MOSFET and talking dog

What mosfet, what mcu, and EXACTLY how are they connected, what power supplies are you useing, and what about grounding?

martin

The MOSFET is a IRL2703. The MCU is an Atmega48. I have 100ohm resistor between the MCU and the gate. I also have 10K pull-down resistor on the gate to limit any false starts while programming the chip. I have an additional Schottky diode across the MOSFET and one across the motor. The motor is on the high side, going from power directly into the drain. The power for the motor comes from an unregulated source, either a battery at 7.2V or a transformer at ~8V. The MCU is powered by a 5V regulator. The source is tied directly to ground. All the components share a common ground.

the Schottky diodes across the fet and motor sound dubious, try 1n4002 instead. What happens when you short the gate directly to ground, while the mega is sending out a 0?

martin

I'd be thinking that on the spring driven return, the motor is running as a generator and forcing current through the diode load across it. Suggest you put a 10V Zener in series with diode. The motor will generate a few volts but not enough to overcome the Zener voltage. Hence it will return unloaded and at a normal speed.

put a pull down resistor on the gate. or program your uC with pull Up/down.. Some uC use open collector outputs.. etc..

put a diode by pass from the uC output to the gate via some small resistance to not over current the uC output so that the off state of the uC will have greater influence on the gate discharge time.

First get rid of the diode across the mosfet. It serves no purpose. Second, what you see is due to the diode across the motor which shorts it when it returns.

Your power sources suggest that the motor current isn't so high. That plus the fact that your mosfet is avalanche rated suggest that you simply use it as is, without any diode. What are your peak current (when the mosfet is switched off) and it's duration and also the motor cemf voltage when it returns (open circuit)? Compare this to fig 12c diagram and also the cemf + Vsupply to the 30V V(BR)DSS and you'll know what to do. My bet is that you can simply go with the mosfet alone. If you want to protect something, then clamping the gate voltage to GND and +5V with a pair of small sig diodes (1N4148/BAV99 or something) will be more useful.

Removing the diode from across the motor did the trick. The motor spring return now works perfectly. I also removed the diode from across the mosfet as suggested. The motor's max current draw is far less than the rated current draw on the mosfet. Thanks!

In what situations does it make sense to use a diode across the motor terminals and what kind of diode should be used?

Sam

samchristy a écrit :

For example if you want to PWM the motor to reduce it's speed. Here the diode is mandatory and serves as a 'free wheeling' diode (allow the current to flow during the mosfet of period, thanks the the motor inductance). Use a fast recovery diode, or maybe a shottky if the supply voltage is low enough. But it is more usual to use bridges (half or full) depending on what you need.