I was surfing through some of the worksheets here, and I have a question about the answer to Question 23 on http://www.allaboutcircuits.com/worksheets/igfet.html. The answer shows a diagram virtually identical to a circuit I'm designing right now, to use an IRF510 MOSFET to turn on a high-power (1 watt) LED. The answer says the diagram shown is not the most practical, and challenges the reader to improve on it. Since I'm not taking an electronics course at this time, could someone help me out and tell me how that circuit would be made more practical? For the record, in my application the MOSFET's gate is to be driven by an output pin of a Parallax BASIC Stamp, with the external pull-down resistor to insure that the MOSFET is turned off if the Stamp's pin is set to input mode and therefore would be floating otherwise.

 

Since you're working with logic levels (the Basic Stamp) you really should be using a logic level MOSFET like an IRL530 or IRLZ24.

The uC to gate connection should have some resistance (10-20 Ohms). If you want to use PWM on the LED, then you'll need a gate driver circuit, as the Basic Stamp can only source/sink around 20mA, which would be woefully inadequate for anything more than a few hundred Hz.

Use a 10k resistor from the gate to the source to keep the MOSFET turned OFF in case you leave the pin floating by accident, or if the output pin itself gets burned out due to overstress.

If you insist on trying to drive an IRF510 from a Basic Stamp, your driver circuit will be rather complicated, I'm afraid. I'm away from home, so I don't have any examples to show you at the moment.

 

Okay, I figured correctly on the 10kΩ pull-down resistor. Can you explain why the series resistor on the gate is needed, and why the Stamp's source current is inadequate? When the I/O pin is high, it would only need to source 0.5mA through the pull-down resistor, since the MOSFET's gate in parallel has a much higher impedance.

Actually, I misspoke in my original post. It's not the Stamp itself driving the MOSFET gate, but a Parallax PWMPAL sitting under the Stamp. Unfortunately the only datasheet available for the PWMPAL is at http://www.parallax.com/Portals/0/Downloads/docs/prod/acc/PWMPAL.pdf, and it doesn't state what the source current on the outputs is. Unfortunately, Parallax is closed until after January 1st, so I can't contact tech support until Friday at the earliest to get that info. But the sample schematic for motor control does show a 220Ω resistor in series with the gate of the MOSFET (2N7000 or similar).

Fortunately, with the PWMPAL I have control over the on and off times on the outputs, so I can set the duty cycle of a PWM drive signal to whatever I want, at a wide range of frequencies, so keeping the frequency below a few hundred Hertz is not a problem. I was planning to use a duty cycle of around 90% to reduce the average power dissipation somewhat, during the operating mode where the light is to appear to be solid on. During the other modes, the LED will be strobed at 10Hz and 2Hz at 10% duty cycle.

 

A-ha... I just read through the MOSFET page on this site, and now I understand the presence of the gate resistor; it limits the inrush current when charging the gate at turn-on. That's a detail about MOSFETs I wasn't aware of.

Okay, here's another design question regarding MOSFETs. Would a typical circuit using the MOSFET as a current driver (in this case sinking the current through a high-power LED) have its on-state operating point on the far left side of the transfer curves in the linear region, or to the right of the knee in the saturation region? What are the pros and cons of either case? I'm going to guess that if the MOSFET remains in the saturation region when turned on as well as off, that it's going to dissipate more power as heat than if it were to go into the linear region when switched on. Am I interpreting the MOSFET's operation correctly?