Hbridge blowing up

Discussion in 'The Projects Forum' started by Shira, May 11, 2013.

  1. Shira

    Thread Starter New Member

    Apr 19, 2013
    16
    0
    Hello

    My task was to make a h-bridge for 100W, 24V brushless DC motor.
    The circuit should be able to draw 4A current from the supply. Also, the motor should be able to go in both forward and reverse directions with variable speed. The PWM that was given to the lower MOSFET was of around 15kHz.

    The circuit was working fine at first but during switchover from forward to reverse the PMOS that was turned off (Vg=15V) was heating up. Afterwards the PMOS started to blow up and I can't seem to understand why.

    Kindly help me out as soon as possible.
     
  2. Shira

    Thread Starter New Member

    Apr 19, 2013
    16
    0
    A better Image of my schematic
     
  3. crutschow

    Expert

    Mar 14, 2008
    13,056
    3,245
    You shouldn't reverse the motor when power is applied. Shoot-through current when both MOSFETs are momentarily on during the transition can zap a transistor.
     
  4. Shira

    Thread Starter New Member

    Apr 19, 2013
    16
    0
    So before changing direction I have to disconnect the power supply?
    How can it be incorporated in this circuit?
    Any suggestions?
     
  5. kubeek

    AAC Fanatic!

    Sep 20, 2005
    4,670
    804
    Simply turn Q5 off before you change the direction.
    Also, why are you using the discharge pin of the 555 instead of output to control Q5? Any way, 10k is way too large to turn it on fast, big mosfets have lots of gate capacitance, and slowly turning it on will cause lot of heat dissipated.
     
  6. Shira

    Thread Starter New Member

    Apr 19, 2013
    16
    0
    Without the resistors on the gate the PMOS having Vgs=15V is still blowing up. For switching the direction I am using a dpdt switch that has a center position that makes all the gates disconnected (NC) so is it still necessary to switch off the power supply?
     
  7. Shira

    Thread Starter New Member

    Apr 19, 2013
    16
    0
    Simulating it has no errors at all but when I tested the circuit on a breadboard none of the MOSFETs blew up although the PMOS did get hot. So I made a pcb and the PCB's first testing was fine but when the pcb was tested again (in front of our project manager) the PMOS just blew up...=[
     
  8. strantor

    AAC Fanatic!

    Oct 3, 2010
    4,302
    1,988
    Which 2 of the 3 wires from the BLDC motor are you connecting to your H-bridge?
     
  9. Shira

    Thread Starter New Member

    Apr 19, 2013
    16
    0
    The DC motor has only 2 wires. One red, the other is black.
     
  10. Shira

    Thread Starter New Member

    Apr 19, 2013
    16
    0
    Should I add a snubber circuit?
    Also I read somewhere that the MOSFETs heat up only when they are not fully turned on or turned off. Meaning that they are not going in the cut-off and saturation region and operating in the triode region. So is Vg=15V wrong?
     
  11. Shira

    Thread Starter New Member

    Apr 19, 2013
    16
    0
    @ kubeek
    The square wave from the discharge pin is much better than the one from the output...as observed on a CRO.
     
  12. Shira

    Thread Starter New Member

    Apr 19, 2013
    16
    0
    Can somebody guide me why only the PMOS (that is supposedly turned off) is blowing up?
     
  13. kubeek

    AAC Fanatic!

    Sep 20, 2005
    4,670
    804
    Which one? Q5 or Q1 or Q2?
    I think I see the problem. The bottom mosfets have Vgs 15V, but when you switch to the opposite polarity the top mosfets get 24V, which is above the Vgsmax= +/-20V. Use a 15V zener across each G-S and you can also get rid of the separate 15V source.
    Ok you can´t beacause it is supplying the 555, but still it is a much cleaner and also safer solution.
     
    Shira likes this.
  14. Shira

    Thread Starter New Member

    Apr 19, 2013
    16
    0
    Q1 and Q2
    Q5 is an NMOS
     
  15. Shira

    Thread Starter New Member

    Apr 19, 2013
    16
    0
    Please bear with me because I don't fully understand it.
    When Vg=0V, the PMOS has Vgs=-24V so it is above the max range.
    When Vg=15V, the PMOS has Vgs=-9V, which is within the range but even then the PMOS blows up ionly in this condition.
     
  16. kubeek

    AAC Fanatic!

    Sep 20, 2005
    4,670
    804
    Put some resistor in series with the 24V supply like 100 ohms or something, put in new mosfets adn the zeners and start measuring voltages and currents, because the mosfet should not blow so something must be different from the schematic.
     
  17. Shira

    Thread Starter New Member

    Apr 19, 2013
    16
    0
    Okay, thank you.
    Will buy the zeners on Monday and start testing then.
    =]
     
  18. ScottWang

    Moderator

    Aug 23, 2012
    4,856
    767
    Assuming the situation as below according to the circuit.
    The circuit should be redesign, because when the S1,S2 provided a 15V high voltage for the g of Q1 and Q3 and then the Q3 wil be on(sat), the Vgs of Q1 has 24V-15=9V, it means that the Q1 is still working, but not in the sat status.

    The Vgs of Q2 is 24V(too high), so the Q2 will be on(sat), and Q4 will be off.

    When the S1,S2 provided a zero voltage for the g of Q1 and Q3, and then the Vgs of Q1 is 24V(too high) and on(sat), the Vgs of Q2 has 24V-15=9V, it means that the Q2 is still working, but not in the sat status, the Q3 will be off.

    The Vgs of Q4 is 15V(a little higher), so the Q4 is on(sat).

    I'm not sure what time you will push the S1, S2 switches, but it could be damage the mosfets as above description.

    Another situation is when S2 switch to GND, but I haven't describe yet.

    You should redesign the circuit, and you should be care about the Vgs of mosfet, maybe you could using the resistors to do the voltage divider, or adding the bjt to solve the problems.

    The values of R2,R13 are too small, and the better values of voltage for Vgs is 12V>=Vgs>=10V.
     
    Shira likes this.
  19. Shira

    Thread Starter New Member

    Apr 19, 2013
    16
    0
    Oh, I get it.

    But do explain the S2 connected to ground part again.
     
  20. Shira

    Thread Starter New Member

    Apr 19, 2013
    16
    0
    So by this : 12V>=Vgs>=10V
    you mean to say that instead of connecting s1,s2 to 15V and ground I should use 12V and 14V?
     
Loading...