Mosfet Prob

[2camjohn]

I am having a strange problem with mosfets.

When I measure the D-S resistance of my N-Chanel Power mosfet out of circuit it comes out at 16000 on the MegOhm scale, so quite alot.

The problem is when I wire it up to a motor like this
**broken link removed**

The motor is constantly on.

I have exhausted all the possibilities I can think of and I have replaced all of the components.

Can anyone explain what it might be?

Cheers
John

 

[Styx]

swap the drain and source connection over. the internal diode is in an orientation to be always conducting

 

[2camjohn]

Cheers

Ill give it a go.

 

[Styx]

And have that 10k going to the Source as well (to hold off the FET)

and put a diode across the motor terminals otherwise when you start to switch you will blow the FET (backEMF from coils)

 

[2camjohn]

and put a diode across the motor terminals otherwise when you start to switch you will blow the FET (backEMF from coils)

I thought the internal diode(my initial problem) was there to prevent this ??

 

[bmcculla]

The internal diode is actually a product of the internal structure of the mosfet. In most mosfets the source is connected to the bulk silicon in the die. This connection eliminates the effects of the bulk silicon having different voltages (bulk voltages can cause the FET to switch) but creates the diode in the FET. It's not really designed to be used as a diode - at least I wouldn't trust it for more than a little current.

 

[Styx]

and put a diode across the motor terminals otherwise when you start to switch you will blow the FET (backEMF from coils)

I thought the internal diode(my initial problem) was there to prevent this ??

No!!!!

This is a big misconception. The diode needs to be across the inductive load. If you draw out your circuit with the intrinsic diode to the MOSFET you will see that there is no diode in parallel with the inductive load.

If you sit down and think about it you will see that the diode is needed across the inductor :- when you open the FET when current is flowing, what path will the current take to keep on flowing...

 

[Oznog]

1. You do not need that 10k resistor on the gate, the gate does not even draw current.
2. I'm assuming you want to use the transistor as a switch, but you have it hooked up as always on, so you might as well use a wire.
Assuming you want to switch it on and off, there should be a 10k-100k resistor from gate-to-source to drain off the charge. Otherwise the transistor will stay on indefinitely once you remove the source voltage.
3. Styx is right about the diode, it has to go across the motor.
4. BTW, the body diode is usually strong enough to handle the max current of the package. It can be very high. The spec sheet will list the current rating. This doesn't change the fact that it's not in the right place to stop the back emf of the motor. It also acts as a zener, as depicted in the symbol. The zener voltage usually establishes the pkg's max vds voltage.

 

[zevon8]

As a side note, the intrinsic diode in many MOSFETs is not as quick at reverse recovery as the MOSFET is at switching, so often you may need to add a fast recovery diode. This switching time can be a problem in H bridges for motors, where the intrinsic diode is used intentionally for freewheeling. If not enough time is allowed between switching, one MOSFET can be on before the diode in the other recovers, allowing shoot-through, and bang goes your MOSFET.

There are MOSFETs that have quicker intrinsic diodes, designed just for h bridge applications, but many garden variety MOSFETs are not always fast enough.

 

[2camjohn]

Crazy things mosfets! and i thought inside it was clockwork!!!

 

[williB]

i am pretty sure that you cant "measure" the resistance , of a MOSFET ,with a DMM ,when it is connected to a circuit..
you could calculate it though.