1. Not finding help here? Sign up for a free 30min tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

RLC circuit overshoot voltage

  1. Apr 16, 2012 #1
    1. The problem statement, all variables and given/known data
    RLC circuit as shown in the attachment. Input is a pulse, of frequency 1MHz, 5v. Rise/fall time 1ns.
    How to find the voltage at the capacitor.
    In simulation, I see overshoot and undershoot. But how do I find the overshoot/undershoot amplitude mathematically?
    Ringing frequency is f = 1/2*pi*sqrt(LC)


    2. Relevant equations


    3. The attempt at a solution

    No idea. I tried VL = L di/dt
    But not sure how to get dt.
    1. The problem statement, all variables and given/known data



    2. Relevant equations



    3. The attempt at a solution
     

    Attached Files:

    • rlc.pdf
      rlc.pdf
      File size:
      5.6 KB
      Views:
      256
  2. jcsd
  3. Apr 16, 2012 #2

    NascentOxygen

    User Avatar

    Staff: Mentor

    The general approach is to form the second-order differential equation, and solve it. You have a series circuit, with a current i(t). You know the equations for voltage across a resistor, across an inductor, and across a capacitor, all in terms of the current through each, so sum those voltages. This sum equals the applied voltage at all times.

    The topic is treated well in most textbooks, and at wiki http://en.wikipedia.org/wiki/RLC_circuit. Possibly your examiner requires nothing more of you than just using the equation the text derives for the overshoot, after plugging in the values for your circuit to determine the damping, ζ.

    That 1pF capacitor is very tiny, though not totally unrealistic. Are you sure the value is correct?

    Good luck! :smile:
     
  4. Apr 17, 2012 #3
    It's not an assignment problem. I made it up.
    The RLC ckt is the equivalent of a scope probe. I am trying to mathematically derive the overshoot amplitude.
    Since the input is a pulse, I can't use the standard textbook approach.
    I have attached a plot of the voltage across capacitor for rise/fall time of 0.5ns.
     

    Attached Files:

  5. Apr 17, 2012 #4

    NascentOxygen

    User Avatar

    Staff: Mentor

    Probes are usually carefully designed. I'm skeptical about your equivalent circuit, have you included the resistance of the source in your simulation? Also include the source capacitance. Try 50Ω and 10 pF if you are just making up values. :smile:

    I wouldn't call that a pulse, your input is a step, repetitively, and that's exactly what the textbooks deal with. If a system rings with a step, it will ring with a delta.
     
  6. Apr 18, 2012 #5
    The values you mentioned are for passive probes.
    I'm trying to play with approximate model of an active probe.
    The L and C values are correct. I added the R to dampen the ringing caused from ideal L and C in simulation.
    Can you point me to resources which will help me find the delta?
     
  7. Apr 18, 2012 #6

    NascentOxygen

    User Avatar

    Staff: Mentor

    Laplace Transforms give you the system response for any input, including Dirac's delta. Look for an electronics textbook introducing Laplace Transforms. Many textbooks will have worked examples. Or search online.

    Good luck!
     
  8. Apr 19, 2012 #7
    Found this - http://csserver.evansville.edu/~blandfor/EE210/MatLab/MLEx5.pdf
    Plugged in values for time and got voltage close to the one in the plot (page3) for underdamped case.

    But in my circuit, the overshoot increases as I make the step input rise time fast.
    How do I take this into account?
     
  9. Apr 19, 2012 #8

    NascentOxygen

    User Avatar

    Staff: Mentor

    Have you learnt about Laplace Transforms in your studies?
     
  10. Apr 19, 2012 #9
    Yes, I did learn LT back when I was in school.
     
  11. Apr 19, 2012 #10

    NascentOxygen

    User Avatar

    Staff: Mentor

    In the time domain, to generate a step input with a finite rise-time:
    at t=t₀ apply a steep ramp of slope g volts/sec, i.e., v(t)=g·t
    at t=t₁ add to this input a ramp of equal but opposite slope, viz., -g volts/sec,

    With your system being represented by a linear second order DE, you can use the superposition of responses to more than one input to find the composite output. (I can't help with the mathematics, have forgotten those details.)

    Your best hope lies in using Laplace Transforms. Maybe here's your opportunity to revise that topic? I only use LT for sinusoidal responses, but maybe someone else can help you with your piece-wise LT. Good luck!
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook