Half Wave Symmetrical Sawtooth Waveform

Discussion in 'Analog & Mixed-Signal Design' started by Carlos Lengua, Jun 30, 2018.

  1. Carlos Lengua

    Thread Starter New Member

    Jun 30, 2018
    4
    0
    I am new to this forum and I hope the question isn't too broad but I am seeking a circuit or even a device/instrument that can generate a half wave symmetrical sawtooth waveform. I do have a basic idea of how I might try to generate such a waveform but I'm not sure if it would work.

    Charles
     
  2. ericgibbs

    Moderator

    Jan 29, 2010
    5,442
    1,019
    Hi Charles,
    Welcome to AAC.
    If you post your draft circuit idea, we could help you with it.?
    E
     
  3. AnalogKid

    AAC Fanatic!

    Aug 1, 2013
    7,003
    1,985
    Please post a sketch of the waveform you want to achieve. Based on what I think a normal sawtooth waveform is, "half wave symmetrical sawtooth waveform" is a contradiction. Symmetrical about what?

    ak
     
  4. MrChips

    Moderator

    Oct 2, 2009
    16,724
    5,127
    Half wave
    symmetrical
    sawtooth

    Those terms conflict. We know what a sawtooth waveform looks like.

    [​IMG]

    Edit: ak beats me to it.
     
  5. danadak

    Well-Known Member

    Mar 10, 2018
    1,433
    293
    An analog approach, add a mux to convert triangle to saw using square output.

    http://www.learningaboutelectronics.com/Articles/Function-generator-circuit.php


    Use a comparator based circuit -




    Then a digital approach using wavetable and DAC

    http://www.cypress.com/documentatio...p-easy-waveform-generation-wavedac8-component

    http://www.cypress.com/file/122721/download

    With this approach you can control freq and amplitude, and using a small amount of other
    onchip resources burst N saw waves with controlled delay between them. Very versatile.
    Or trigger it synchronously with external input or generate simultaneously a tri or sine......

    This project shows a switchable sesign. sine or saw. two wavedacs could have been used to gen
    them simultaneously. See attached.

    What freq and amplitude do you need ?



    Regards, Dana.
     
  6. Carlos Lengua

    Thread Starter New Member

    Jun 30, 2018
    4
    0
    I had in mind something pretty much what's posted in this video:

     
  7. crutschow

    Expert

    Mar 14, 2008
    19,518
    5,443
    Is this what you want?
    What is the frequency and amplitude you need?
    If you have an idea on how to do it, then post it.

    upload_2018-7-3_10-50-48.png
     
    Carlos Lengua likes this.
  8. Carlos Lengua

    Thread Starter New Member

    Jun 30, 2018
    4
    0
    Yes, something just like that. I would like the amplitude to be variable from about 0-5 V +/- and frequency no greater than about 1-5 kHz for the time being.
     
  9. danadak

    Well-Known Member

    Mar 10, 2018
    1,433
    293
    Here is the circuit to implement sign change, feed it with
    a saw and it creates the waveform crutschow posted. Note
    gate of switch feed by saw generator square waveform if
    available, otherwise use a comparator.

    upload_2018-7-4_9-29-17.png

    Choice of MOSFET mainly low RDSon, << 1K ohm

    Regards, Dana.
     
    Carlos Lengua likes this.
  10. Kjeldgaard

    Member

    Apr 7, 2016
    270
    97
    I have an idea for the classic 2 operational amplifier Integrator / Schmitt Trigger circuit:
    TriSquWav_1.jpg
    At the point I have called OUT, there will be a signal that reminds of the TS wishes.

    But these are just thoughts, and there is a lot of detail with regard to supply, common point / GND and buffer on OUT.
     
    Carlos Lengua likes this.
  11. crutschow

    Expert

    Mar 14, 2008
    19,518
    5,443
    Here's the LTspice simulation of a circuit to generate the desired waveform.

