Hbridge switching losses reduction idea

[Jamie Morken]

Hi,

I was thinking about how to decrease switching losses in an Hbridge, ie.
one with a 600V supply.

If the switching time is ~50ns for the fets, could you reduce the
switching losses by using a series inductor between the 600V supply and
the top of the Hbridge that saturates soon after the switching is
complete, so that it restricts current flow during switching, and then
saturates once switching is complete so that it doesn't restrict current
flow to the load?

cheers,
Jamie

 

[legg]

Hi,

I was thinking about how to decrease switching losses in an Hbridge, ie.
one with a 600V supply.

If the switching time is ~50ns for the fets, could you reduce the
switching losses by using a series inductor between the 600V supply and
the top of the Hbridge that saturates soon after the switching is
complete, so that it restricts current flow during switching, and then
saturates once switching is complete so that it doesn't restrict current
flow to the load?

In practice, this function is only practical if the device can be
'desaturated' in time for the next switching period. In the process,
the flux swing produces losses that have to be taken into account to
ensure that material limits and magnetic characteristics are not
compromised.

The volt-seconds that such a device needs to exhibit in a 600V
switching circuit is not small, while the choice of suitable core
materials and shapes are.

Most attempts at switching loss reduction using saturable reactors
employ the parts in circuit branches where impressed voltage reversal
occurs naturally during the full switching cycle, or in auxiliary
assisted switching cells that work to recycle the magnetizing energy.

In such circuits, the limitations of practical rectifiers soon becomes
evident; don't be the first on your block to discover reverse recovery
charge amplification. Large currents in small rectifiers accumulate
larger stored charge and foward over-voltage than the same current in
a larger rectifier of similar 'speed'. Sad but true.

RL

 

[legg]

google 'loss-free snubber', "passive snubber", "soft switching" or
"assisted switch"

RL