I am trying to build a controller for a DC gearhead motor: 3V to 18V compatible, draws less than .5A with no load. I don't need speed control, but I do need to reverse direction. I plan to use a 5V μC for the logic, and 12V for motor power. Please comment/correct my schematic. Thanks.

 

I would add some clamp diodes.
I'd kind of worry about shoot thru - where the top transistors turn on before the bottom one is fully off. Both would be a lot worse if you were pwming it.

 

I agree- there will be heavy shoot-thru currents during the switching interval.

Clamping diodes are inherent in the FET's - the body diode.

 

If you turn off the power before you reverse direction, there shouldn't be any problems. If you try to reverse with power applied, then there will be stress on the motor and FETs.

 

Thanks for the input.

I should have said in my original post that the application is (yet another) automatic door opener/closer for a chicken coop. The power will always be off before the motor is reversed.

So, am I correct in my understanding that the design is satisfactory for my application?

 

The input current Ii = (5v-0.7V)/1K=4.3mA, then the Ic will be 43mA, it's too much, the R1 can change to 3.9K and adding a Rbe as 2.2K to avoid when the input is floating.

 

The input current Ii = (5v-0.7V)/1K=4.3mA, then the Ic will be 43mA, it's too much, the R1 can change to 3.9K and adding a Rbe as 2.2K to avoid when the input is floating.

Do you mean that I should change R3 and R4 to 3.9k, and add a 2.2k from the base of Q5 and Q6 to ground? Thanks.

 

Do you mean that I should change R3 and R4 to 3.9k, and add a 2.2k from the base of Q5 and Q6 to ground? Thanks.

Yes, If the I/O current of pic is enough then you should do.

 

Clamping diodes are inherent in the FET's - the body diode.

Has something changed in the makeup of mosfets? I see more and more people saying that the intrinsic diode present in a mosfet is all that's needed. It hasn't been that long ago(a year or two) that people were saying to ignore the intrinsic diode for protection, that it was of a unknown/inferior value.

While this is what was said for mosfets, an IGBT is another matter. In an IGBT the diode present is a separate known value. It, the diode in an IGBT, is a real added discrete component in the IGBT, matching the characteristics of the IGBT.

Help me to understand this new thinking on mosfet please.

 

Why screw with it? I needed several, so I just bought some of these. They work for regular DC motors, not just steppers.

I needed PWM for speed control, and reversing of a PM 12V motor which draws a maximum of 2.5A. I used an PWM output pin from an Arduino for the speed control and a separate output pin for direction.

 

Why screw with it? I needed several, so I just bought some of these. They work for regular DC motors, not just steppers.

You are correct, of course. That seems to be the case more and more often: cheaper to buy from China than to build. Still, I had all the parts for this, so I built it anyway. I will probably order some of the ones from China for future projects. Thanks.

 

Do you mean that I should change R3 and R4 to 3.9k, and add a 2.2k from the base of Q5 and Q6 to ground?

A resistor from base to ground is not needed for normal operation.

 

The L298 is convenient, but you lose a lot of power due to the voltage drops across the internal components. Take a look at the Source and Sink saturation voltages near the bottom of page 3: http://www.st.com/web/en/resource/technical/document/datasheet/CD00000240.pdf
Total drop (worst case) with 1A current is 3.2v, with 2A, 4.9V.

 

Has something changed in the makeup of mosfets? I see more and more people saying that the intrinsic diode present in a mosfet is all that's needed. It hasn't been that long ago(a year or two) that people were saying to ignore the intrinsic diode for protection, that it was of a unknown/inferior value.

They aren't very good diodes. For example the one in the IRF4905 can take over 100 ns to turn off and has a large voltage drop. I learned this one the hard way in a high power motor driver. Everything was ok until the diode was still conducting from the inductive kick when the fet opposite it turned on and poof.
But in this case where there is no pwm and no concern over the voltage drop it should work ok.

 

Below is the simulation of a simple bridge circuit with a single control input that should work fine if you don't switch when power is applied. It does require the lower N-MOSFET to be a logic-level type if you want to control it with a 5V logic signal.

 

Crutschow, try replacing the 10 Ohm load resistor with an inductor of about 2H, and a resistance of perhaps 6 Ohms, and increase the battery source resistance to around 20m Ohms. Then take a look at the Vgs of the upper MOSFETs during ON/OFF transitions. Remember, if Vgs excursions exceed ±20v, the MOSFET gets smoked.

 

They aren't very good diodes. For example the one in the IRF4905 can take over 100 ns to turn off and has a large voltage drop. I learned this one the hard way in a high power motor driver. Everything was ok until the diode was still conducting from the inductive kick when the fet opposite it turned on and poof.
But in this case where there is no pwm and no concern over the voltage drop it should work ok.

Interesting!

I have not heard much discussion about this myself.

If the body diode is paralleled with another external diode, it must have a lower forward drop than the body diode, to ensure that the body diode is not carrying current.

It seems that schottky diodes would be essential for this role, but I never see any mention of this concept?

 

Yes schottky or often another FET that turns on shortly after the bottom one turns off then the FET becomes the clamp and the diode can be small because it only carries current for a short time.

 

Crutschow, try replacing the 10 Ohm load resistor with an inductor of about 2H, and a resistance of perhaps 6 Ohms, and increase the battery source resistance to around 20m Ohms. Then take a look at the Vgs of the upper MOSFETs during ON/OFF transitions. Remember, if Vgs excursions exceed ±20v, the MOSFET gets smoked.

The op stated (Post #5) that the power is off when the bridge is switched. (I did show the bridge switching under load but that was just for simulation purposes). But the turning off of the switch to the battery would generate a large transient so the bridge power connection should be bypassed with a reverse-biased diode to ground to absorb the negative transient (which I should have included but did not). Thanks for bringing that to my attention.

I don't see a problem with turning it ON. And 2A * 20m ohms is only 40mV so I don't see how the battery impedance is a factor.

 

Interesting!

I have not heard much discussion about this myself.

If the body diode is paralleled with another external diode, it must have a lower forward drop than the body diode, to ensure that the body diode is not carrying current.

It seems that schottky diodes would be essential for this role, but I never see any mention of this concept?

All of the older mosfet application notes and many electronics books bring this up. The body diode is an "artifact" of the mosfet building process. It is supposed to be very inferior compared to the mosfet ratings. Was always mentioned in the explanation of a circuit but not so much lately.

As a newbie/self learner/hobbyist in electronics I'm constantly trying to keep up with stuff, this is why I was asking.

@ronv, thank you for confirming my belief on this, things are changing so much I wasn't sure if it still applied.