Power MOSFETs

I am using the alternative Arduino Mega/Pololu Carriers Electronics

Anyone have any advice for driving the extruder heater without the extruder controller?

Some Questions:
-Is there something specific I should be looking for in a power MOSFET (I assume sufficient voltage and current are the main points)?

-Do I need anything other than the MOSFET?

-Wiring advice. I assume i'm driving this off the PWM output of the mega, and the 12V/Ground from the power supply, how do I tie the grounds together safely. The Mega will be running off USB power so the grounds are isolated right now.

-What am i missing? i'm a programmer not an EE, just had the basics in school.

Edited 1 time(s). Last edit at 03/16/2010 04:49PM by mfsamuel.

Are the steppers wired up and working? If the grounds really are isolated, you won't be able to send any commands to the stepper controllers. You're right that MOSFET voltage and current are the things to look out for, its worth getting ones that can handle a much higher current than you need, because they'll get much less hot and may not even need heatsinking. You should pull the MOSFET gate low with a resistor as well, or the MOSFET partially conducts and gets hot.

The other thing with MOSFETs is to try to pull them high as fast as possible; typically, you them driven at lower than threshold voltage, or higher than saturation voltage. The shorter amount of time spent dwelling between these two voltages, the less heat will be generated by current flowing thru the gate during the transations.

Also, N Channel MOSFETs typically have about 1/3 resistance than comparable P Channel at saturation, so should generate less heat.

Adrian has published a complete circuit here: [objects.reprap.org], just follow that.

It uses a protected MOSFET suitable for logic drive for the heater.

Quote

The shorter amount of time spent dwelling between these two voltages, the less heat will be generated by current flowing thru the gate during the transations.

No it isn't current flowing through the gate that causes heating. When the MOSFET is off the current is zero and when it is on the voltage is close to zero, so in both states the power dissipated is small. During the switching transition the MOSFET is half on so has significant current and voltage, hence a lot of power is dissipated for the time of the transition.

---

[www.hydraraptor.blogspot.com]

looks good for the heater. it is exactly what i was expecting, and the led is a nice touch.

i will need to cut the thermistor part out since I have a thermocouple.

---

-----------------------------------------------
[mfsamuel.blogspot.com]

The key points for a suitable mosfet are low Rds(on) at Vgs=4.5v and Tj=175c. I'm using the IRL3803 to drive my heater, and others have mentioned the STP55NF06 and FDB8880. All these mosfets have Rds(on) of 0.02 ohms or less under the above conditions, and can pass several amps with no heatsink.

Note that many common power mosfets (such as MTP3055E and IRF540) have unacceptably high Rds(on) at Vgs=4.5v and don't even mention the figure at that gate voltage so you must read the datasheet carefully.

The advantage of a mosfet rated for Vgs=4.5v is that it can be driven directly from your arduino with no buffers or voltage shifting.

Read FET Current Ratings from Chuck's Robotics notebook if you want to calculate the exact current limits, or conversely how much heatsinking you need.


As for turn on and off time, I wouldn't worry too much in this case. At 10mA, the IRL3803 can turn on or off in about 14 microseconds and I think the atmega can supply a bit more than that in pulses. I'm driving my IRL3803 with a PWM frequency of 62.5KHz and it barely gets warm with a 2A load. If I was overengineering, I'd put a 10 ohm resistor between atmega and gate just to limit peak gate current.

---

-----------------------------------------------
Wooden Mendel
Teacup Firmware

nophead Wrote:
> No it isn't current flowing through the gate that
> causes heating. When the MOSFET is off the current
> is zero and when it is on the voltage is close to
> zero, so in both states the power dissipated is
> small. During the switching transition the MOSFET
> is half on so has significant current and voltage,
> hence a lot of power is dissipated for the time of
> the transition.

Perhaps I did not get the terminology correct? ... I meant to express that MOSFETs acts as a resistive element in the power circuit -- power dissipation is proportional to resistance. 12 volts flowing through two resistive elements, the nichrome element at 6 ohm, and a unsaturated MOSFET at 1 ohm will dissipate ~10 watts thru the heating element, and another ~2 watts thru the MOSFET. Compared to 0.1 ohm resistance saturated MOSFET - ~11 watts of heat produced in the heater vs. 0.3 watts thru the MOSFET. You want the MOSFET to either have an extremely high resistance (overall power extremely low), or a very low resistance. Dwell time between these two values should account for the majority of heat generation within the MOSFET, if using PWM, and pulse rate is high.

I suspect we are saying the same thing here, though.

Well yes it was "current through the gate" that is not correct. It is current through the channel that causes the heating. Gate current only flows to charge and discharge the gate capacitance and isn't a significant cause of heating.

But now you have made it worse by using "saturated" incorrectly ;-)

Unlike a bipolar transistor which is "saturated" when the voltage can't get any lower, a MOSFET is "saturated" when the current can't get any higher and the drain source voltage has therefore started to rise. So you want a FET to be unsaturated for minimum dissipation. Confusing terminology but that is the way it is.

---

[www.hydraraptor.blogspot.com]

hi
give the MOSFET information . how to creat the saturation & non saturation region