Mosfet trouble

K

Kiwi Bruv

Member
I have built a PWM drive with a 555 and an op amp. The output from the op amp drives a trans which drives an n channel mosfet.

I have the emitter of the trans oscillating, so the pwm is working and I can see the duty cycle changing as I tweak a pot.

But the motor isn't changing speed.

I removed the trans from the circuit and the motor still went. ie the motor is being driven by the supply into the drain and the source connected to gnd ie the mosfet is shorted.

So I am assuming my mosfet is toast? But I have a replacement and it behaves the same way.

How can I test them? It must be a faulty mosfet or am I missing something? It is possible I fried them when trying to build an H bridge a while back.
 
E

evanh

Senior Member
You say the motor runs even without the MOSFET. To me, that means the FET is not tour immediate problem, more likely wrong wiring or a bridge in your soldering.

MOSFETs are simple enough to test out-of-circuit, use the diode test on the multimeter and place the probes across the Source-Gate pins, positive probe on the gate for N-channel, then shift the positive probe to the Drain and it should read as a short (0 volts) then put a finger across the Source-Drain pins and the meter should show the FET turn off across the Source-Drain pins.


Evan
 
E

evanh

Senior Member
Be wary. All plain MOSFETs are exceedingly static sensitive! To give an example of how raw the gate pin connection is: If you try my above function test of a MOSFET with just a small zenor doide soldered across the Source-Gate pins you no longer need to use you finger to discharge the Gate but instead you can watch the FET turn off within a few seconds after shifting the probe to the Drain pin as the zenor leakage discharges the Gate.


Evan
 
B

BeanieBots

Moderator
Did you fit a catch diode?
Without one, the flyback voltage from the motor will toast your FET after the first few pulses.
When testing, use an incandescent bulb, they don't give the MASSIVE voltage spikes that a motor does.
 
K

Kiwi Bruv

Member
To be clear. If I take the trans driving the mosfet out the motor keeps going. If I take the mosfet out the motor stops.

So I'm sure it's the mosfet. I am guilty beanie bots of not using the flywheel diode first time I fired it up. Too eager, inexperienced with mosfets and I used a tiny 12V motor - thought it would be OK.

Just wondered how to test a mosfet so I can be sure. Initially the circuit didn't work because of the op amp too. When I changed it for a new one it worked fine.I have a lot of bits and bobs I experiment with and they get used and abused. I might need just to throw it all out.
 
K

Kiwi Bruv

Member
evanh,

One Fet shows O/L across the G/S and also across the D/S

The other shows 0.02V across the G/S and O/L across the D/S.

 
B

bobrayner

Member
It sounds as if you are turning the mosfet on at the first pulse and it is not turning off. Probably due to your driving circuit. A mosfet has quite a bit of input capacitor. You need to charge this cap to turn the mosfet on but you also HAVE TO DISCHARGE IT to turn it off. If you don't arrange to do this the mosfet will stay switched on forever and a day. I think this could be your problem.
cheers BobR
 
B

BeanieBots

Moderator
The FETs should be easy enough to test.
For N channel type, connect drain to 0v, source to a load, other end of load to power supply. Connect drain via 10k to 0v. Load should be off. Connect drian via 10k to power rail, load should be on.
For P type it is the same but swap the polarity of the supply.
If you make the load quite small such as a 1W 12v bulb, the FET will be safe even if you have all the pins connected wrong. This is how I test unmarked or unknown three terminal devices that I get from scrap boards. Current rating is guestimated from physical size and the circuit it was taken from.
 
K

Kiwi Bruv

Member
Ok it must be the mosfet, will get another tomorrow. Using a bulb driven by the mosfet drive tran, the circuit is fine.

The motor runs with the mosfet physically isolated, no gate drive just 12V supply to motor to mosfet drain to mosfet source to OV!!!!!

It must be the mosfet I think.
 
B

BeanieBots

Moderator
The most common failure mode is when the junction melts. This leads to a short between drain and source. (load is always on). It often includes a short to the gate as well which can propogate down through all your drive circuitry destroying that as well. That is one of the reasons why it is important to drive the gate via a resistor.
 
K

Kiwi Bruv

Member
Ok, I have done some experimenting and this is the result.

The mosfets work fine. I made a simple astable at fixed duty using a 555.

If I use a bulb as load, the mosfet is latching on. At 50% duty the lamp is half bright but when I remove the gate signal the mosfet stays on and the bulb goes full brightness.

I have searched the net but can't find out why this is happening or how to fix it.
 
E

evanh

Senior Member
BeanieBots: You got the Source and Drain swapped in your polarity description. A FET's Source is equivalent to the BJT's Emitter and Drain is equivalent to Collector. And N-channel is equivalent to NPN so the N-channel's Source goes to ground/negative while the Drain goes to supply/positive.


Evan

Edited by - evanh on 10/23/2005 9:50:57 AM
 
B

BeanieBots

Moderator
evanh, thanks for pointing that out. I often get my emitters and collectors the wrong way around too, not to mention anodes and cathodes. That's why they put arrows on them, so that people like me can remember which way the current flows!
 
You must log in or register to reply here.
Top