[SOLVED] P-Channel Mosfet woes. Can someone tell me what's wrong with this?

No matter what the signal from the microcontroller is, the device is always being supplied with power (always on). The "4V ??" is what I measured with a meter when the signal from the microcontroller was off and the device had been removed. Surely the voltage there should have been 8.4 not 4???
Can't use an N-Channel, device has to be on the ground rail.
Microcontroller signal is 5V

Any help appreciated.

Surely the voltage there should have been 8.4 not 4?

Click to expand...

Yes, can't be understood from the schematic, which is basically fine. Probably defective parts or a wiring error.

As a first step, you should verify that the Q1 base-emitter voltage is actually zero, respectively < 0.3 V.

I completely removed Q1 and just left the gate pulled up to +, the device was still on. So looks like the Mosfet is defective.

So looks like the Mosfet is defective.

Click to expand...

It might be blown if the "device" is inductive with a stored energy exceeding the MOSFET's avalanche energy capability.

The device is just a heating coil, it shouldn't have any inductance. Besides, it shouldn't have come on at all the very first time I connected the power anyway. Just tried another p-mosfet, this time with just an led and a resistor as the device. Same, led on all the time. Oddly tho' it's not as bright as it is if I connect the led+resistor directly to the power.

- - - Updated - - -

This is how I have it breadboarded to test the mosfet, now unless I'm going mad, that LED shouldn't be on, but it is. With all 4 p-mosfets I have.

Just out of curiosity, does a power mosfet like the 5305 need a particular voltage before it works properly? All the examples I have seen they use 12V or more, I've never seen a power mosfet example with less than 12.

EDIT: Source and Drain mixed up, DOH!! Sorry. Problem Solved.