There is any alternative to the inverter cricuit?

[AMSA84]

Hi guys,

As the topic says, there is any alternative circuit to implement the same functionality of the inverter? But without having the issue of the simultaneous conduction? This would be to drive a bigger transistor.

Regards.

 

[dick_freebird]

Split the gate drive of the N and P legs of the final, and
maybe their predrivers as well, and tailor the drives to
get nonoverlap (or minimized-overlap) switching. I call
it "ballistic anti-shoot-through" (as opposed to a feedback
logic anti-overlap).

You'll never be rid of crossover current entirely but you
can eliminate 95% of it pretty simply.

 

[AMSA84]

Hi freebird,

From what I understood from your suggestion is to use a non-overlapping scheme in the last 2 drivers, is that it? If so, I have done that already. I have improved the dissipated power by 75%. But I would like to improve even further.

Taking the advantage of the post, I was looking to this topology: http://analoglib.net/2011-03-18/fisher-jssc87-class-ab-buffer-amplifier/ don't know if it makes sense to use "as the" last two drivers.

Regards.

 

[erikl]

I think it's a bit exaggerated to use a rather complex, highly linear driver for a digital driving purpose. And don't forget the quiescent power which is needed for such an analog driver, even for class AB stages.

dick_freebird's method is better for this purpose, I'd think. If you already managed to save 75% of current-shoot-through power, you could perhaps achieve still a little bit more saving.