Connect with us

Impedance Matching (A Match made in heaven)

Discussion in 'Electronic Design' started by Kevin Doyle, Jul 19, 2005.

Scroll to continue with content
  1. Kevin Doyle

    Kevin Doyle Guest

    Hi all,
    I'm browned off from reading unhelpful literatuire about impedance matching
    networks.

    I have two questions that need answering if possible.
    Q1. I have a CE Amp with Rc=576ohms so Zout =576 ohms. The CE Amps a CC
    Buffer with Zin=15Kohms.
    How do I match these two circuits for transferring the full and propper RF
    energy.

    Q2. How do you know if the input or output impedance has reactive components
    in it.
    If the CE amplifier above has capacitor on the output does its Zout become
    576-j1/WC ohms?

    Regards,
    Kevin.

    PS the avtice devices I'm using are the mps 5179 npn Transistor ft=900Mhz.
     
  2. Andrew Holme

    Andrew Holme Guest

    You don't always need to. Impedance matching is typically required between:
    1. antennas and recivers;
    2. transmitters and antennas;
    3. high-power stages in a transmitter

    See the recent thread "How come its OK to mix impedances in a radio
    system??" in s.e.b.

    Judging by the impedance levels you mention, those are low-power stages.
    They can be voltage driven. Between higher power stages, the matching
    network is designed to present the driver with just the right dynamic
    collector load to suck out the correct amount of power needed to drive the
    following stage. You've probably seen examples using P = Vpk^2/2R to
    calculate collector load.

    At the other end of the power range - in receivers - impedance matching is
    important for totally different reasons: to preserve precious pico-watts of
    energy, maximising the signal-to-noise ratio. At low and intermediate power
    levels, it doesn't matter.

    You can work out how much drive a stage needs by subtracting the transistor
    manufacturer's stated gain from the desired output power. Or you can do it
    by trial and error. It's good to have a way to adjust drive.
    If the manufacturer's datasheet doesn't tell you, you can always include
    trimmers in your circuit to allow for some latitude of adjustment.
    Not if the capacitor and resistor are (effectively) in parallel.
     
  3. colin

    colin Guest

    Theres been a few of these questions lately.

    With Zout << Zin the power going into the CE stage will be Vout/Zin.
    If you increase the impedance of Zout by adding a series resistor to match
    Zin you will see a reduction in the power transfered.
    If you decrease zin by adding a parallel resistor to match zout you will
    also see a reduction in power transfer.
    In both cases you will see a reduction in voltage at the cc amp what is
    probably not what you want.

    So impedance matching does not necessarily mean you get full and proper
    power transfer.
    I dont know why they seem to teach this as if its always necessary.

    although you might get maximum transfer if zout = zin for high power systems
    you wil also get maximum smoke !

    as has been said before if you have a multi killowat transmiter with an
    internal resistance of 50ohm it will melt, it will usualy have much less.

    Impedance matching is usualy only necesary at the receiveing end of
    transmision lines or filter stages to ensure they perform properly, you have
    no such thing here that needs to be matched.


    However you can use a rf transformer to increase the output voltage and
    power transfer wich will at the same time increase the Zout by a factor of
    n^2.
    If its a largish decoupling capacitor in series with the output it can
    probably be ignored, the stray capacitances, inductances and transister
    charecteristics become an issue at high rf frequencies, like a capacitor in
    parallel with the Rc it will add a reactive component.

    Colin =^.^=
     
  4. Kevin Doyle

    Kevin Doyle Guest

    Thanks for that guys!
    Kevin.
     
  5. John  Larkin

    John Larkin Guest


    What frequency are you working at? What are the input and load like?
    Ordinary low-freq transistor designs, like resistive-loaded
    common-emitter amps and emitter followers, tend to not work at high
    frequencies.

    John
     

  6. Kevin.

    Time to get a book or two.


    Solid State Design for the Radio Amateur - published by ARRL

    RF Circuit Design - Chris Bowick - published by SAMS.

    Introduction to RF Design - Wes Haywood - Prentice Hall


    Flipping thru Bowick's book I see a chapter on Small Signal RF Amplifiers
    and one on Power Amplifiers - nice discussion on the difference. Also a
    chapter on Impedance Matching". Clear worked examples. You can build
    transmitters with this information! ;) Recommended.

    Perhaps this thought will help. Mostly, we are not impedance *matching* -
    we are impedance *transforming*. The same networks and maths is used for
    both. Mostly, we transform the impedance of the load into the impedance
    required by the amplifier. The impedance required by an amplifier depends
    on its output voltage swing and the power it can safely deliver. In the
    rarer cases where we do impedance match, we use transform the load to the
    correct value - often 50 ohms - stamped on the amplifier output etc.

    Roger
     
  7. I dont think they do. Its usally a case of a smart arse proffessor
    trying to be clever with a bit of calculas. You know, the bit about
    diferentiating and setting to zero. Then the dumb shit students pay no
    attention at all as to the conditions of the derivation.

    The max power theorem is when, given a *fixed* source impedance, and the
    load is varied, then value of load gives max power when RL=RS. If the
    load is fixed and we want to find the max power into it, by varying Rs,
    the max occurs when Rs=0!

    Kevin Aylward

    http://www.anasoft.co.uk
    SuperSpice, a very affordable Mixed-Mode
    Windows Simulator with Schematic Capture,
    Waveform Display, FFT's and Filter Design.
     
  8. colin

    colin Guest

    Maybe, just seemed to me there does seem to be a surprising number of people
    who think this though.
    I cant remember how they taught it at uni, maybe I wasnt paying attention !
    yes thats a better way of explaining the situation actualy, youl get least
    smoke too ! :D

    Colin =^.^=
     
  9. Kevin Doyle

    Kevin Doyle Guest

    Hi there,
    The circuit I have is in the FM Band 88-108.
    Yes I think I need to get a good book on the topic.
    I remember my lecturer in college using the system of transfer of max power
    for every
    bit of theory we did "Thanks Michael!" It looked great and and clean cut on
    the board but
    now I'm finding it doesn't work in the real work. I am getting some good
    help here though, thanks!

    Kevin.
     
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

-