Bode and stability of an amplifier (using PSPICE)

Discussion in 'Homework Help' started by Lintukori, Jan 13, 2010.

  1. Lintukori

    Thread Starter New Member

    Jan 13, 2010
    Didn't see any threads considering problem that I have.

    I have a task of designing an amplifier with following qualities:
    -inverting op-amp circuit for voltage amplifying
    -a b-class push-pull circuit for current amplifying
    -bandwith 20-20000Hz
    -voltage gain 10

    So I managed to design a fairly good amp (well i think so, because the PSPICE simulations look quite good) using guidelines from the
    B Class Audio amplifier example that can be found in this website
    But the problem is, that I have to show its stability by simulating the
    circuit with PSPICE and plotting the bode diagram and find its gain and phase margin. So how do I do that.
    Sorry that I don't have the images of my particular design, but you can
    refer to the circuit in the example above url (because its quite similar, except I'm using only one op amp and some other minor things).

    If I understand correctly, the stability margins of a feedback system should be looked from the OPEN-LOOP system Bode diagram. So how do I simulate the open-loop system gain and phase of the circuit using PSPICE?
    Should I just cut the feedback wire and simulate and that's it? Or I'm I totally wrong about the whole stability thing?
    Last edited: Jan 18, 2010
  2. Lintukori

    Thread Starter New Member

    Jan 13, 2010
  3. Audioguru


    Dec 20, 2007
    Your output transistors have a bandwidth that is much higher than the opamp and have a low input capacitance so stability should be no problem.

    Are you making a class-B amplifier with crossover distortion or a class-AB amplifier without crossover distortion?
  4. Darren Holdstock

    Active Member

    Feb 10, 2009
    If the circuit uses the same feedback loop to stabilise the DC operating point then breaking the loop will also break this. A workaround for this is to split a net by inserting a stupidly large inductor (1 GH, say), which will allow for a stable DC bias while opening the loop for AC signals.

    A good stability margin in a cromulent Bode plot is not an absolute guarantee of stability (Routh is better in this respect), but it's pretty good in most instances.