Help with O/A problem

Discussion in 'General Electronics Chat' started by waltsar, Mar 7, 2015.

  1. waltsar

    Thread Starter New Member

    Nov 10, 2012
    10
    0
    This circuit is supposed to create a clean, linear, symmetrical saw tooth wave. The problem is that there is a small "glitch" in the waveform as shown in the attached drawing. Anyone have any idea why this small "step" appears near the top of the positive-going slope of the wave?

    I'm using an LM324 quad O/A chip. I've used this circuit in the past using a longer (10 second) ramp time and had no noise problems.
     
  2. #12

    Expert

    Nov 30, 2010
    16,345
    6,831
    I'm not sure, but I'd try increasing the 2.4k resistor so the output doesn't get all the way to the positive supply before the timer flops.
     
  3. cmartinez

    AAC Fanatic!

    Jan 17, 2007
    3,574
    2,549
    Electrolytic capacitors tend to do that sort of thing. Try changing it with a tantalum one.
     
  4. blah2222

    Well-Known Member

    May 3, 2010
    554
    33
    I have a suspicion that the RC time constant (2.4kohm * 2.2uF = 5.28 ms) is causing the issue. As suggested try increasing 2.4kohm so that RC constant is much larger than 8ms.

    What seems to be happening is that the negative input of the first op-amp is ~98% (220k/224.7k) of 11.5V when the 555 timer is in ON state. With the output initially at zero, this causes the output of the first op amp to swing hard to the negative rail (-11.5V). This in turn causes current to be pulled from the capacitor thru the 2.4k resistor, thus, the output starts ramping positively based on the formula I=C*dv/dt. It takes the output ~4.8ms to hit the 98% mark which then causes the first op amp's output to swing to the positive rail once it passes thru that threshold. Note that the 555 timer has not gone LOW yet. Since the output of the first op amp is at 11.5V and the 555 timer output is 11.5V, there is no current flow through R1 and R2 so its negative input goes to 11.5V. Since the output of the second op amp is not greater than 11.5V the first op amp output swings low again, thus, entering in a battle causing the op amp to flip-flop b/w being low or high. What this means is that the output stays relatively constant at 98% of 11.5V until the 555 timer goes LOW.

    This seems to match up fairly well with what you were seeing on the scope, though your timing scales seem to be a bit different. Someone please correct me if I am missing something though...

    Is there a reason why you are using the 555 timer? You could generate the same thing using just the two op amps and capacitor: http://www.qsl.net/zl1an/Downloads/JanFeb11_solution.pdf
     
    Last edited: Mar 7, 2015
  5. crutschow

    Expert

    Mar 14, 2008
    13,056
    3,245
    That circuit is rather strange.
    What is the feedback from the output to the plus input of the left op amp supposed to do? :confused:
     
  6. #12

    Expert

    Nov 30, 2010
    16,345
    6,831
    I thought I was merely ignorant about that configuration. :D
     
  7. cmartinez

    AAC Fanatic!

    Jan 17, 2007
    3,574
    2,549
    Well, if you're merely ignorant... that would make me just plain ignorant... :confused:
     
    #12 likes this.
  8. waltsar

    Thread Starter New Member

    Nov 10, 2012
    10
    0
    When the input from the 555 goes high it saturates Q1 in the (-) direction because the (+) input is low. When the output of Q2 which is connected to the (+) input of Q1 and is charging through C1 and the 2.4K resistor, exceeds the (-) input of Q1, Q1 instantly reverses polarity and becomes saturated in the (+) direction. Long way of saying that the input of Q1 is differentially connected.
     
  9. WBahn

    Moderator

    Mar 31, 2012
    17,777
    4,805
    You're railing the amplifiers and all kinds of demons live there.

    Like the others, I don't see what the feedback from the output to the first stage in all about, except possibly to create a Schmitt-trigger like circuit. If you want the first stage to act like a comparator, then use a comparator.

    Try reducing the signal amplitude so that the opamps remain in their linear regions and see if the artifacts still remain.
     
  10. crutschow

    Expert

    Mar 14, 2008
    13,056
    3,245
    Here's a two op amp triangle wave generator (LTspice simulation below) that requires no 555.

    If you want an adjustable frequency use a pot for R1.

    Triangle Gen.gif
     
    Last edited: Mar 7, 2015
Loading...