cdp clock??

Discussion in 'Hi-Fi and General Audio' started by DavidF, Jan 14, 2007.

  1. DavidF

    DavidF

    Joined:
    Mar 31, 2006
    Messages:
    3,296
    Likes Received:
    1
    Location:
    Shropshire
    Hello,

    Can someone explain what the "clock" in a cdp does+what its function is?

    Tried google...didn't really turn up much.

    Thanks,



    David
     
    DavidF, Jan 14, 2007
    #1
  2. DavidF

    julian2002 Muper Soderator

    Joined:
    Jun 19, 2003
    Messages:
    5,094
    Likes Received:
    1
    Location:
    Bedfordshire
    digital audio relies on the data being read off the 'media' (the cd) at a known rate. the clock makes sure that this data is read and fed to the audio electronics (dac, filters, etc.) at the exact moment necessary to provide an accurate representation of the original.
    in cd's case a sample must be read off the cd and passed to the dac 44100 times every second. very simple there is a bit of crystal that when it has a current applied to it vibrates at that frequency (this isn;t exactly the case but will do as an explanation - it actually vibrates at a much higher frequency and is divided down to provide the correct frequency) this is then used to move the data read from the cd to the dac as it's needed. this crystal is therefor a kind of clock providing the timing that is essential to making the cd player work.
    this 'clock' is not however 100% accurate and can 'wander' slightly due to electrical noise, temperature variation or just through being cheap, this simplistically is 'jitter' - the clock jitters arround it's ideal frequency. understand though that this jitter is measured in picoseconds (a proper english billionth of a second i.e. a million millionth of a second, where cdp clock vibration speeds are usually measured in megahertz or millions of vibrations per second - giving you some insight into why some say jitter is just marketing bollocks).
    clock upgrades provide better crystals with power filters and other refinements to make them ore accurate and less prone to jitter.
    hopefully i've got that vaguely right - it's a bit simplified but should start you off.
     
    julian2002, Jan 14, 2007
    #2
  3. DavidF

    Tenson Moderator

    Joined:
    Nov 12, 2003
    Messages:
    5,947
    Likes Received:
    1
    Location:
    Kent, UK
    Tenson, Jan 14, 2007
    #3
  4. DavidF

    DavidF

    Joined:
    Mar 31, 2006
    Messages:
    3,296
    Likes Received:
    1
    Location:
    Shropshire
    ok...interesting; thanks very much.

    Havent full read through Simons link yet but is the clock housed inthe transport...for some reason imagine it there, not the pcb?



    thanks again
     
    DavidF, Jan 14, 2007
    #4
  5. DavidF

    felix part-time Horta

    Joined:
    Dec 13, 2003
    Messages:
    757
    Likes Received:
    1
    Location:
    dead
    Usually, yes. There's not necessarily a lot to see - a crystal (in a small metal can), a tiny pair of small value ceramic caps and a logic gate (often part of the dac or digital filter chip) are all it takes to make an oscillator. Julian hass given a nice summary of what it does and the typical problems.

    The significance of the 'the clock' is simply that, in digital audio, the audio information is effectively stored as a series of fixed values to be dished-out at constant intervals. This constant 'clock signal' simply measures out equal intervals to the chip(s) doing the converting. Due to the potential sources for error(s), the clock is ideally generated right at the DAC chip and distributed elsewhere as required. In fact, for mundane reasons, the oscillator is often associated with the disc-reading mechanism control circuitry instead, with the timing signal piped across the board to the dac (in a one-box cd player) or munged together with the data and sent (via SPDIF usually) to an offboard dac. This last gives more potential for variance in the encoding and decoding...and noise sensitivities... hence differences between transports!

    A very few standalone dacs do it right and have their own master clock which the transport is then 'slaved' to; DPA and Pink Triangle are some of the original proponents of this method. Another current alternative is to have a local clock in the dac and simply ignore the timing data coming in from the transport ('asynchronous reclocking') - though there can still be problems with this, especially when combined with resampling/upsampling.

    PS - Simon's link is a joke ;)
     
    felix, Jan 15, 2007
    #5
  6. DavidF

    DavidF

    Joined:
    Mar 31, 2006
    Messages:
    3,296
    Likes Received:
    1
    Location:
    Shropshire

    ahh.....ok
     
    DavidF, Jan 15, 2007
    #6
Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.