6000W By IRS2092

[darkfenriz]

I am a bit suspicious about mosfets rated 200V and more. For a given technology level and current rating, the capacitances and miller charge seem to go propotionally to VDS rating, while body diode charge seems to rise more or less proportionally to third power of VDS.
You may be much more "headache free" when switching at 60-100V VDS and bridging.
For my next project I plan to use a "bridgeable bridge" topology with four floating half-bridges to get +-100V output swing with 60V devices (Qgd=6.3nC, Qrr=17nC)

[ChocoHolic]

No doubt, MosFets and especially their body diodes are quite 'ugh' above 100V.

It would be nice, if you could post a simplified principle of what you call a bridged bridge. My brain does not deliver a prompt translation from 'bridgeable bridge' into a schematic.

I am aware of staggering the power stages by using multiple floating power supplies. And such concept would again allow to be bridged (ref. attachement). Theoretically this concept would even allow multi staggering, not only 2 times, but 3..4..5...
But also there - probably the lunch is not for free.
If you skip the filter inbetween, you will get large circuit areas with giant dv/dt versus earth... generating an EMI hell. If you put the filter inbetween - you get more phase shifts and delays, resulting in less freedom for feedback and stability. Of course a trade off with a fast filter might be an option.

Curious about you conceptual schematic...

[lumanauw]

Is it something like patent#6097249, where the supply can be switched to ground?

[81bas]

Originally Posted by darkfenriz

For my next project I plan to use a "bridgeable bridge" topology with four floating half-bridges

What you will do against the bus pumping effect in floating PSUs?

[cccv]

What do you think about this mosfet STW13NK100Z

[81bas]

Originally Posted by cccv

What do you think about this mosfet STW13NK100Z

Rds-on is too high (0.56 Ohm)... Current capability is too low (13A)...

[fredos]

body diode way too slow...

[Eva]

Originally Posted by cccv

What do you think about this mosfet STW13NK100Z

For ~1000V class D?

[cccv]

just for safety. but why not?.. if I take this parallel?

[Eva]

With good PCB layout and gate drive, MOSFET can be used at nearly it's rated Vds, and on the other hand, higher voltage devices have higher capacitances and impose a lot of restrictions on how their body diode can be used. So using parts with higher Vds than required is not a good idea.

Most MOSFET rated at 100V and less will handle body diode hard switching without problem. Some 150V parts are suitable too. With well controlled di/dt, a few 200V/250V ones, and very few 500V/600V ones will handle it too, but charge storage results in high losses on the latter. All higher voltage parts would need external diodes.

IRFB4227 and IRFB4127 are rated at 200V and still can handle hard switching with reasonably low losses. Also, low Rds-on allows to bypass the body diode up to a certain current level (depends on temperature). In my current project, body diodes are only used with 2 ohm load.