Auto Battery Polarity

Discussion in 'Electronics Resources' started by Don P, Sep 17, 2014.

  1. Don P

    Thread Starter New Member

    Sep 17, 2014
    5
    1
    I am trying design an auto polarity for a 9V battery. This allows the user to insert the battery either direction and the circuit will compensate and switch the polarity. I found this H-Bridge design that works some of the times. However, every now and then, the mosfet will smoke and I am not sure if it is caused by a race condition in switching the mosfets.

    Does anyone have any idea of a better design or a modification to this design to stop the mosfet from smoking. :). Attached is the schematic.
     
  2. #12

    Expert

    Nov 30, 2010
    16,321
    6,818
    Just build a full wave bridge out of Schottky diodes.
     
    rogs likes this.
  3. Don P

    Thread Starter New Member

    Sep 17, 2014
    5
    1
    I looked at this design and it is simple and sweet. I was worried about the voltage drop using the diodes. I am hoping for every last bit of juice from the battery and went with the MOSFET. As for the mosfet, I think I might of stumbled onto something. I don't have any inline resistor to the gates and on another forum they mention that the mosfet gates are extremely sensitive and any transient spike will fry the mosfet gate to close position causing that will result in a short. I am going to try the above circuit with inline resistor and will report back.
     
  4. BR-549

    Well-Known Member

    Sep 22, 2013
    2,004
    394
    Is this a 9 VDC transistor battery? The little rectangular unit with polarized connectors? This is new. Have you ever connected a 9 volt battery backwards? Did it do any damage? A most curious situation. Is the power receptacle plug out of view? Is the battery difficult to retrieve and flip? Of course #12 has the proper solution. But with a polarized plug and a protection diode....what is the need?
     
    panic mode likes this.
  5. Don P

    Thread Starter New Member

    Sep 17, 2014
    5
    1
    Yes, it is a standard transistor battery. We are not using standard 9v battery clips (with different sizes for +/-), but springs in handle that the battery drops in. Even though we have a label for polarity and a diode for protection, the customer is wanting to put the battery in either way and have it work. That is the need, to meet the customer wishes.
     
    BR-549 likes this.
  6. #12

    Expert

    Nov 30, 2010
    16,321
    6,818
    BR-549 likes this.
  7. crutschow

    Expert

    Mar 14, 2008
    13,028
    3,237
    You can use a 10k ohm resistor or larger in series with each pair of gate connections, since the switching speed of the bridge is of no consequence.
     
  8. inwo

    Well-Known Member

    Nov 7, 2013
    2,433
    315
    Isn't that per diode?
     
  9. #12

    Expert

    Nov 30, 2010
    16,321
    6,818
    Yep. You caught me in a mistake. I assumed the curve was for the device they are selling, not each part inside the device. :(

    At 20 ma, it's about .175V per diode...0.35 volts per bridge.
    Definitely something you can beat with mosfets.
     
  10. MrChips

    Moderator

    Oct 2, 2009
    12,447
    3,363
    Keep it simple. Go with a Schottky bridge.
     
    rogs likes this.
  11. alfacliff

    Well-Known Member

    Dec 13, 2013
    2,449
    428
    a possible reason your mosfets smoke is that the source to gate juction has a large amount of capacitance, you must discharge it when the thing is off, otherwise the mosfet s still turned on till it discharges.
     
  12. Don P

    Thread Starter New Member

    Sep 17, 2014
    5
    1
    This makes a lot of sense. See the proposed schematic. Is this the concept? I am concern about leakage with the resistance. Would should the Rgs be for 9V battery? I seen some comments recommending 47K, but with 4 gates, this is serious leakage and will drain the battery quickly.
     
  13. crutschow

    Expert

    Mar 14, 2008
    13,028
    3,237
    Edit:
    A possible source of failure is high shoot-through current as the bridge is turning on. My simulations showed a very high current when the bridge initially had power applied to it due to two MOSFETs in series being momentarily on. A small resistor (1 ohm) in series with the battery will minimize this surge. Also reduce the value of the resistors in series with the gate to perhaps 100 ohms or so to reduce the duration of the shoot-through.
     
    Last edited: Sep 18, 2014
  14. crutschow

    Expert

    Mar 14, 2008
    13,028
    3,237
    I don't see "this H-Bridge design".

    Marmosets are small new world monkeys. :confused:
     
  15. Don P

    Thread Starter New Member

    Sep 17, 2014
    5
    1
    All,

    Thanks for the help. I have built up a few different designs and the attached is what I am going to go with. One think that slip through the crack was the on the FDC6305 / FDC6312 MOSFET, the Vgs is max at 8V. I was driving this with 9+ which exceed the mgf limits.

    I selected a higher power MOSFET, might be an over kill, but, seems to work. I am using a Complementary MOSFET with N & P channel sharing a common drain (AON3613). This MOSFET has max Vgs of +/- 20V. I added the 1 ohm resistor in series with the battery and added zener diodes (15v) BZV55C15 to protect the Vgs to not exceed it maximum.

    I have pounded this circuit with variety of voltage and spiked the voltage and it seems to hold up. I have yet to smoke the mosfet.

    Thoughts?
     
Loading...