Charge pump for high side mosfet

Discussion in 'General Electronics Chat' started by shortbus, Feb 18, 2013.

  1. shortbus

    Thread Starter AAC Fanatic!

    Sep 30, 2009
    4,019
    1,542
    Will the modifications I've shown in the attachment be OK for a boot strap charge pump? The App Note (AN 978) it came from shows it only working when the mosfet source is turned on, and it gets its power from the source voltage. I want to power it from a available 12V on the board.

    Will it work or is there another way that's better?

    App Note 978 page # 18. http://www.irf.com/technical-info/appnotes/an-978.pdf
     
  2. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    How is the additional 12V source related to the other 12V and HV?
    If it's completely independent then why not just connect it to VB and VS? Then you wouldn't even need the 555.

    In the original circuit, when the IGBT is ON, you have HV over the load and therefore over the zener diode and the 100k resistor (- Vce of the IGBT).

    That means the whole 555 circuit voltage levels are near HV. You have HV - Vzener on the 100k/1W resistor. It is only active when the IGBT is ON.

    Your circuit will always be active. You used the same ground symbol which would mean HV RETURN and the 555s ground would be the same. Then the bootstrap capacitor will only be charged when the IGBT is OFF, through the load. But this happens anyway through the diode 11DF6 in the original circuit.
    You also could not use the 1N4148, it would need to be one with a voltage rating higher than HV. The 555 circuit is now near ground.

    Why do you want to change the circuit?
     
  3. shortbus

    Thread Starter AAC Fanatic!

    Sep 30, 2009
    4,019
    1,542
    Thanks for looking at my post. The reason I thought a charge pump was needed is that my circuit will be charging a cap bank and not getting source to ground even when off. Will add a schematic to the end circuit to show what I mean. It was in this thread - http://forum.allaboutcircuits.com/showthread.php?t=81184
     
  4. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    Yes, for the two n-FETs you need a high side driver. Any specific reason why you can't use p-FETs? Should be ok for your case, no high frequency switching except for fast turn on of the second FET.

    You also need to limit the current through the first FET when charging the caps.
     
  5. shortbus

    Thread Starter AAC Fanatic!

    Sep 30, 2009
    4,019
    1,542
    Thanks again praondevou! I knew about the current limiting just didn't show it in the sketch. The circuit will be switching in the 5kHz to 50kHz range, depending on the electrode being used.

    Another dumb question for you.:) Can the Nmos be connected like a Pmos? By that I mean with supply voltage on the source. Aren't mosfets really a bipolar switch? I know the intrinsic diode will conduct if connected backward, but will the mosfet work that way?
     
  6. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488

    Oops, 50kHz is too fast for the circuit I had in mind (I think).
    So you'd need the Nmos.

    Yes I think you can connect them backwards but then you have either current running through the body diode or through the Mosfet, i.e. it would always be on, somehow.
     
  7. shortbus

    Thread Starter AAC Fanatic!

    Sep 30, 2009
    4,019
    1,542
    Thank you for answering.:)
     
  8. kubeek

    AAC Fanatic!

    Sep 20, 2005
    4,670
    804
    Actually when you turn the moseft on you can completely bypass the diode drop, the mosfet works in both directions.
     
    shortbus likes this.
  9. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488

    Yes I agree with that, he just couldn't completely cut the current path, since the diode will conduct when the FET is off.
     
    shortbus likes this.
Loading...