H-Bridge Problems

Are you running a program on the Arduino to make sure the outputs are not turning on all the transistors at once before you connect the 12V ? If not this could

produce a short between 12V and ground across the transistors which would cause problems. Are you using the Arduino PWM library ? If I remember correctly

the Arduino PWM library code turns the outputs off for the programmed time interval (outputs start out high) which is inverted from what you would normally

expect. So if you start out with the lowest value for the PWM time (0x00) the outputs would always be on.

Here was the original code. Control1 was connected to Q5, Control2 was connected to Q1, Control3 was connected to Q6, and Control4 was connected to Q2. Although I connected each transistor gate to a PWM pin on the Arduino, I used digitalWrites in my code.

int control1=3;

int control2=5;

int control3=6;

int control4=10;

void setup()

{

  pinMode(control1,OUTPUT);

  pinMode(control2,OUTPUT);

  pinMode(control3,OUTPUT);

  pinMode(control4,OUTPUT);

}

void loop()

{

  digitalWrite(control1,LOW);

  digitalWrite(control2,LOW);

  digitalWrite(control3,HIGH);

  digitalWrite(control4,HIGH);

  while(1);

}

Sorry I had to adjust myself to you using P and N channel MOSFETS. I am more used to integrated solutions that normally just use

N-channel with charge pumps for the the high side gate voltages. Aren't P-channel transistors normally used on the high side and

N-channel on the low side ? The program looks like it should be working to turn the motor on (Q1 and Q6 on) just make sure its

running before you connect the power or the motor.  With pull down resistors on the P-channel inputs the transistors will be on if

the Arduino is not actively pulling them high. I'm not sure if this is what you want or not.

I have the P Channels on the high side, and N channels on the low side. I thought that by setting pull down resistors on all of the transistors, the motor will constantly be off until the Arduino apply voltage at the gates.

Hi David,

Unfortunately the circuit as shown won't work, because as you say, the P-ch should be at the top, and N-ch at the bottom, however with MOSFETs, the arrow sticking out signifies P-ch. Basically, the top and bottom MOSFETs need swapping round.

Once that's done, have R2 and R3 going to the same supply as the motors, not to ground. Then, ensure that the microcontroller outputs are 'open drain' types for the top two devices (personally I would use a small resistor in series with the gate too - not too high, maybe 47ohms or so.

This should get you some way toward getting this going. There are further refinements you could do, but the adjustments described above will get it working. Personally I would stick an inductance in series with the motor too.

The built-in diodes should be ok, although not all devices specify if they will be fine (so some people use external schottky diodes too). I think you may be fine.

I thought by setting pull down resistors on each gate sets the transistor initially off. Also, I was using 1/16 watt resistors. If I have to use pull resistors connected to a 12 9AH battery, I will need to replace the resistors with higher power rated resistors.

I'll make sure to try this out and see what happens. Thanks guys!

Hi David,

The gates consume no current (to a first-level approximation) so 1/16W resistors are fine. (Besides, the high resistance (10k) means that 1/16W won't be exceeded).

This circuit may still not be ideal  - the N-ch MOSFETs may not fully turn on (the motor will work, but may not run at full power). This one would be more ideal, and

is in a TO-220 package too.

Hi David,

The gates consume no current (to a first-level approximation) so 1/16W resistors are fine. (Besides, the high resistance (10k) means that 1/16W won't be exceeded).

This circuit may still not be ideal  - the N-ch MOSFETs may not fully turn on (the motor will work, but may not run at full power). This one would be more ideal, and

is in a TO-220 package too.

This is  a really bad motor driver circuit. Even if you make the corrections others have suggested (P channel FETs at the top, R2 and R3 to +12V) then the Arduino outputs, even if open drain, are not rated for 12V so you'll damage the Arduino. If you get round that by buffering the Arduino outputs then you still have the issue that there is no dead time (time when all the FETs are off as you switch) and no "shoot through" protection (a way of ensuring that a top FET and a bottom FET are never both turned on at once). On top of that you need more gate volts (10 not 5) to turn on the FETs properly and you should have schottky diodes across each FET to catch reverse voltage when switching off. Then you need some capacitors across the power supply.

You can get almost everything on one chip but it's in a "difficult" package - ST VNH3SP30TR (Farnell 1909372). An easier part but only just up to your spec is the Infineon TLE5205-2 (Farnell 2215546) in a TO 220 package.

MK