I'm not sure if this is what you'd call "floating voltage" or not. I'm lacking on some terminology so bear with me. Here's the circuit with explanation below:
NOTE: I updated the image. There was an error previously. A PNP BJT was shown but both BJTs should be NPN.
The premise of the circuit is this: I want to power L1 coil with square pulses. You can see I've got a freewheeling diode to handle the back emf field collapse to protect the circuitry.
From previous experiments and talking to others on here, I learned that using an N-channel mosfet on the high side of the coil like this doesn't work because the large negative voltage during the back emf coil collapse puts negative voltage on the mosfet source which causes the gate (at zero volts) to turn on while the field is collapsing.
I experimented with BJTs and they make it easy to fix the problem, but they also get hot. I really want to use mosfets for the ability to pass a lot of current with minimal resistance. So I've switched to a P-channel mosfet in the circuit shown. This seems pretty simple but I'm adding a level of complexity.
I want to have arbitrarily high voltage in the main circuit (80volts shown here). Since the p-channel mosfet requires 80volts to be on the gate in order for it to be turned off, I have added a small resistor network to supply the required voltage during off-state. You can see I'm using a 555 and some BJT's to pull the voltage low at the gate to turn it on.
Now the tricky part. From my understanding, I can't drop the mosfet gate all the way to 0V because that would make an 80volt difference between the Drain and the Gate which would destroy it. I need to make sure the gate drops to about 12 volts less than the Drain but drops no further. THe solution I've come up with is to add a 12 volt zener diode and adjust the resistances accordingly so once the gate starts to become more than 12 volts less than the drain, the diode will reverse bias and bring the gate back up. It should stay right around 68volts if the power supply is 80 volts -- as long as the BJT is biased on anyway. Once the BJT turns off, the gate will recover to 80 volts and shut off the mosfet.
So my question is simply this: Am I doing this right? Am I over-engineering it or perhaps doing something wrong? Let me know. Thanks!
- Jim