Analysis help needed: MOSFET control circuit for -12V?

[ankeakala]

Can someone explain what is the purpose of D2 zener in this circuit? As far as I understand, both of those enhancement type MOSFETs are off because both VGS are at 0V when PF0 is floating.

And if PF0 is pulled to 0V, P-Ch would then conduct, and presumably D2 zener would also conduct because the delta is greater than 5.6V and would pull N-Ch G1 to roughly 5V (turning it on), but how will that R17 to -12V affect that? Would it not pull N-Ch G1 to -12V and the N-Ch would stop conducting? Or am I lacking in my understanding of how negative voltage works?

Or is there something else going on?

[Ian.M]

Somebody decided that 17V  (5V - (-12V)) was too much for the N-MOSFET gate, so they put in a Zener to drop the gate drive by about 5.5V.  The MOSFET is actually rated for +/-20V Vgs for both the N and P channel devices ([datasheet] page 2) so, unless the -12V supply has negative going spikes of over 3v, its not essential.  Personally, I'd use a 4K7 resistor there which would form a potential divider with the 10K gate pulldown limiting the gate voltage to about 11.5V , but then I'm cheap like that!

You asked what happens with the gate pulldown R17.  When the P-MOSFET turns on, the Zener wins (as its slope resistance is *MUCH* less than 10K), pulling up the N-MOSFET gate to approx -0.5V i.e. 5V supply - Zener voltage, (nom. 5.6V, but reduced due to low current through it)

[T3sl4co1l]

That. ^

FYI, the fuse won't do anything.  Fuses blow very slowly, milliseconds.  Transistors blow much faster, hundreds of microseconds perhaps.

Well, I mean the fuse will still protect the circuit, but only after the transistor has failed shorted, allowing full fault current to dump into the load, through the fuse, clearing it.

A better solution is to use a current sensing circuit, and when current goes well beyond limits, simply turn off the transistor for some time.  Alternately, a much bigger transistor can be used, that can pass enough current to clear the fuse (but this will be somewhat more expensive, and will have much more gate capacitance, so will switch slower).  Also, a bigger transistor can be used in a different way, either with an analog control circuit (a few transistors) or an IC (typically called a hot-plugging, wired-OR or load switching controller), to limit current actively rather than permitting fault currents to flow.  Downside is this can dissipate a lot of power (not quite as much as the unlimited fault condition), so some other means of protection is usually a good idea (a time delay or thermal cutout).

Tim

[ankeakala]

Thanks for the tips and comments. Still not quite clear how exactly it works but at least it seems to work exactly like said. With the Zener the N-MOS gate drive is that 5.5V lower.