H-bridge motor driver need help here

Discussion in 'The Projects Forum' started by ericyeoh, Aug 31, 2009.

  1. ericyeoh

    Thread Starter Active Member

    Aug 2, 2009
    58
    0
    Please take a look at this link,
    http://www.robotroom.com/BipolarHBridge.html

    If i exchange the Transistor of PNP into NPN.
    The operation is change into

    Forward
    R1= GND, R2= Vcc, R3= VCc ,R4 =Gnd
    Reverse
    R1= Vcc, R2= Gnd, R3= Gnd ,R4 =Vcc

    The problem, i facing at this moment is that when the transistor is on, The transistor cannot endurance for long period due to the heat. The transistor that i been using in the circuit is NPN(2N2222A). The resistor connect to the each of the Transistor is 1k ohm.
    The Voltage supply to this circuit i use is 5V, the input volatge to each transistor also 5v(when need it).

    But i exchange the resistor into 100 ohm to each Transistor(base), it seems to be last long a bit, but also cannot endurance for long period, the transistor will heat up and burn....

    I need help here. please reply the solution step by step tq.
     
  2. Wendy

    Moderator

    Mar 24, 2008
    20,766
    2,536
    Try posting a schematic, it will help with the explanation.

    From what little I read, and think I understand, you don't quite understand saturation and why it is important. The two most efficient states for a transistor in digital mode is on and off. When a transistor is fully on (and is in common emitter mode, which is what is shown in the link you gave) it only drops around 0.1 Volts between Collector to Emitter. The other mode, Common Collector, which is what I think you are talking about, the CE drops 0.6. This can make a huge difference in heating.

    We'll also need to know how you're driving them.

    Check out my profile and AAC blogs, I think you'll find I have some background in the subject.
     
  3. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    In other words, the PNP transistors need to stay PNP.
    You cannot simply change them for NPN transistors and expect it to work.
     
  4. ericyeoh

    Thread Starter Active Member

    Aug 2, 2009
    58
    0
    Actaully it can work, but it cannot last long due to the heat.
    I'm using PIC microcontroller to drive it.
    The input of each (base)transistor is control by the output of PIC microcontroller, the PIC microcontroller output is rather 5v.
     
  5. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    OK, so if you replace the upper transistors with 2N2907 PNP transistors, your heating problem should be much reduced - but that also depends upon how much current your motor will draw when supplied with 5v.
     
  6. bertus

    Administrator

    Apr 5, 2008
    15,649
    2,348
    Hello,

    Did you read the contrilong part?

    Greetings,
    Bertus
     
  7. ericyeoh

    Thread Starter Active Member

    Aug 2, 2009
    58
    0
    Well, original i use PNP transistor at R4 and R2, it works very well, there is no problem at all to the heating issue, but in order to connect to PIC to let it function, PIC never supply 0v or GND, it is just disconnencted or 5v.

    To overcome this problem i change the design abit i try it by adding 2 NPN transistor to the base of the PNP transistor picture as shown below, this was advised by my friend.

    but after i tried, the motor is not working and which ever supply 5v input to base (NPN)transistor is hot.
     
  8. ericyeoh

    Thread Starter Active Member

    Aug 2, 2009
    58
    0
    IF the motor is forward condition the current in the circuit it might look like this.
    please tell me/say out if my theory is wrong!

    And please tell me wheather this circuit can be work or not. It is just theoretical solution.
     
  9. bertus

    Administrator

    Apr 5, 2008
    15,649
    2,348
    Hello,

    If R4 is also high as in your schematic, the transistor on R6 will also be active.
    The current will flow from the powersupply to ground, heating the transistors.

    Greetings,
    Bertus
     
  10. jj_alukkas

    Well-Known Member

    Jan 8, 2009
    751
    5
    Your H-bridge needs inputs which is either High or low. It cant be floating. When you change transistors from PNP to NPN, the bridge enters a state in which either or both the inputs are floating, resulting in the heat and short circuit.
     
  11. Wendy

    Moderator

    Mar 24, 2008
    20,766
    2,536
    Actually it is worse than that. Even the transistion between levels can damage the transistors.

