how to reduce switching ime

Discussion in 'General Electronics Chat' started by justmailmahesh, Aug 21, 2009.

  1. justmailmahesh

    Thread Starter New Member

    Aug 11, 2009
    9
    0
    hi all
    I am using an IGBT IRGPC50U in my circuit and for this i have designed driver. And i am getting a 5V signal (pulse width 1ms and frequency 25hz) from optical fiber reciever HFBR 2531 and i am amplifing this voltage to 15V using an transistor switch and this 15 is being given to IGBT through push pull circuit. But i am getting around 10us delay from base to collector of amplifying transistor and so from signal to IGBT.
    When i simulate this circuit in orcad it was giving less than 1us delay but when i assemble it practically it is giving around 10us delay which is not tolarable.
    I have attached my orcad circuit and practical waveform by CRO.
    Plz help how can i reduce this delay
    Thank you in advance
     
  2. n9352527

    AAC Fanatic!

    Oct 14, 2005
    1,198
    4
    That channel 1, is it the input signal or the signal at the base of the IGBT?
     
  3. justmailmahesh

    Thread Starter New Member

    Aug 11, 2009
    9
    0

    Channel 3 is the output of CD4049 inverter i.e. voltage at the input of R5 i.e. signal and channel 1 is voltage at collector of transistor BC547.
     
  4. justmailmahesh

    Thread Starter New Member

    Aug 11, 2009
    9
    0
    i think normally transistor does not give delay but here it is giving 10us and what is the reason i am not understanding. If any one give any suggestion how to reduce this delay.
     
  5. peranders

    Well-Known Member

    May 21, 2007
    87
    0
    The switch times aren't that slow. Have you checked any application notes for IGBT drivers? If you use a driver (IC) you will have much more switching power with very few parts. Check for instance Intersil and International Rectifiers for drivers.
     
  6. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    Reduce the value of R1 to 100 ohms, or even zero. Reduce the value of R4 to 1K. Or use a driver IC, as suggested by peranders.
     
  7. justmailmahesh

    Thread Starter New Member

    Aug 11, 2009
    9
    0
    Actually i have made PCB so i can make some small changes but cannot use driver. iReduce the value of R1 to zero and R4 to 1K but it is not showing any effect in delay. Instead when i used QN2222 transistor in place of BC 547 the delay got increased from 10us to 15us.
    So is this delay is due to transistor only i am somewhat confused. If this so then plz let me know any easily available fast transistor.
    Thank you in advance
     
  8. justmailmahesh

    Thread Starter New Member

    Aug 11, 2009
    9
    0
    And another thing i think that this delay may be due to some capacitive effect near transistor but as i have not used any capacitors it may be soma parasitic capacitance . If it is so then is there any method to eliminate this parasitic capacitance.
    Thank you
     
  9. rjenkins

    AAC Fanatic!

    Nov 6, 2005
    1,015
    69
    I'm guessing it's the low-to-high edge (at the output) that is delayed?

    If so, it's saturation effects in the BC547.

    You need to control the drive to load ratio to prevent saturation for optimal switching, or you can just cheat and put a small schottky diode from collector to base on the transistor (anode to base, cathode to collector).

    That's how LSTTL ICs work; when the collector get down to around 0.2V it diverts current from the base and stops saturation.

    Just as a note, it's always good practice to have a resistor between base and emitter on a bipolar transistor, it bypasses any collector leakage and gives a more positive turn-off.
     
Loading...