Mosfet h-bridge

Discussion in 'General Electronics Chat' started by us202000, Jun 3, 2012.

  1. us202000

    Thread Starter New Member

    Jul 5, 2011
    13
    0
    I am trying to make a simple h-bridge with all n-ch mosfet and would like to get your comments and suggestions. Attached is a draft circuit for your review. The high side uses bootstrap and pwm input. I plan to use a microcontroller to control each of the mosfets. The mosfets could be something like IRF3205 or IRFZ44n. Will this work? Any comments? Thanks.
     
  2. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    There will be increased switching losses due to the high gate resistor values for the upper FETs.

    There is nothing to charge the lower FET's gates.

    Maximum gate voltage for the 3205 is +-20V. C1 is charged to 24V when Q4 conducts.
     
  3. us202000

    Thread Starter New Member

    Jul 5, 2011
    13
    0
    Thanks for your comments. You are right the gate voltage is too high. How about adding a resistor between the gate and source on the high side? Will this reduce the voltage? I redraw the left half in the attached. Hopefully it also fixed the problem in the lower half. Is this any better? Also note I changed the location of s1 on the high side.


     
  4. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    Now the gate voltage for the upper FET would not be exceeded. For the lower it would be a little more than 18V.

    The gate resistors are too high. You cannot expect to be able to drive the motor with a few kHz PWM switching frequency. The FETs are going to dissipate too much heat.

    this may help you: http://www.ti.com/lit/ml/slup169/slup169.pdf
     
  5. us202000

    Thread Starter New Member

    Jul 5, 2011
    13
    0
    Thanks again. I will have to study the document you recommended.

    Is the issue lowering the value of the resistors (say to 100 ohms, or this type of design for high side doesn't work at all for motors? Previously, I was able to use the same gator resistor value when used on the low side of the motor for PWM switching. Now I am confused. Please advice. Thanks.
     
  6. us202000

    Thread Starter New Member

    Jul 5, 2011
    13
    0
    Based on the article I read, I changed the gate resistor to 50 ohms (or could remove it), see attached. I remember I borrowed the 1k gate resistor (and slow rise time) from audioguru's old post (http://www.electro-tech-online.com/...channel-mosfet-driver-bootstrapped-mosfet.png) where the resistor value can't be too low in that design. I also changed the power to 12v (hard to make it work for 24v). However, I am not sure about the fall time in this design. Will this work?
     
  7. praondevou

    AAC Fanatic!

    Jul 9, 2011
    2,936
    488
    Well the fall time depends on the resistor to discharge the gate. That's R2 and it's very high. You can't lower it to much in this configuration though.

    A dedicated FET driver IC (can also be built with discrete components) would be better.
     
  8. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,292
    1,255
    Keep in mind with the boot strap you cannot run a 100% duty cycle.
     
  9. shortbus

    AAC Fanatic!

    Sep 30, 2009
    4,019
    1,540
    Unless you use a 'charge pump' circuit.
     
  10. us202000

    Thread Starter New Member

    Jul 5, 2011
    13
    0
    The input is PWM at a few kHz but the worry is that the gate may not turn off due to the slow fall time before the next cycle. What's the estimated fall time with a pull down R2=1k resistor for IRF3205?
     
  11. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,292
    1,255
    Pretty slow. A little over 12 usec. Will your isolator let you use a smaller resistor?
     
  12. shortbus

    AAC Fanatic!

    Sep 30, 2009
    4,019
    1,540
  13. us202000

    Thread Starter New Member

    Jul 5, 2011
    13
    0
    rov: yes, I could lower the resistor values, although I don't know what the Rgs low limit is. I haven't seen any Rgs values below 1k.

    shortbus: thanks for the tip. I was debating whether to use IR2184 or IR2110. Which one is a newer chip?

    Thanks.
     
  14. shortbus

    AAC Fanatic!

    Sep 30, 2009
    4,019
    1,540
    While in the same chip family all of the IR21xx chips have different attributes for different circuits. Though the IR2110 started it all. The IRF series is the newest though. To learn more about them you'd have to look on the IRF site - http://www.irf.com/product-info/cic/fsgatedriverics.html
     
  15. us202000

    Thread Starter New Member

    Jul 5, 2011
    13
    0
    I saw a couple of designs using IR2184 but also have opto-isolators to the microcontroller. So how important is it to add opto-isolation between IR2184 and microcontroller?
     
  16. shortbus

    AAC Fanatic!

    Sep 30, 2009
    4,019
    1,540
    Don't know what your making, but something I'm familiar with can maybe shed some light. In home shop CNC machines they some times use a "break-out board" between the computer and motor drivers. Not so much in the newer designs though. They have found out that the gate driver chips are so effective at protecting the computer that optos aren't used in the break outboards any more.

    The removal of the optos has the benefit of allowing faster switching to the motors, increasing their speeds. But you do what your comfortable with.:)
     
  17. RRITESH KAKKAR

    Senior Member

    Jun 29, 2010
    2,831
    89
    May this help.........
     
  18. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,292
    1,255
    Sometimes the drivers are used without isolation from the power line. In those cases the opto provides a level of protection against having power line voltages on your circuits.
     
Loading...