H-Bridge Buck-Boost topology

[red_alert]

I'm trying to implement a buck-boost converter using a H-bridge (non-inverting, single inductor) topology.

The input voltage range is 0 - 100 V DC (three-phase rectified AC from a wind turbine generator) and the output should be around 30 V (or 60 V, after I reconfigure the battery string from 24 V to 48 V). The switched power will be around 1500 W.

The schematic diagram looks like this:



Actually, I should use MOSFET switches instead of those freewheeling diodes so the modified diagram should look like this:



I'm going to use an MCU to generate those four PWM signals but my main concern is how to keep a high-side MOSFET (SW1) in ON state while the circuit is running as boost converter. I don't have separate power supplies for the high side driver so I wonder how to implement a bootstrap operation in this situation.

Any help is kindly appreciated.

 

[FvM]

Yes, you can use bootstrap gate driver supply for the high side switches. It works as with any other bridge converter. As a prerequisite, the duty cyle must be limited slightly below 100 % to keep the bootstrap working.

 

[red_alert]

But the corresponding low-side switch will be in OFF state all the time. How/when to recharge the bootstrap capacitor?

 

[mtwieg]

If the low side switch is off "all the time" then what is its function in the circuit?

The low side switches will have to be turned on at some point to refresh the bootstrap caps once in a while. All control ICs based on this topology implement it in some way, even when the switches should nominally be off.

 

[FvM]

You got the point, the actual problem is lowside minimal on-time. Of course it's not optimal to keep the inactive half bridge switching, but it's the most simple way to maintain bootstrap operation. The alternative is to implement an auxilary circuit (e.g. a charge pump) to power the driver. This method is implemented in Linear buck-boost switch controller LTC3789.

 

[red_alert]

When the circuit works as a boost converter (looking at the diagram in my first post) the high-side switch (near the Vin) has to be always in ON state. It is PWM driven only when the circuit runs as a buck converter (and that's dynamically changing).

Anyway, maybe I should put its corresponding low-side switch to ground (the freewheeling diode in the first diagram) durring a particular phase of boost operation (I'm still looking for informations about that). That means I have to put that inductor terminal to ground (instead of Vin).