    It uses LM339/339 comparators to generate a unipolar sawtooth, with current-mirror Q2/Q3 providing a constant-current to give a linear sawtooth ramp at C1, adjustable by pot U2 to change the frequency.

    The quad LM6484 opamps provide signal buffering, inversion and adjustable gain.

    Flip-flop U2 changes state at the end of each sawtooth retrace pulse, which controls switch U7.
    This reverses the phase/polarity of U5's output (the circuit posted by danadak) on alternate sawtooth periods to generate the bipolar symmetrical sawtooth.

    Pot u9 adjusts the output between 0V and ±5V (maximum with 5V supplies).

    The simulation is shown for the two extreme settings of pot U2 which gives frequencies from about 700Hz to 7kHz.

    (Note that the circuit in post #10 generates a triangle wave, not a sawtooth.)

    upload_2018-7-4_11-59-19.png
     
    Last edited: Jul 4, 2018
    Carlos Lengua and Bordodynov like this.
  12. MrAl

    AAC Fanatic!

    Jun 17, 2014
    5,120
    1,109
    Hello,

    It seems to me it would easier to just create a normal sawtooth generator, then invert the output with anther op amp stage, then use an analog switch to 'choose' which output gets linked to the final output using a flip flop triggered by the original sawtooth generator.
    I think you can do a normal sawtooth gen with two op amp sections, then one more for the inverted output, then one flip flop and one analog SPDT switch.
    So that could happen in as little as one quad op amp package, one flip flop package, and one analog switch package, and a few small components like resistors and caps.
     
    Carlos Lengua likes this.
  13. crutschow

    Expert

    Mar 14, 2008
    19,518
    5,443
    I liked to see that circuit.
    I know you can do a triangle-wave with two op amps, but a sawtooth is a little more difficult.
     
  14. MrAl

    AAC Fanatic!

    Jun 17, 2014
    5,120
    1,109
    Hello,

    I have seen a design using zeners but did not analyze it yet.

    But couldnt you just create a standard triangle generator and use diodes to control the duty cycle separately for the rise and fall times? A sawtooth is just a triangle with faster fall time than rise time, or vice versa.
     
  15. crutschow

    Expert

    Mar 14, 2008
    19,518
    5,443
    In theory yes, but typically you want the sawtooth fall time to be very short, and an op amp maximum output current is not usually sufficient for that, so a separate, high-current switch is used to generate the fall-time (rapidly discharge the capacitor).
    In the circuit I posted, the LM339 output transistor switch is used for that purpose.
     
  16. Kjeldgaard

    Member

    Apr 7, 2016
    270
    97
    Here are a few more sketches in continuation of my post #10.

    I thought the hysteresis feedback signal was the waveform that was desired, but here the square component is inverted.

    So here is an update of the schematic:
    TriSquWav_2.jpg

    And a set of graphs:
    TriSquWav_3.jpg
     
    Carlos Lengua and Bordodynov like this.
  17. crutschow

    Expert

    Mar 14, 2008
    19,518
    5,443
    Okay.
    MrAl jogged my thinking and while I was doing that, Kjeldgaard came up with a circuit similar to the one I was working on.
    Basically it's just adding a period square-wave offset to the triangle-wave.
    Requires just one quad op amp.
    Here's the LTspice simulation of the result.

    upload_2018-7-5_0-14-29.png
     
    Carlos Lengua and Bordodynov like this.
  18. MrAl

    AAC Fanatic!

    Jun 17, 2014
    5,120
    1,109
    Hello,

    Depends on the requirements of the application.
    Requires one dual section op amp. Maybe LM358 which is a 8 pin package.
    If you want the inversion then another op amp section is needed.
     
  19. Kjeldgaard

    Member

    Apr 7, 2016
    270
    97
    @crutschow and post #17

    As I see it, you have the square signal in the wrong polarity, like the signal C in my second picture from post #16.
     
  20. danadak

    Well-Known Member

    Mar 10, 2018
    1,433
    293
    Do you need the waveform to have controllable offset or is it
    always symmetrical about ground ?

    Regards, Dana.
     
Loading...