Team MedBots

[Shakey]

Motors like the pololu HP stall at around 1.6A on 2 cell (2.4A on 3 cell) so really anything just over that should be fine.

[Max]

I thought when using PWM you need a mosfet rated substantially higher than the stall current of the motor as when driving not at 100% duty cycle the midget can see much more current (for shorter times) than the motor is seeing? ie if say the motor was drawing 1a but is running at 50% duty cycle the FET will see 2a half the time and 0a the other half?
Or have I got that wrong?

[Shakey]

You may be thinking of when a motor is reversed which can see extremely high peak currents over the stall of a motor (as the voltage is the battery voltage + the generated back emf of the motor). A motor that draws 1A at 100% with draw 1a at 50% but only half the time making the average current 500mA. There is tiny variation in this by nature of capacitance/inductance of the motor but it's basically not a concern particularly at these levels.

The real thing to make sure is that you can switch the FET on and off fast enough. While it is on fully it has a very low resistance, if it isn't on fully it will have enough resistance that it will heat up quite a bit. Though normally in an antweight just driving the FET off of a microcontroller (PIC/Atmel etc.) is fast enough. It's only really beetle/feather/heavy esc's that need to worry about this.

[Rapidrory]

Or just don't worry about any of that and just use a suitably rated motor driver chip :L

[Max]

Ok cool. So something like this should work fine:
http://pages.ebay.com/link/?nav=item.vi ... 1712256525

Not sure what the 4th pin is for, I guess it's just connected to the drain.
Then a fly back diode like this:
http://pages.ebay.com/link/?nav=item.vi ... 0964398644

To protect the mosfet from back emf

[Shakey]

What do you intend on using it for?

[Max]

Melty brain spinner! I only need to drive each motor in 1 direction so no real point using a motor driver. So plan is to have a mosfet driving each motor then a fly back diode across the MOSFET so that when the MOSFET is off you get braking of sorts.

As in writing this I'm realising that the diode will try to 'brake' the motor between each of the PWM pulses which probably isn't good.

I'll have an arduino mini running the thing and an accelerometer to measure the radial acceleration and hence the rotational speed.

I doubt it will work very well but it should be a fun project.

[Shakey]

Yeah you may aswell just go for a regular driver chip. Or a half bridge driver chip if you really only need single direction. It will probably make the coding behind it a bit simpler and it worries about all the brake/coast/power etc itself.

[Max]

So something like this:
http://www.ti.com/lit/ds/symlink/sn754410.pdf

I guess that doesn't have built in braking but it's not the end of the world if the motors don't need brake. Would I still need a fly back diode?

[Shakey]

From the data sheet I don't think that one is built for PWM switching. They normally say about stuff like that.
https://www.pololu.com/product/713
https://www.pololu.com/product/2135

Those two chips are pretty decent dual motor driver chips. They are bi-directional but that doesn't matter, just only tell them to turn one way. And they come on a breakout board (I had some of the drv8835's not on boards. I think 2 of them are in my carpet now :S ).