design of flyback converter issue

Discussion in 'The Projects Forum' started by dark night, Feb 26, 2012.

  1. dark night

    Thread Starter New Member

    Jan 20, 2012
    5
    0
    hi alll
    i am finding a problem with timer 555 IC.....when i run the simulation the output of the timer disconnect..plz help me solve this problem...find my attached circuit diagram..

    thnx in advanced...
     
  2. crutschow

    Expert

    Mar 14, 2008
    13,039
    3,243
    It appears the simulator thinks the 555 is being overloaded. I see no obvious reason for that.

    What simulator are you using?
     
  3. DickCappels

    Moderator

    Aug 21, 2008
    2,659
    632
    There is something to be said for the realism a physical model has to offer.
     
  4. bountyhunter

    Well-Known Member

    Sep 7, 2009
    2,498
    507
    At least I lived long enough to hear somebody besides me point that out.

    We are raising a whole generation of engineers who seem to be terrified of electrons and soldering irons.

    I designed and built a 9V - 30V boost converter for one of my scope calibrators. You do have to actually build things sometimes to see how they work.

    I don't know why the sim is barking at the 555 output, but that PNP/NPN arrangement will have high shoot through current driven that way.
     
    Last edited: Feb 26, 2012
  5. MrChips

    Moderator

    Oct 2, 2009
    12,449
    3,364
    "But it worked ok on the simulator" - is what I get from my students all the time.
     
  6. bountyhunter

    Well-Known Member

    Sep 7, 2009
    2,498
    507
    Tell them they got a simulated A on the test.......:p
     
  7. Nicholas K. Heinrich

    Member

    Feb 25, 2012
    61
    1
    I plan to become an engineer, and am only 17, and I would agree that a lot of engineers I see are afraid to actually build things. I sometimes use the elctronics depatment at our career center to design and build something I need. Nobody seems to notice me while im designing something on the computer, but once I pick up a soldering iron and start attatching components, all the students are like " OMG, he's actually building something!" The looks on their faces are priceless.
     
  8. #12

    Expert

    Nov 30, 2010
    16,337
    6,820
    Electrons are never wrong.

    Applying electricity and seeing the circuit work is the only undisputable proof that you did it correctly.

    There. That sounds much better than displaying my attitude about simulators.
     
  9. Nicholas K. Heinrich

    Member

    Feb 25, 2012
    61
    1
    Exactly #12
     
  10. crutschow

    Expert

    Mar 14, 2008
    13,039
    3,243
    I always simulate a circuit before I build it because it finds a lot of my dumb mistakes, and allows me to easily tweak the design, try different circuit variations, or see the effects of part tolerance. But you have to build the circuit to prove that it works. It usually works close to the simulation, but not always.

    I remember one time simulation saved me a bunch of redesign. I was designing a window comparator with a couple of op amps but the simulation did not seem to perform properly, with my set points being offset from my calculated values. After checking the op amp data sheet more closely I realized it had internal diodes across the two inputs for over-voltage protection and they were causing feed-through that was affecting the set points. After redesign to allow for this the simulation worked properly and so did the circuit when I actually built it.
     
  11. crutschow

    Expert

    Mar 14, 2008
    13,039
    3,243
    If you simulate the circuit you will see there isn't any shoot-through since there is an input dead-band of about 1.4V between the turn-off of one transistor and the turn-on of the other. ;)
     
  12. bountyhunter

    Well-Known Member

    Sep 7, 2009
    2,498
    507
    If you build it you will see there probably is shoot through, it depends on how fast the drive signal is switching. Remember the bipolar transistors have stored charge and stay ON for a period of time after the drive voltage is taken away and the other transistor is turned on (They turn ON way faster than turning off, hence the problem both transistors ON at the same time for a short interval). This shoot through problem is classic for this type of driver, how bad depends how fast you switch the drive voltage up and down.
     
    Fshah likes this.
  13. crutschow

    Expert

    Mar 14, 2008
    13,039
    3,243
    If it were the classic push-pull driver with the NPN on the bottom and the PNP on top, what you say is true. But these transistors are operating as emitter followers and never saturate, thus do not exhibit stored charge delay. That's the principle of ECL and why it's so much faster than TTL. So if you simulate it or build it you will likely see no overshoot either way.
     
  14. Brownout

    Well-Known Member

    Jan 10, 2012
    2,375
    998
    Back to the topic... that's really bizzare. I've never seen anything like it before. I wonder if it's telling you that you need some base resistors for the transistors. As the output is configured now, the bjt's probably won't ever turn off all the way.

    I suggest you try a different simulator, like LTSpice.
     
  15. bountyhunter

    Well-Known Member

    Sep 7, 2009
    2,498
    507
    Which would cause a problem like shoot through current, ie one still being on when the other one turns on.....

    I was going to suggest the OP might want to get a protoboard and build the thing but that would take a good two minutes so it's apparrently not feasible. It will forever remain a mystery.... but to build it, I would add a 1K resistor from the emitter to base of either transistor.
     
Loading...