Build a H-Bridge Circuit: Tips & Info

[ha700]

As part of a group project I have been given the task of building a H-Bridge circuit to replace a previously built one in a fighting robot.
I have no access to what this H-Bridge looked like or what transistors where used etc.
I understand only the basics of a H-Bridge about how it works and what type of transistors I will need to make it work.
Since the motors that need to be driven are high powered (dc motors and do not know any specs but can find out) I am struggling to decide on a circuit design.
I have found multiple circuit designs online where some use diodes and capacitors and others don't.
What I do know is that the transistors I need to use are MOSFETS (both npn and pnp) that need to be around 60V/70A.
The input signals will come from a pwm speed controller.
I have been advised to look into using a H-Bridge driver chip as well but I don't know where this could be connected in a circuit I design or find.
I just need a bit of a push and help on where to begin and why.

This is one of the circuits I have been looking at and wondered if I changed the transistors for more powerful MOSFET ones I have ordered whether it will work the same or not for what I need.

If there is any more information that I would need to provide I will gladly be able to find out and all comments and inputs will be appreciated greatly!

 

[Baluncore]

Welcome to PF.
You will use four mosfets as power switches, rated at 80V and 80A or better. The gates of those mosfets will be driven by H-bridge drivers capable about 1 amp output to the gate.

Since N-chan mosfets are being used as high side switches, there will need to be an over-voltage available above V+.
Firstly you select one or more driver chips to control the gates of 4 x N-channel mosfets.
Find out what control logic voltage will be used, is it +3.3V or +5V ?

Use the selection tables available on component distributors websites such as Digikey, Mouser, Farnell and RS, to find H-bridge drivers that meet the required specifications. Download the data sheets and application notes.

You will need to justify your decisions so you need to understand H-bridge drivers and mosfets. See;
http://www.egr.msu.edu/classes/ece4...e Regarding H Bridge Design and Operation.pdf

 

[Windadct]

Also - this is a very common hobby / diy project with literally thousands of resources on line, many are open source. It all starts with the motor spec.

 

[rbelli1]

What I do know is that the transistors I need to use are MOSFETS

Why do you need to use MOSFETS? The higher efficiency you gain using them has a cost in complexity of the drive circuit especially for high powered loads. A modified version of your circuit will work in a pinch but can only be used on fairly low powered devices as you need to add resistors to mitigate the shoot-through you will inevitably get.

Since N-chan mosfets are being used as high side switches,

P-channel can be used as well. This avoids the necessity of needing the higher voltage. Integrated drivers will often provide these voltages for you so when using integrated drivers it is best to find one that allows N-channel only operation as Baluncore describes. If you need static operation (100% duty cycle either direction) The drivers get more expensive.

BoB

 

[Baluncore]

Why do you need to use MOSFETS?

A MOSFET with a very low channel resistance will have a very low Vds when conducting, so it is efficient with low voltage, high current motor H-bridges.
An IGBT will drop more voltage, but at a lower current, so it is better applied to high voltage, lower current inverter H-bridges.

A modified version of your circuit will work in a pinch but can only be used on fairly low powered devices ...

I do not trust that circuit. What are the 1N4148 signal diodes, rated at 100 mA, 75V doing there? A fast Schottky diode capable of handling 80 amp is needed for freewheeling in the proposed application.
It may be good to have that diode integrated into the MOSFET.

BJTs are not good for high currents. A single transistor will require a continuous base current of over 1 amp, expect about 40 watt while conducting. A Darlington transistor is inefficient in high current switching due to the high Vce of about one volt that will generate 80 watt whenever the switch is turned on.

P-channel can be used as well.

P-channel mosfets cost about double the price for an equivalent N-channel power device. It is better to stock and use only the N-channel devices for high current switching.

 

[rbelli1]

BJTs are not good for high currents.

Yes. I somehow missed the 70A part of the OP's question.

P-channel mosfets cost about double the price

True but for a one off build the easier gate drive could compensate for the increased cost.

BoB

 

[Baluncore]

True but for a one off build the easier gate drive could compensate for the increased cost.

That may be true below 24V, but above that, the gate drive of high side switches require level translation for both P and N channel devices.

