Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Crossover distortion?

  1. Jan 16, 2010 #1
    I'm getting a bit of what appears to be crossover distortion in my audio pre-amplifier simulation.
    Tried a bunch of different fixes, it was still there. Could it be that the simulation's transistors are so exactly matched?

    I'll continue searching for an error in my model.

    Thanks for any help.


    Attached Files:

  2. jcsd
  3. Jan 16, 2010 #2


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    What's it like at lower frequencies? You seem to have some ringing, too.
    Isn't there some adjustment in the driver circuit to get the bias right and minimise crossover problems?
    But the glitch is not where I should have expected crossover distortion - i.e. half way up. There seems to be too much asymmetry for it to be simple C/O distortion. Could one leg of the driver circuit be faulty - like an electrolytic connected the wrong way round. Feedback may be doing its best to correct a really major distortion somewhere.
  4. Jan 16, 2010 #3
    It seems to have something to do with the middle tone control. As I lower it, the distortion decreases until it's almost gone with the control at 0%

    If I remove the middle control from the circuit entirely the 'crossover' disappears.

    Now that I have the problem so isolated I feel confident I'll be able to weed it out. If not, I'll be back.

    Where is the ringing in the wave? It does sound a little doubled or 'fat' as the synthesizer guys like to use. But again, with the middle control lowered, that effect goes away.

    Thanks for the help!
  5. Jan 16, 2010 #4
    Hello erer44

    Could you please post a schematic of your circuit, so that we can get a better idea of what is going on. It would also be very helpful to know the input signal, and output load (if any).
  6. Jan 16, 2010 #5
    Of course. This is the preamplifier section I am working out the kinks on right now. I'm putting 1.5vP into it from the function generator, and it goes through a power amp and output transformer for 8 or 16 ohm speaker loads.

    Thanks again! Let me know if there's anything you cant make out.


    Attached Files:

    Last edited: Jan 16, 2010
  7. Jan 16, 2010 #6
    And here is the oscilloscope with the middle pot all the way down and all the way up.

    Attached Files:

  8. Jan 16, 2010 #7
    Why don't you put 1 millivolt peak into it and see what you get? 1.5 volts peak input seems a bit much. Make sure you are getting less than 1 volt peak out (TR9 collector). Also, measure the dc steady state voltages of all transistors, and compare them to the schematic.
    Bob S
    Last edited: Jan 16, 2010
  9. Jan 16, 2010 #8
    That did it, thanks!! Now I can get to building the real thing.

    I'm in this just a bit over my head. Starting to understand more and more though.
  10. Jan 17, 2010 #9
    I think it would be advisable for you to find out what input level this circuit was designed to handle, and also what output to level expect. [Alternatively, what input level and what gain.]

    This would help you with setting up more realistic simulations, and interpreting the results.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook