Connect with us

RF transistor biasing

Discussion in 'Electronic Design' started by Jamie Morken, Dec 30, 2004.

Scroll to continue with content
  1. Jamie Morken

    Jamie Morken Guest

    Hi,

    Why do many RF amplifiers use an external biasing resistor instead of a
    push pull halfbridge configuration? Is the external biasing main
    function to pull up the output since the amplifier only can sink current
    on its output?

    example of an RF amplifier that requires external biasing:
    "http://www.minicircuits.com/appnote/an60010.pdf"

    Also can the same MMIC (monolithic microwave integrated circuit)
    amplifier be used either as a LNA or as an RF power amplifier?
    ie. the "Gali series":
    "http://www.minicircuits.com/dg03-166.pdf"

    cheers,
    Jamie
     
  2. Don Pearce

    Don Pearce Guest

    Several reasons. It saves a pin and lets you use a four-pin package.
    It provides flexibility to use whatever rail voltage you have. And it
    removes one of the bigger power dissipation components from the
    package.

    d

    Pearce Consulting
    http://www.pearce.uk.com
     
  3. John Larkin

    John Larkin Guest


    The technologies only produce fast N-type devices... silicon or InGaAs
    NPN bipolar transistors so, yes, they can only pull down. There have
    been a few MMIC-type amps with active pullups (NEC maybe? Maxim?) but
    they tend to be slow.

    There's nothing inside most of these parts but a Darlington and a
    couple of resistors.

    We had a thread a few weeks ago on the problems associated with
    creating a super-wideband inductor for use in the pullup.

    John
     
  4. push pull ? They'd to be driven anti phase. Can you supply
    antiphase drive over the mentioned bandwidths ? From basically
    DC to several GHz ?

    Rene
     
  5. Active8

    Active8 Guest

    what was the subject if you remember?
     
  6. Andy

    Andy Guest

    Andy replies:
    I've use the MMIC amps for a long time as both LNAs and
    Power amps.
    However, they are really great at either.

    The noise figures, IM, and power diss aren't close to what you can get
    with a discrete design, when doing LNAs. However, if
    your requirements are not "best that can be done", they are
    extremely convenient.

    For power amps, I think some of the manufacturers have
    devices in the 100 mw range, tho I haven't used one for that
    yet. The current requirements are pretty high, and the
    IM specs aren't great. But, if your specs aren't too
    stringent, it's very handy to plop one or two down on
    a 100 mil track on G-10 (FR4) to get 50 ohms, and move on...

    They can even be directly paralleled to almost double the
    power output (but not the gain).

    To use them as LO buffers which put out +10 dbm or so
    to a balanced mixer is a slam dunk...


    In short, they answer to your question is YES, BUT.

    The convenience and speed of design must be weighted against the
    specifications. Often , it is a good method.

    Andy
     
  7. John Larkin

    John Larkin Guest

    "Wideband inductors", Dec 14.

    Oh, one of my guys did disect a PSPL bias tee...

    John
     
  8. Jamie Morken

    Jamie Morken Guest

    Hi,
    Is that for 0.062" thick PCB? So you can run 100mil traces for all the
    RF signals if everything is 50ohms? :) Does this apply to a 2layer
    board with a ground plane on the bottom?
    Do you use a common bias resistor/wideband inductor for this or one per
    amplifier?

    cheers,
    Jamie
     
  9. John Miles

    John Miles Guest

    Have you seen the Mini-Circuits ADCH-80A? Nice wideband part with
    multiple series inductances, characterized to (IIRC) 8 GHz.

    Might have been mentioned in the other thread; I didn't see that one.

    -- jm
     
  10. John Larkin

    John Larkin Guest

    Those are nice, but L is low and drops like a rock as the DC current
    increases. For the amp we're doing now, we needed 100 uH maybe at 200
    mA, so we wound up with a string of three ferrite beads of increasing
    size, followed by two axial inductors, each with a shunt resistor.
    Inductors are sure non-ideal gadgets.

    John
     
  11. Andy

    Andy Guest

    ******** 100 mils on .062" FR-4 , ground one side, track on
    the other, forms a 50 microstrip (very very close). I have
    often just used an exacto knife to cut it out.
    I have also used adhesive copper tape on the ungrounded side.

    This is
    known as "micro-strip" and is well defined by graphs of
    Wheeler's Equations. It's just as easy to make 35 ohm or
    70 ohm, but I forget the widths right now.... Higher impedance
    means narrower track, and by 300 ohms the track is smaller
    than I want to mess with using an exacto knife.

    Note that this subject is a complex one, but 50 ohm track
    is so common that RF engineers, who do their own assembly and
    building (getting rarer these days) remember certain dimensions as
    "rules of thumb".....
    While the mathematical types will argue till hell freezes
    over about 2 or 3 ohms difference, as a practical matter it
    is immeasureable in performance....




    ***** I believe that the mini-ckts lab designer handbook has
    some examples. I, personally, would uses separate biases
    and cap couple the outputs so one amp can't
    current hog, but , I may be being overly conservative....

    The inputs MUST be cap coupled separately. Use chip caps.

    Understand, in my opinion, this will work OK but I would use
    it only in non-critical applications, where the purpose is
    to get some quick increase in power output where there are
    no OVERRIDING other considerations. Quick and Dirty. But
    that's just my opinion...... Often certain parts of a
    design will not require any more than this, like line
    amplifier, LO buffers, etc. Experience usually is a good
    guide as to whether this method will cause unforseen
    problems.....

    One advantage is that an exacto knife will open up the
    line so you can just solder coupling caps in. Also easy to
    tap the line with bias resistors and chokes........You will
    see immediately the advantages for quick assembly once you
    try it...



    Good luck with your project. I would reccommend that you
    check out EDN's "Ideas for Design" each month to get a
    feel for some innovative ways to do different things.
    Also, LOTS of circuits are in The Radio Amateur's Handbook,
    and the UHF Experimenters Handbook...... If you have a copy
    of each of these in a stack in your bathroom, you should
    be in good shape...

    Andy W4OAH
     
  12. John Larkin

    John Larkin Guest

    Agilent's free AppCad program is great for calculating microstrip,
    stripline, and cpw impedances.

    John
     
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

-