    The H bridge is a digitial circuit, and connecting the inputs of the transistors is usually manditory. Generally circuit are added to prevent both transistors being on at the same time, a condition called shoot through. Sometimes some shoot through is unavoidable, but it is never good.

    Case in point, this is a true H bridge, even though it is driving a speaker instead of a motor. The zeners try to prevent the transistors being on at the same time, and both transistors are off during the transistion from high to low, or visa versa.

    [​IMG]
     
  12. Wendy

    Moderator

    Mar 24, 2008
    20,766
    2,536
    BTW, those diodes (D1-4) don't really do anything for drive. They absorb back EMF spikes from the motor and prevent it from damaging the transistor. Otherwise they don't exist in the current paths.

    The whole point of a H-Bridge is to turn one transistor (say Q1) on while the other is off (Q2). The other side can be inverted, but if it is on the same state then the motor (or speaker) gets no voltage at all.
     
  13. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    The PIC should be able to source or sink up to 20mA per I/O pin, meaning that it can output nearly 5v (Vdd) or 0v (GND), except for the MCLR pin. There is also a total current source/sink limit; you will have to consult the datasheet for the particular PIC you are using to determine that.
     
  14. jj_alukkas

    Well-Known Member

    Jan 8, 2009
    751
    5
    Is there any reason for you to not try an ordinary H-bridge? Build your H-bridge the way you can, then adapt the PIC's output for that as H-bridges give you only limited options for construction and connectivity whereas a PIC port can be adapted, inverted, amplified or switched.
     
  15. ericyeoh

    Thread Starter Active Member

    Aug 2, 2009
    58
    0
    finally i find out the way already... the schematic diagram is shown below...

    Please give some opinion!! thx ya...:D
     
  16. hobbyist

    Distinguished Member

    Aug 10, 2008
    764
    56
    I don't think it will work because
    Your 2 PNP driver transistors are connected from base to +Vcc through resistors R5 and R6 and the collector of your 2 inverters are connected to +Vcc. instead of to the base of your drivers.

    If you put resistors between the collectors of the top 2N222 transistors and the +VCC then it would work to that point.
    Or just connect the collectors directly to the base of there drivers.
    That would work like an inverter than.

    You have the proper concept just need to get the components in the right place

    Here is a H-bridge driver of been breadboarding of late for my next robot project, see if you can use any ideas from this to get you on the right track.

    .See if this helps. The Back EMF diodes are not in this schematic, but can be implemented as needed.

    robot motor controle H-bridge for small hobby motor..jpg
     
    Last edited: Sep 11, 2009
  17. aboamal

    Member

    May 29, 2008
    14
    0
    try this circuit

    [​IMG]
     
  18. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Aboamal,
    Your circuit has no current limiting resistors for any of the bases. This will result in excessive current and a good deal of smoke.

    Suitable current-limiting resistors would need to be inserted at nodes 1, 2, 5, and 7.
    270 Ohms would be about right for nodes 1 and 2 (bases of Q5, Q6) if the drive source is a PIC with a Vdd of 5v.
    A 2N3904 would be a limiting factor in saturating the TIP41/TIP42 pairs. A 2N2222 would be a significant improvement, just as available and about the same cost.

    85 Ohms would be about right for nodes 5 and 7 (bases of Q1, Q2) for a Vcc of 30v. They would have to be rated for 15 Watts or more. If you stayed with the 2N3904, you'd need to use 270 Ohm resistors rated for 5w.
     
  19. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Hobbyist,
    I'm afraid that the H-bridge circuit you posted would give very poor performance indeed.

    You have 10 MEG resistors as pull-downs for the MOSFET gates, and 1 MEG resistors in the gate charge path. It would take a long time to charge the gates fully, and ten times as long to discharge them. While the MOSFET gate is slowly charging/discharging, the MOSFET will be operating in the linear region, and generating heat.
     
  20. hobbyist

    Distinguished Member

    Aug 10, 2008
    764
    56

    Thankyou very much,
    I'll have to look into that, and work with the data sheets, specs , ect... this will be a good excersise in designing with mosfets, great information about the capacitance, that's something I never knew about, this is a great forum to learn from.

    Thanks again..
     
Loading...