Zero current or zero voltage drop makes for zero heating. The trick to using MOSFETs is to switch them quickly to avoid the VI losses during transition. That is especially important for high frequency PWM switching. Use the lowest frequency consistent with motor speed and torque control.

P-channel MOSFETs cost twice that of N-channel because they need twice the area of silicon for the same conduction. That doubles the gate capacitance, so it takes twice as long to transition, the switching losses are double and the maximum switching frequency is halved.

It is the lower collector capacitance of IGBTs that give them the faster dv/dt switching advantage at higher voltages when their higher saturation votage drop becomes less important. For low voltages, MOSFETs with high current gate driver chips give higher dv/dt and so have the advantage.

 

[rbelli1]

to avoid the VI losses during transition.

It also reduces issues with shoot through if you don't have a gate driver with a dead band.

level translation

That doubles the gate capacitance,

These two conspire together to make me wish I never heard of using P channel in the high side of a bridge and for non-static operation the boost function is essentially free.

I have noticed that load switches are moving to using N channel FETS and built in boost converters.

BoB

 

[Windadct]

I am sure for this application MOSFET is fine. I do not know how OP came up with the 60V/70A Spec for the MOSFET - if this is the motors spec, the MOSFET need a higher rating - etc. Issues like regen current, and DC overshoot at turn off require considerable headroom over the motor and power supply values. There are plenty of MOSFET in the 60V / 70A range. IGBT generally not considered until you need 600V + devices. (> 250 VDC).

As for the driver - there are Dual drivers with bootstrap ckt for the high side simplifying design - even for a one off build.

 

[ha700]

I have chosen the N-channel MOSFETs as they are cheaper and I only have a £50 budget for my entire project as I need to build 2 H-Bridge circuits.
I went to my tutor for advice on transistors that I found and they suggested a range of 60V / 70A.

At the moment I have purchased both N and P type MOSFETs to test which I know are not all 60V / 70A as some were bought before I asked my tutor.
IRFZ40PBF N-channel 60V / 50A
2SK3845 (Q) N-channel 60V / 70A
IRF9540NPBF P-channel 100V / 23A
IRLR2905ZPBF N-channel 55V / 60A
They do not necessarily have to work for the intended purpose first time around as I want to be able to test and improve the circuit (the main target of being a group project).
I have also bought 1N4004 - E3/54 400V / 1A diodes, would these work? If not, what types would be better? and why?

Also, if I needed capacitors and resistors within this circuit how would I be able to figure out the right values I need? I still do not know the control logic voltage but will try and find out soon if this is an important factor.

http://www.egr.msu.edu/classes/ece4...e Regarding H Bridge Design and Operation.pdf
For the diagram in figure 2 of this page, say the load was a motor, would you turn on the high side transistors to make the motor rotate in one direction and the low side for the opposite direction? It doesn't make it clear as ost diagrams I have come across, you turn on the diagonal ones as apposed to the ones next to them to make the motor rotate.

Thank you so much for all your responses! I wasn't expecting as many and have all been extremely helpful into helping with my understanding of H-Bridges and being able to build and improve one of my own.

 

[Baluncore]

For the diagram in figure 2 of this page, say the load was a motor, would you turn on the high side transistors to make the motor rotate in one direction and the low side for the opposite direction? It doesn't make it clear as ost diagrams I have come across, you turn on the diagonal ones as apposed to the ones next to them to make the motor rotate.

Look at figure 1; which labels S1, S2, S3 and S4. Transfer those labels to figure 2.
The safe situations are:
Forwards = Turn on S1 and S4.
Backwards = Turn on S2 and S3.

You must avoid the power supply short circuit situations that occur when either S1 and S3, or S2 and S4 are turned on at the same time. Make sure it is not possible by using gate drivers that logically lock out that possibility.

The special case of anyone switch on will have no effect.

The free wheeling diodes with the inductance of the motor can have interesting effects. All four switches off, will allow the current through the motor to fall rapidly to zero which will quickly reduce motor torque without reversal.
S1 and S2 on, or S3 and S4 on, will allow the motor current to spin for slightly longer.