Low Duty cycle generator

Discussion in 'General Electronics Chat' started by er_sjsu, Nov 2, 2010.

  1. er_sjsu

    Thread Starter New Member

    Nov 1, 2010
    5
    0
    Hi,
    I need pulse generator to Drive MOSFET. My requirements for duty cycle is Ton=1mS (5V,DC) and Toff=60 Sec (0V, DC).
    Can any one please suggest me the simplest possible solution to the problem.
    I am trying to avoid Timer and counter circuit.
    I want to use some kind of oscillator and schemit trigger.I am not sure how?
    Please help me.
    Thanks for every one in Advance.
     
  2. R!f@@

    AAC Fanatic!

    Apr 2, 2009
    8,754
    760
    So go for a transistor based ASTABLE.

    Obviously u know how to built this, if not try AAC database, or google
     
  3. jpanhalt

    AAC Fanatic!

    Jan 18, 2008
    5,698
    905
    Here:
    [​IMG]
    IC2A generates a very asymmetric pulse (i.e., short duty cycle). Don't know if you can get to 60 seconds for the OFF, but it's worth a try. Additional stages for buffer and for scope sync, if you want to see it. Original from Dekker used in his inductor test bench.

    John
     
    er_sjsu likes this.
  4. er_sjsu

    Thread Starter New Member

    Nov 1, 2010
    5
    0
    Thanks John,
    I'll try it and see how it works
     
  5. jpanhalt

    AAC Fanatic!

    Jan 18, 2008
    5,698
    905
    Please accept my apologies, if I am stating the obvious to you. Since you are new here, I don't have a clue as to your experience level. 1mS and 60 sec is a 60,000:1 ratio. R2 vs. R3+R4 controls that ratio. Because you are aiming for 60 seconds, you need a large time constant (i.e., R2 X C8+C9). I might increase R2 to 2.2 M, try R3 at 33 ohm and R4 10 ohm variable (or just use a 100 ohm variable for both). C2+C3 will need to be increased to (probably) 22 to 47 uF . One could calculate it closer, but the calculations don't take into account leakage during the slow charge through R2. BTW, You only need one capacitor. I used 2 to get the value I needed. I hope it oscillates for you. Do you have an oscilloscope?

    John

    Edit: All relaxation oscillators will have the same difficulty with the 60,000:1 ratio. You can add additional circuitry or go to an MCU, if needed. The simplest of MCU's will do this quite easily, if that is something you want to take on.
     
    Last edited: Nov 2, 2010
  6. er_sjsu

    Thread Starter New Member

    Nov 1, 2010
    5
    0
    John,
    Please refer the attachment, I made it work with simillar scheme on simulation tool but the component values are unrealistic so I am looking for another option.
    Thanks
     
  7. jpanhalt

    AAC Fanatic!

    Jan 18, 2008
    5,698
    905
    See my edit. All relaxation oscillators will have the same problem. Maybe an oscillator followed by a divide by 4 (etc.) would work? With an MCU (e.g., 12F508 etc.) it would be super easy.

    John
     
  8. Wendy

    Moderator

    Mar 24, 2008
    20,766
    2,536
    Instead of a single oscillator, think a slow oscillator (60 second) feeding a monostable (1ms). The 1ms would be much more precision.
     
  9. er_sjsu

    Thread Starter New Member

    Nov 1, 2010
    5
    0
    Bill,
    I already have slow oscillator (schmitt trigger) (60sec). Can I able to make monostable out of schmitt trigger or I need to use timer IC to make monostable?
    Can you please suggest me the simplest possible solution of monostable for my application?
    Thanks again.
     
  10. jpanhalt

    AAC Fanatic!

    Jan 18, 2008
    5,698
    905
    I would also ask, how accurate you need this thing to be? The simple, single oscillator will be affected a lot by leakage (i.e., the final resistors will be smaller than calculated) and leakage (diode and capacitor) will be affected by temperature.

    John
     
  11. Wendy

    Moderator

    Mar 24, 2008
    20,766
    2,536
    Amen to that. There are chips with crystal oscillator circuits that could do this with 3 chips that would be extremely accurate.

    Of course when the µC (micro processor) crowd see this we can expect other inputs.

    Rifaa was right about a simple 2 transistor multivibrator working too. All depends on the precision needed.
     
  12. er_sjsu

    Thread Starter New Member

    Nov 1, 2010
    5
    0
    Please review the attached waveforms. I almost got the desired results (still on Simulation tool). Do you think this circuit will work on actual bench?
    Now, the problem I have is I can't drive capacitive load with attached circuit. Do I need to add buffer stage to drive MoSFET?
    The reason why I am asking simple circuit because the circuit is part of control logic of Power supply design. so minimum power consumption and minimum and cheaper part count is first priority.
    Accuracy is not that important.
    Toff can be vary 5~6%. Its ok.
     
  13. jpanhalt

    AAC Fanatic!

    Jan 18, 2008
    5,698
    905
    If it works in practice, yes you can drive a mosfet with it, or a mosfet driver. That is what I used it for when I was interested in studying the saturation behavior of various inductors. To directly drive a mosfet, just parallel the unused gates as shown in the original schematic. In my case, I did end up using a simple mosfet driver though.

    John
     
Loading...