High side circuit simulating strangly..

[Dreamer1]

I have designed the this high side drive, it's meant to dive a load attached to a battery

 

To turn it on I just leave the charge pump on and to turn it off I turn the charge pump off and use the optocouplers to quickly drain the flying capacitor. Needless to say, I don't have must experience with this sort of things.

Anyhow, when the gate of the high-side mosfet is at 12V (same as the batter and drain) for whatever reason the mosfet conducts? Why? I have attached a graph. I don't see how this is possible, the body diode is facing the other way so I don't know what it could be? Simulation error?

Plus, is this a good cheap high side drive?

By the way, the graph shows the voltage at the gate of the mosfet and the current through the 1 ohm resistor.

Thank you for your kind help and suggestions.

[dannyf]

I have designed the this high side drive,

Before you get it fancy in simulation, ask yourself how does the circuit work? What is the essence of it?

To turn it on I just leave the charge pump on and to turn it off...

Does it actually do any of that? Should it do any of that?

for whatever reason the mosfet conducts?

The mosfet would have failed if it did NOT conduct.

Ask yourself under what conditions the mosfet would conduct? Are those conditions true in your circuit? ...

It helps a lot if you start simple.

[nickm]

Your source voltage is a little under 10V so Vgs is around 2V and your FET is going linear.  The problem is your opto coupler transistor is trying to pull down to 0V but you have your battery voltage pulling it up through the two diodes and the battery is winning.  Put some impedance like 1kohm in between the gate and your charge pump and leave your opto transistor where it is.  This way the opto transistor only has to sink about 10mA to turn the transistor off. 

Also you might consider using a normal transistor in an open collector configuration rather than an opto.  You've probably already thought about it but you could use a P-channel FET without having to bother with the charge pump and drive the gate directly with your logic.  Downside is lower RDson.  Other than that I think it looks fine. 

[Richard Head]

Circuit simulation is an amazing tool that so often is abused.
99% of the time a simulator is not needed or required, just common sense.
Forget the simulator and understand what you want to acheive and how it can be realised.

Dick

[dannyf]

Forget the simulator and understand what you want to acheive and how it can be realised.

Couldn't agree with you more.

Simulation is only useful if you understand how the circuit to be simulated works AND the limitations of simulation.