The MagPi Magazine - Aimed at learners - Printed edition Kickstarter

On page 6 of the latest magpi, there is a picture of a H bridge motor drive with npn transistors driven from the Pi gpio lines.

How can the upper transistors ever conduct when they are driven from a 3V3 gpio line?

Sorry if it's a bit off topic. Just want to warn people who try to build this. 

I have been reading The MagPi since the first issue and I find that it has a good mix of beginner, intermediate, and more advanced articles.

I definitely recomend it.

Luc Cool wrote: How can the upper transistors ever conduct when they are driven from a 3V3 gpio line?

It will work with bipolar junction transisters. With MOSFET transisters it requires a gate voltage that is higher than the source voltage.

Gary Stewart wrote:

 

I have been reading The MagPi since the first issue and I find that it has a good mix of beginner, intermediate, and more advanced articles.

I definitely recomend it.

 

 

Luc Cool wrote: How can the upper transistors ever conduct when they are driven from a 3V3 gpio line?

 

It will work with bipolar junction transisters. With MOSFET transistors it requires a gate voltage that is higher than the source voltage.

From my understanding of the circuit, you can get the upper transistors to conduct, but you'll only be able to get a volt or so across the motor.  Assuming the upper left and lower right transistors are switched on and the other two are off, your current path from RasPi is through the upper left resistor (maybe a 1V drop), across the upper left transistor's BE junction (0.6V drop), through the motor, and then through the saturated lower right transistor (0.3V drop).  With a 3.3V supply, that leaves only 1.4V across the motor which won't run very fast unless it's a really low voltage motor.  If you need more voltage across the motor, the upper transistor can't switch on.

Wikipedia's article on the H-bridge says:

A solid-state H bridge is typically constructed using opposite polarity devices, such as PNP bipolar junction transistors or P-channel MOSFETs connected to the high voltage bus and NPN BJTs or N-channel MOSFETs connected to the low voltage bus.

 

The most efficient MOSFET designs use N-channel MOSFETs on both the high side and low side because they typically have a third of the ON resistance of P-channel MOSFETs. This requires a more complex design since the gates of the high side MOSFETs must be driven positive with respect to the DC supply rail. However, many integrated circuit MOSFET drivers include a charge pump within the device to achieve this.

You can't directly use PNP transistors with RasPi if the motor voltage is above 3.3V, but you could add two additional NPN transistors to act as inverters to control upper PNP transistors.

Your conclusions are the same as mine John. I posted this on the official pi forum as well, but so far no serious response.

Another addition I would strongly recommend is an over current protection. This can be a small resistor in the negative line. If the voltage on that exceeds 0.7V, it can drive a transistor that pulls the base drive of the 2 lower transistors low. Another good protection would be diodes preventing the higher motor voltage to return to the pi gpio lines in case one of the transistors breaks. Darlington pnp and npn transistors might be a better choice as they will have a much higher current gain. MOSFETS could be used as well, but they usually require a gate source voltage of at least 5V (the high current ones), so they will need a more complex drive circuit.

It's a truly dreadful ciruit and has obviously not been tested - a motor driver chip would be the best way to go - if you choose the right one you get all the protection and level shifter bits built in.

Michael Kellett

I agree, but motor driver circuit are often expensive and difficult to find.

Besides that, for the purpose of learning how things work, the setup with separate transistors is better.

If you want to add the protections, it quickly becomes a more complicated circuit.

Luc Cool Wrote:

Wikipedia's article on the H-bridge says:

A solid-state H bridge is typically constructed using opposite polarity devices, such as PNP bipolar junction transistors or P-channel MOSFETs connected to the high voltage bus and NPN BJTs or N-channel MOSFETs connected to the low voltage bus.

 

The most efficient MOSFET designs use N-channel MOSFETs on both the high side and low side because they typically have a third of the ON resistance of P-channel MOSFETs. This requires a more complex design since the gates of the high side MOSFETs must be driven positive with respect to the DC supply rail. However, many integrated circuit MOSFET drivers include a charge pump within the device to achieve this.

You can't directly use PNP transistors with RasPi if the motor voltage is above 3.3V, but you could add two additional NPN transistors to act as inverters to control upper PNP transistors.

Google L298, a  bipolar transistor H-bridge driver IC and download PDF. It clearly shows NPN transistors high and low side.

The datasheet only shows the block diagram.

Since they show the drive of the upper and lower transistor inverted, it makes sense to show 4 npn transistors.

John Beetem's article from wikipedia looks correct and describes it accurate.

I don't see a voltage charge pump in the block diagram of the L298, so either the block diagram is highly simplified or the actual

silicium uses pnp transistors for the upper ones.

Gary Stewart wrote:

 

Google L298, a  bipolar transistor H-bridge driver IC and download PDF. It clearly shows NPN transistors high and low side.

You can have upper NPN drivers provided that you can get the base voltage higher than the emitter voltage.  The L298 diagram doesn't show the details of the AND gates that drive the upper NPNs.  I suspect they convert TTL input levels to "Vs".  Preferably you'd like those base voltages higher than "Vs" so the upper NPN turns on fully, but that requires a charge pump.  Delving into the L298 data sheet, I see voltage drops of 1.35V (typ at 1A) on the high side and 1.2V (typ at 1A) on the low side.  That's a total of 2.55V at 1A = 2.55W, hence the large ground pads.

I sent my rpi or rip back to cpc cos the ethernet and updates didn't work and I got fed up with reflashing img. Might buy another when it's more mature.I liked PIC usage and this looks interesting.good luck to whoever is involved in providing info,it is most welcome.stan