• State
  • Locked Locked
  • Replies 7 replies
  • Subscribers 48 subscribers
  • Views 3698 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

1 MHz to 1.5 MHz PWM or triangle wave generator

David Johns
Prodigy 40 points
Other Parts Discussed in Thread: TLC555I'm looking for a simple solution to creating a 1 MHz to 1.5 MHz PWM controller or triangle wave generator. I'm looking for a peak to peak voltage of no less than 5 V, no more than 8 V. Does anyone have a a suggestion? I've searched the databases, data sheets, etc - and cannot find anything looks right for the low MHz operating regime. THX.
  • 0
    TI__Guru**** 168485 points

    Hi David,

    One 'simple' way to create a PWM generator is to use one of the timer outputs on the MSP430.  Here are a couple app notes to get you started:

    http://focus.ti.com/lit/an/slaa116/slaa116.pdf

    http://focus.ti.com/lit/an/slaa120/slaa120.pdf

  • 0 in reply to Tom Hendrick
    Prodigy 40 points
    I'm looking for something a little more fundamental. My goal is to obtain a triangle ramp wave form with a peak to peak voltage of between 5 and 8 volts. This Vpp range defines the usable signal amplitude for the error signal of the high power PWM that I want to build. The triangle wave form will feed into one input of a comparator, the control error signal goes into the other input of a comparator, and the comparator output wave form duty cycle is then proportional to the error signal = a very basic PWM controller, but with a large signal amplitude and higher-than-typical frequency. I really need large signal swing on the error signal to drive down the signal-to-noise ratio. This is the primary controller for a 1 megavolt, 100 kW power conversion system switching at 1 MHz. Regards, Dave
  • 0 in reply to David Johns
    Intellectual 410 points

    Seriously David, 1 000 000.00V @100 000.00W   and switching  frequency 1 000 000.00Hz? I am impressed! Give me more details, like: What will  be sources of energy you wont to convert  and  in which form is going to be output energy (  i mean DC voltage or maybe  electromagnetic/transmitter )?

    To produce triangle ramp you have excellent advice in previously post.

    Dragan

  • 0 in reply to dragan svalina
    Prodigy 40 points

    Dragan,

    We are the only power supply company licensed to produce the 'James Cross' high voltage transformer, subject of US Patent 6,026,004.  The transformer is a stack of cores and boards w/ the secondary windings in the boards, and a capacitor across each secondary winding to cancel the leakage MMF flux in the gaps along the length of the magnetic circuit.  James Cross ran this design well over 1MV.  To date, we've shipped systems up to 750 kV at 75 kW using two 35kW driving supplies (one passive rectified bus with four parallel phase shift modulators per 35 kW driver).  The transformer resides in a large pressure vessel at 5 ATM SF6.

    The math in the theory of ops in the patent requires a purely sinusoidal drive for the transformer for the flux compensation to work equally well up the XFMR stack.  Phase shift modulation, hard switched H bridge, Series Resonant, and other classic inverter topologies do not reproduce a harmonically pure sine wave at the transformer input.  I would like to try to drive this transformer stack w/ a high power Class D amplifier with a 100 kHz carrier wave, reproduced by a 1MHz switching stage.  I can multiplex an array of  power FETs to keep power dissipation down in the individual switches. 

    This is a pure R&D exercise right now.  I don't know if I can push so much current in a FET array with such short switching time, and I don't yet know how I will compensate the output Class D LC filter from zero load to full load conditions.  The first task is to come up with the PWM and control circuit.

    But, I figured out the PWM ramp.  It's a pseudo-triangle ramp set up with a high current driver charging a timing cap.  The voltage on the cap feeds into two comparators arranged as a window comparator.  The comparators set / reset a flip flop which in turn drives the high current driver.  I can adjust the DC offset and the peak to peak spread of the PWM, and thereby control my full scale signal swing from the overall error amp.

    -Dave

  • 0 in reply to David Johns
    Intellectual 410 points

    Hi David,

      I red " Patent 6,026,004."  and have better picture of project . Give me more time to study "patent" and e-mail address  for contact .  I have some experience with high power  resonant  amplifiers and maybe I can give you some useful advices. 

    Dragan


  • 0 in reply to dragan svalina
    Prodigy 40 points

    Cool. work email: djohns'at'kaisersystems.com

  • 0
    Genius 5430 points

    David;

    Check the TLC555 Timer; I think it will run at that frequency but I haven't looked at the data sheet for a while. Replacing the RC timing resistor with a constant-current source will give you a linear ramp or you can run the square wave output of a TLC555 in the asynchronous mode into an integrator to get a triangle wave. You may need another op amp to boost the amplitude.

    Regards, Neil  ex-Burr-Brown