Where is the gain?

Partial dV_DS/dV_GS.

Cheers

Phil Hobbs

There is nothing spectacular about that high side driver! In fact, I think it's flying on a wing and a prayer !

Those diodes are very much needed! with out them, the charge on the gate would dissipate rather quickly from L1 or the load. Remember that voltage differences is going to cause current until they balance and I don't see the 3.8V side moving for any one!

Lets get back to how the charge is working there in the first place. Short turn on time with QM3 will discharge the gate there by turning QM2 off, because there wasn't enough time to magnetize L1 (enough).

A longer on time of QM3 will allow for L1 to start magnetizing enough energy so that when QM3 turns off you'll get enough collapsed energy (+), that is more than enough voltage to put the gate above the 3.8Vs that sits on the drain side, this of course also puts the gate above the source side to ensure it is on.

So, as you can see, the diodes are very much needed to prevent drain off of the gate from either L1 or the load.

Personally, I would be concerned with leakage for long on times.

Jamie

Even better with the original 12 volt supply.

That's due to the capacitance of the load when the device is turning _off_ (by choosing VG as the X-axis during a transient analysis, you see the "hysteresis" due to the load capacitance slowing the fall of the source). ...Jim Thompson

I was just quoting the gain of any FET amp whatsoever. Whether one considers that an input/output gain depends on what one considers the input, of course.

Cheers

Phil Hobbs

That doesn't make it "Miller". If I run "data-restricted" so that I catch only the turn-on, there is no slope > 1.

You seem to be defining the source so that it wags the dog... doesn't happen ;-)

But if you want to call it "Miller", go ahead. You have a perfect right to be wrong-headed and obstinate about it >:-} ...Jim Thompson

I wasn't doing anything of the sort. I merely pointed out that there was only one kind of FET behaviour, and that the nomenclature depended on which electrode you took as AC ground--i.e. it's a purely semantic issue.

Cheers

Phil Hobbs

You can have an apparent Miller effect, and in this case it is the Cgd dependence on VDS, the capacitance increases at turn-on significantly, looks like a factor of x4 in datasheet.