1. This forum section is a read-only archive which contains old newsgroup posts. If you wish to post a query, please do so in one of our main forum sections (here). This way you will get a faster, better response from the members on Motherboard Point.

PLL and noise questions

Discussion in 'Embedded' started by Jay, Jul 8, 2003.

  1. Jay

    Jay Guest

    Hey gurus,

    I'm trying to understand how PLLs work, so I have some (basic?)
    questions...

    1. When a PLL has "locked" to an incoming siqnalg, is the phase error
    between the output signal and the input signal 0 deg/radians or is the
    phase offset just fixed?

    i.e. Assuming PLL lock condition, if I have the input to a PLL up on a
    scope and I bring the output up on a scope, will the rising and falling
    edges of both waveforms be aligned up, or will I see some offset between
    rising(and falling) of input and output?

    2. Related to above: If the phase error is fixed and non-zero, is there
    any way (with PLLs/DPLLs or DLL?) to time-align the waveforms so they
    are close to perfectly in phase?

    3. If I understand correctly, PLLs can track phase variations AND
    frequency variations (dTheta/dT) just by adjusting a VCO(with a control-
    loop)?

    4. I was looking at specs for noise and I see it's specified in

    nano-Volts/(square_root(Frequency)) along with another units, dBFS.

    Is there any article or website, or section of a book that actually
    explains these units, how they came about and how they're used? I have
    several engineering books, but I don't ever see discussion of noise,
    especially with those units.

    And no, this isn't for homework!

    -- With warm regards

    Jay.
     
    Jay, Jul 8, 2003
    #1
    1. Advertisements

  2. Jay

    Robert Scott Guest

    There are many kinds of PLLs and each kind has a different locked
    condition. They are generally classified by the type of phase
    detector used. If the phase detector is the type that produces a
    fixed voltage for a fixed phase error, then the PLL will lock with a
    static offset sufficient to maintain the required control voltage to
    the VCO. If the phase detector is the type that integrates the phase
    error, then even a small phase error will eventually accumulate the
    required integral of VCO feedback voltage, so that kind of PLL will
    lock with zero phase error.


    -Robert Scott
    Ypsilanti, Michigan
    (Reply through newsgroups, not by direct e-mail, as automatic reply address is fake.)
     
    Robert Scott, Jul 9, 2003
    #2
    1. Advertisements

  3. Jay

    Tauno Voipio Guest

    The classic 4046 in nearly any CMOS series has detectors for both types. The
    multiplier (xor) detector locks at 90 degrees and the digital detector locks
    at zero. It's an easy thin to start with.

    HTH

    Tauno Voipio
    tauno voipio @ iki fi
     
    Tauno Voipio, Jul 9, 2003
    #3
  4. The phase is usually 90degrees or pi/2, as only then the two
    signals multiplied are zero.
    The phase error is assumed zero when the signals are 90degrees
    to each other. Yes, there is some timing to lock involved and it
    is dependent on the loop gain function.
    There are various application notes from eg Analog Devices and others.
    The noise of a PLL is called phase noise and is also expressed
    as 'ps Jitter RMS'.

    Rene
     
    Rene Tschaggelar, Jul 9, 2003
    #4
  5. Jay

    maxfoo Guest

    Take a look at th HMC439 and 440 Phase freq detector data sheet
    and the PLL calculator on Hittite Microwaves website
    http:www.hittite.com
     
    maxfoo, Jul 10, 2003
    #5
    1. Advertisements

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.