Connect with us

characteristic impedance

Discussion in 'Electronic Design' started by Krish, Mar 28, 2005.

Scroll to continue with content
  1. Krish

    Krish Guest

    characteristic impedance of a line is square root of L/C. It does not
    depend on the length. But the losses occuring in resistance, no matter
    how small, is dependent on the length. Any literature to read on how
    the length affects the charactristic impedance.
    When doing circuit analysis, will this charactreritic impedance be
    added to the series connecting element?
    I have a device that says input impedance 50ohm, i am not using it.
    But to use other part of the device i have terminate it with 50ohm.
    How can i do it? can i just connect a 50ohm characterisitc impedance
    cable and leave it?
    thanks in advance
  2. Active8

    Active8 Guest

    Something like that.
    Most texts leave out the resistance as negligible, but I've seen it
    figured in - just don't remember where.
    Read up on VSWR, reflection coefficient, return loss, ... If a 50
    ohm source is connected to a 50 ohm line terminated with 50 ohms,
    you'd ignore the line impedance, but the line loss is another story.
    There's different kinds of analysis, so exactly how you do it is
    another story.
    No. You'd want a 50 ohm resistor. Whether or not you use the line is
    up to you, but it would need to be terminated. You can buy
    terminators that mate to different connectors. Good connections like
    that keep water out and RF leakage down. You should probably
    terminate the input whether or not you use it.

    And your secret device isn't worth keeping secret, so if this
    doesn't answer your questions, you'll have to spill the beans.
  3. PeteS

    PeteS Guest

    Zo is sqr(L/C), to a close approximation, provided the following is

    1. The frequencies of interest are well above the skin effect knee
    frequency. Above this frequency, R(s) is fairly constant (but not

    2. Rs is negligible (per unit length) relative to Zo.

    As noted, you should read up on transmission line theory and filter
    theory (which is where that equation comes from). For most circuits,
    provided the above are true, you can simply terminate to a resistor,
    although many circuits designed for high frequency have terminators
    internally. Some do, some don't.

    This is something you can spend years on and still not understand
    completely (there are many things in this field which have been found

    Note that to get a decent 'feel' for characteristic impedance (and
    transmission line theory in general), you should be comfortable
    analysing in the frequency domain and dealing with S-parameters
    (there's some more reading).


  4. I read in that Krish <>
    The full expression for characteristic impedance is (use Courier font):
    /R + jwL
    Zo = /---------
    \/ G + jwC

    where R = series resistance per unit length
    L = inductance per unit length
    G = shunt conductance (dielectric loss) per unit length
    C = capacitance per unit length
    No. Use a 50 ohm resistor.
  5. Mac

    Mac Guest

    In real coaxial cable, the effect of resistance can be specified as a
    certain amount of dB power-loss per foot of cable. For example, if the
    cable has 10 dB of loss per foot, then 5 feet of cable would
    absorb 50 dB of the total power.

    Another effect to keep in mind is that as the frequency goes up, the dB
    losses per foot usually go up. So if a particular cable has 10 dB of
    loss per foot at 500 MHz, it may have 50 dB of loss per foot at 1.5
    GHz. Note that I'm just making these numbers up.

    So, to answer your question, this power loss doesn't really effect the
    characteristic impedance of the cable. To a first approximation, the loss
    can be modeled separately from the characteristic impedance.

    Note that the above discussion assumes you have a resistive load at the
    end of the transmission line whose impedance is the same as the
    transmission line's impedance. For example, if it is a 50 Ohm transmission
    line, you should have a 50 Ohm resistor (or 49.9) at the end as a load.

    And to answer your other question, the best way to terminate a device at
    frequencies up to UHF is to use a small surface-mount resistor (49.9 Ohms
    in this case) as close as possible to the device. If you can't get close
    to the device, for some reason, use a 50 Ohm cable with a 49.9 Ohm
    resistor at the end of it. You can also buy 50 Ohm terminators for cables
    and connectors. If the device has some kind of an RF connector on it, it
    may be easier to use a terminator for it.

    If the frequencies are down around 100 MHz or less, then you don't really
    have to be all that close to the device or do anything fancy. Just put any
    type of 50 Ohm resistor across the terminals somehow.

  6. Mark

    Mark Guest

    in theory yes, a VERY long cable will look like 50 Ohms due to it's
    losses ...

    in practice, what you want is a termination plug

    just by a 50 Ohm termination plug to connect to your source.

  7. Don Pearce

    Don Pearce Guest

    It will indeed look like 50 ohms - but this is nothing to do with its
    losses. It is simply that power is propagated rather than dissipated.
    Provided none is reflected back while you are measuring, the cable
    will appear to be a 50 ohm resistor.

    Pearce Consulting

  8. A bit of thinking is required.
    Apart what has been said, there is DC termination and
    there is AC termination. The AC termination being an
    additional cap in series with the 50 ohm.
    It all depends whether a DC is required or not.

    For more info, you need to give some more information.

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.
Electronics Point Logo
Continue to site
Quote of the day