MOSFET full bridge switching

Introduction

This is part 7 of the series on motor control. It starts to introduce full bridge control of motors.

What is a full bridge

A full bridge is a collection of four 'legs' which are switched in the correct way to control the motor. The diagram below is repeated from the www site.

07-brdge

Some of the logic behind this conventional full bridge has been described on the 4QD www site. However, such an approach has problems when it comes to fully protecting the bridge: the logic of providing effective hi-side current limiting is not simple in a conventional full bridge and many designs fall down when it comes to overcurrent protection, especially in regen mode. But why be conventional when you can be original?

Two Half-Bridges

A full bridge is, logically, two half bridges.

In previous articles we have seen how a very simple half-bridge controller (the 2QD) works, and how it is current limited in both quadrants, and how the current limits look after themselves by altering the modulator.

So - logically, to make a full bridge, you simply use two half bridges, back to back, with the motor between them.

This of course, flies in the face of conventional full bridge motor control where you have four bridge legs and you therefore require four logic outputs to drive them. We only have two inputs - admittedly analogue, not digital, and all four legs of the bridge can, quite possibly, be switching at the same time.

07-bridge

The drawing above shows the simplest possible such full bridge controller, using two 2QD controllers.

You may be concerned that two controllers are effectively fighting each other: one is switching one end of the motor at one frequency, and the other is switching the other at a similar frequency, but with totally random timing.

But we have seen that, if the switching frequency is high enough, the motor current is essentially pure d.c. If only a constant current flows through the motor, how can one controller even 'know' that there is another at the far end of the motor?

This 'full bridge' was originally suggested by us to our users many years ago in a slightly different guise, as a position servo system (maybe for an aerial rotator or similar). In this, the two controllers have different pots. One pot is the 'demand' angle and the other pot is connected mechanically to the rotator as position feedback. Very simple and works well.

When we designed the Uni, we hoped it would replace the 2QD. However, since it does not have an active hi-side voltage pump (relying instead on only the output bootstrap pump), it cannot be used in a full bridge. In a bridge, it is necessary for the gates of the hiside MOSFETs of the controller which is 'off' to be fully biased on, even when the controller is not switching. In a half bridges, under these conditions, there is no voltage being fed to the motor, so there can be no motor current. It follow that the hiside MOSFET gates can be allowed to turn off.

But in a full bridge, this is not the case. It follows that only a specially designed half-bridge controller can be used for used for this and, at the present time, only the 2QD is suitable.


Other relevant pages


Part 1
Part 2
Part 3
Part 4
Part 5
Part 6
Part 8

Page Information

© 2001-2011 4QD-TEC
First published: