you the WAV files and you can burn a test disc to do this.

    I assume you have a scope based on your avatar of an eye pattern.
     
  5. cizeta

    cizeta Member

    Messages:
    78
    Location:
    Milan - ITALY
    My sample has a harmonic distortion of 0.02% at 1KHz 0dB.
    The group delay is, for me, a type of unacceptable distortion.
    For example, the TDA1543 is also an economical DAC, but is considered an excellent DAC. I kept the original filtering, but I buffered and reversed the electrical phase of the DAC output signal, to get a positive impulse response as a standard. The sound is much better than the CX20017 and this must be heard, I can not communicate with a writing. Next step will be digital filtering with 4x oversampling (SM5807EP). The CX20017 is a slow DAC, had been supplanted by the CX20152, of better quality.
     
  6. cizeta

    cizeta Member

    Messages:
    78
    Location:
    Milan - ITALY
    My cd test is "CBS CD-1"
     

     

    Please register to disable this ad.

  7. restorer-john

    restorer-john Addicted Member

    Messages:
    5,372
    Location:
    Australia
    Cool, the CBS disc has pretty much what you need. Let's see the scope shots. :)

    I'll post mine when I get back to the bench tonight.
     
  8. cizeta

    cizeta Member

    Messages:
    78
    Location:
    Milan - ITALY
    Ok. trace 10: 19997 2018-01-19 00.07.24.jpg Hz 0dB group delay
     
  9. cizeta

    cizeta Member

    Messages:
    78
    Location:
    Milan - ITALY
    Track 16: 2018-01-19 00.09.50.jpg Square wave 1 kHz 0dB
    The elongations are caused by the analog MURATA anti-alias filter on the CDP 101
     
    Last edited: Jan 18, 2018
  10. cizeta

    cizeta Member

    Messages:
    78
    Location:
    Milan - ITALY
    Send me your sinusoidal 20kHz and 1KHz square 0dB both channel,please.
     
  11. onwardjames

    onwardjames Hoardimus Maximus Subscriber

    Messages:
    15,384
    Location:
    Kentucky
    I have no freaking idea what you guys are saying, but I'm always amazed at the ingenuity. Keep rocking!
     
    mstrane21 likes this.

     

    Please register to disable this ad.

  12. cizeta

    cizeta Member

    Messages:
    78
    Location:
    Milan - ITALY
    Impulsive signal before the anti-image filter (up) and signal after filtering (at the bottom) , that is, what we really hear at the output of CDP-101....
    2018-01-19 00.37.04.jpg
     
    Last edited: Jan 18, 2018
  13. cizeta

    cizeta Member

    Messages:
    78
    Location:
    Milan - ITALY
    Hi,onwardjames!
     
  14. cizeta

    cizeta Member

    Messages:
    78
    Location:
    Milan - ITALY
    Your shots will be the same as mine, except for the group delay that is absent from me.
    Try to listen to MONO one of your CDP-101 and you will feel a great disaster ...
    I worked as a technician at SONY for ten years in the 80s, I know these devices as my pockets .. for example, Sony tried to attempt to limit this group delay by modifying the value of a delayed channel time constant ( the right) ... if you look at the wiring diagram you will notice that the R530 (I / V left channel) is 15Kohm, while the symmetrical R641 (I / V right channel) is 16Kohm. This resulted in an output voltage that is not the same for the two channels .. no one has ever noticed this?
     
    Last edited: Jan 18, 2018
  15. cizeta

    cizeta Member

    Messages:
    78
    Location:
    Milan - ITALY
    However, the CDP-101 D / A conversion section is obsolete and must be modified.
     
    Last edited: Jan 18, 2018
  16. FileFixer

    FileFixer Well-Known Member

    Messages:
    545
    Location:
    Zagreb/Croatia

     

    Please register to disable this ad.

  17. cizeta

    cizeta Member

    Messages:
    78
    Location:
    Milan - ITALY
    ?
     
    Last edited: Jan 19, 2018
  18. cizeta

    cizeta Member

    Messages:
    78
    Location:
    Milan - ITALY
    Last edited: Jan 19, 2018
  19. cizeta

    cizeta Member

    Messages:
    78
    Location:
    Milan - ITALY
    NAKAMICHI MB-4S: LC78840M Digital filter - TDA1311A Dac

    Naka dac.jpg

    TDA1311A DAC: PROTON AC424 - NAKAMICHI MB-4S
     
    Last edited: Jan 19, 2018
    FileFixer likes this.
  20. FileFixer

    FileFixer Well-Known Member

    Messages:
    545
    Location:
    Zagreb/Croatia
    Why Nak use these three resistors between digital filter and DAC (R801, R802, R803)?
     

Share This Page