Maker Pro
Maker Pro

Peculiar MOSFET operation ??

Pessimist

Jan 5, 2016
6
Joined
Jan 5, 2016
Messages
6
I have a P-channel enhancement mode MOSFET which I am trying to use for high-side power switching. (specifically, the MOSFET is the IRF9Z24N). I have an 18K resistor between gate and source and a switch from gate to 'ground'. There is a 100 ohm load resistor between drain and 'ground'. Applying a positive voltage to the source gives 'normal' operation where if the switch is open-circuit (Vg = Vs, so Vgs = 0), the drain voltage is 0 volts (FET is off), and if the switch is set to ground the gate, the drain voltage equals the source voltage (FET is on). So that's all as expected.
Where I get confused, is if no voltage is applied to the source terminal, but IS applied across the 100 ohm load instead.
By my reckoning, the Vgs is still 0 volts and the FET should be off. However, the voltage applied across the load appears at the source terminal implying that the FET must be ON !!
I've no doubt that there is a rational explanation to this, but it defeats me !!

Any advice will be welcomed !!
 

duke37

Jan 9, 2011
5,364
Joined
Jan 9, 2011
Messages
5,364
Have you got the fet the right way round? It has a diode in it which will pass current if Source and drain are interchanged.
Also, the load should be on the drain Not the source as you say.
 

CDRIVE

Hauling 10' pipe on a Trek Shift3
May 8, 2012
4,960
Joined
May 8, 2012
Messages
4,960
He said it's a P channel. In that case the Drain Source orientation would be the same as a PNP when used for high side switching. IE, the Emitter (Source) would be connected to the positive rail and the base (Gate) forward biased by bringing it closer to GND just as he's doing. His load is where it should be ..in the Drain leg.

You're correct about the internal Diode though. It's connected opposite of an N channel. This puts the Diode's anode on the Drain and the Cathode on the Source. It explains his measurements but not why he did it. :confused:

Chris

Duke, never mind. I misread your post. I'm leaving this as is because I don't delete my misteaks! I eat them.:p
 

Pessimist

Jan 5, 2016
6
Joined
Jan 5, 2016
Messages
6
Hello everybody,
Since my original posting, I have revisited the datasheet and seen that "body diode" - the voltage drop that I measure between drain and source would confirm that this is the path that the current is taking, as the diode is forward biased under this condition.
Chris, the reason why I'm interested in this situation is that, as part of a larger project, I need a battery backup facility, and am therefore using two FET's with their drains dot-OR'ed. When the main supply fails, the battery FET switches in and raises the voltage at the dot-OR (commoned drains). The problem then arises that the other FET's body diode is then forward biased and allows that FET's source to rise in voltage. This "upsets" the control circuitry upstream of the FET.
I hope that I've explained this clearly enough ! My current solution is to include an extra diode in series with the source terminal to block this "reverse" condition, but I would prefer NOT to lose the 0.7 volt drop across the external diode. What is the reason for the body diode (or is it an inevitable consequence of P-channel device construction ) ? Are any p-channel devices available without this diode ? (I need a max current spec of about 1 amp and would prefer a TO92 package (although I haven't found any so far at this current)
Thanks to everybody for their interest !
 
Last edited:

Harald Kapp

Moderator
Moderator
Nov 17, 2011
13,699
Joined
Nov 17, 2011
Messages
13,699
I would prefer NOT to lose the 0.7 volt drop across the body diode
Use a second MOSFET back to back, cf. my post #4.

or is it an inevitable consequence of P-channel construction
This diode is not P-channel MOSFET specific. N-channel MOSFETS have the same diode, reverse polarity. Yes, this is a by-product of the manufacturing process.

Are any p-channel devices available without this diode ?
I don't know any, see above answer.
 

Pessimist

Jan 5, 2016
6
Joined
Jan 5, 2016
Messages
6
Thanks Harald ! I think that I may have to live with the external diode as I'm a little pressed for physical space. Incidently, DO you know of a TO92 p-channel device that will handle 1 amp (or thereabouts) ?
Thanks again !
 

Harald Kapp

Moderator
Moderator
Nov 17, 2011
13,699
Joined
Nov 17, 2011
Messages
13,699
Use a Schottky diode to minimize voltage drop across the diode.
 

CDRIVE

Hauling 10' pipe on a Trek Shift3
May 8, 2012
4,960
Joined
May 8, 2012
Messages
4,960
Check out the FDN358P. It's not a TO92 but it is a small SST-3 package.

Harald, there's always more to learn. I had no idea that the body diode was a manufacturing byproduct. All these years I've been laboring under the mis-impression that it's an integral protection diode that shouldn't be relied upon. :)

Chris
 

Pessimist

Jan 5, 2016
6
Joined
Jan 5, 2016
Messages
6
Check out the FDN358P. It's not a TO92 but it is a small SST-3 package.

Harald, there's always more to learn. I had no idea that the body diode was a manufacturing byproduct. All these years I've been laboring under the mis-impression that it's an integral protection diode that shouldn't be relied upon. :)

Chris
I TOO have learnt a lot from this, Chris ! ( I must admit that I haven't a great deal of experience with MOSFET's)
Thanks for the pointer to FDN358P !
 
Top