I keep killing my MOSFET, why?

[fouadalnoor]

Hello guys,

I am currently trying to run a motor off a chip (PICAXE 08M) and I am using a 2N7000 N-Type Enhancement MOSFET. The gate of the MOSFET is connected to the output of the chip (~5v) while the source is connected to Ground and the Drain to 5v supply. (I have also connected a diode in case of any back emf..)

Every time I try to test if the MOSFET is working it seems like it dies. I dont actually connect a resistor in series with the gate though, but I thought that no current would go into the gate anyway so why would I need to?

Can someone please explain why I must connect a resistor in series and if so what value? (the highest one I have is 400k, is that enough?)

Thanks.

 

[ericgibbs]

hi,
With a 5V Vgs voltage the 2N7000 will not be hard ON.

What is the current drawn by the motor.???

 

[crutschow]

Where did you connect the diode? It has to be directly across the motor (cathode to positive) to do any good. Otherwise the back emf of the motor is likely what's blowing the transistor.

 

[fouadalnoor]

see. at this point I'm not even connecting the motor to it. I would like to just "see" that the voltage between the supply and the drain goes from 0v to 5v as I change the gate voltage from 0v to 5v.

Can I just put 5v on the gate (if I dont have anything else connected to the MOSFET) and then put 5v on the Drain and Ground the Source?

 

[Mosaic]

Hello guys,

The gate of the MOSFET is connected to the output of the chip (~5v) while the source is connected to Ground and the Drain to 5v supply.

Thanks.

You say the Drain is connected directly to 5V? How is the motor connected. Sounds like you driving the FEt into a dead short situation.

 

[fouadalnoor]

You say the Drain is connected directly to 5V? How is the motor connected. Sounds like you driving the FEt into a dead short situation.

Well initially I connected the a diode from the Drain to the supply. The Anode is connected to the drain and the cathode to supply. The motor is connected from the Anode to the cathode.
I connected the PIC's output (the output is 5.17v) to the gate directly (without any series resistor).

Does this seem right?

 

[fouadalnoor]

Ok, now I have fixed the problem with teh MOSFET (I guess you can just put 5v on the gate without a resistor) but I am having major problems with my PIC. I programmed it's output to be high as soon as input3 is high and input4 is low. But for some odd reason the output flickers is low. and then after about 14 seconds it goes high...

now I just dont understand why it would do that...

when I program it with nothing (so its just blank) it takes about 16 seconds of flickering before it goes low.

Is this due to noise or? I mean I am using long wires and I dont have any decoupling caps (but I did try them..didnt seem to help much)

Any ideas?

 

[Mosaic]

Well initially I connected the a diode from the Drain to the supply. The Anode is connected to the drain and the cathode to supply. The motor is connected from the Anode to the cathode.
I connected the PIC's output (the output is 5.17v) to the gate directly (without any series resistor).

Does this seem right?

You mean Supply +ve is going to the motor, then motor to Drain then Source to ground and gate to MCU output? With a rev. biased fast acting diode (1n4001..not 1n4007) across the motor?

That's what you should have. Simple way to verify this safely is with an LED & 1k resistor as the Nfet load.

Yes u need decoupling 0.1 uf across the MCU power pins. Is your supply regulated so that the motor does NOT introduce noise onto the MCU power lines? Bad idea to share MCU power with inductive loads without decent regulation and a decent pi filter plus a sprinkling of ferrite beads and decoupling caps.

 

[fouadalnoor]

You mean Supply +ve is going to the motor, then motor to Drain then Source to ground and gate to MCU output? With a rev. biased fast acting diode (1n4001..not 1n4007) across the motor?

That's what you should have. Simple way to verify this safely is with an LED & 1k resistor as the Nfet load.

Yes u need decoupling 0.1 uf across the MCU power pins. Is your supply regulated so that the motor does NOT introduce noise onto the MCU power lines? Bad idea to share MCU power with inductive loads without decent regulation and a decent pi filter plus a sprinkling of ferrite beads and decoupling caps.

ahh.. I didnt knwo noise made SUCH a difference. Now it seems to work (I cut my wires much shorter and used the cap..)

 

[fouadalnoor]

I have now added a 0.1uF Cap between the supply/gnd pins of the PIC and have cut most of the wires short... The cips output works fine, but as soon as I connect it up to the motor it will start/stop constantly..

it must be a noise problem..but I dont know how to fix that now? Even shorter wires maybe? (also I have added a 470uF polarised electrolytic cap between the 5+ and the gnd rails..)

 

[ronv]

You can try some .01 ufd caps. 1 across the motor leads (at the motor) and one each from the motor lead to the motor case. Also note Eric's post about the 2N7000 not being turned on all the way.

**broken link removed**