Can the L298 be used to drive a discrete MOSFET H-Bridge?

[ConnorGames]

I am trying to drive the motors on a robot (4*2A Stall current), and am trying to find a way to make a cheap h-bridge. Sadly, FET drivers are quite expensive, especially when you need to drive 16 FET's. I am wondering if anyone here knows how/if the outputs of the L298 (2 h-bridges/chip = 4 half-bridges per chip so I would need 4) could be used to drive the FETS in a discrete h-bridge(im planning to use all n-channel, NXP PSMN022-30PL. If not, can you come up with an solution to drive these motors? At the low voltage I am using(7.2v) I cant afford much voltage loss, so the lower the on resistance the better. I am willing to spend about $35. Thanks - Connor

[shebu18]

Why do you not use 2xL298?

[Psi]

What frequency/pwm do you plan to drive the motors at? Or is it just for on/off reverse?
2A isn't much, you could probably drive them direct from a microcontroller i/o pin.
Maybe with a transistor or two to control the two high side fets.

Or drop the micro and just make a driver out of discrete components.

[PSR B1257]

planning to use all n-channel

That's not working, from a single supply voltage. N-channel MOSFET need a positive Gate-Source voltage, which you can either achieve with a bootstrap-circuit (but than you cant switch the highside FTEs continuously), or with an voltage at least another 7V above the 7.2V rail.

can you come up with an solution to drive these motors?

Yes. Build the H-bridge with P- and N-channel MOSFETs. That's much easier.

If you're not operating on high frequencies, you don't need a special driver. A simple bipolar transistor an a resistor, that's it.

[hlavac]

You could use two pins of the micro to make a simple charge pump to get the higher voltage for top  N-MOSFET.
You know, that diode/capacitor ladder. Then switch it onto the gate with a PNP driven by NPN driven by micro. You could even tap the middle for the voltage for the lower MOSFET. If you have PWM unit left you can use that, or toggle the pins in a timer interrupt routine.

Only drawback is it needs to charge first and you really don't want to switch the MOSFET on in the immediate vicinity of its threshold voltage or you could burn it (high resistance=lots of power dissipation!)
 I would stick a high resistance voltage divider from the high voltage storage cap to an input pin to the micro and monitor the voltage as a failsafe...

With a continous PWM you could even be able to drive a charge pump directly from the PWM signal...


BTW driving MOSFETs from a micro would be a nice topic for a video - Dave?, and a nice technical discussion topic...

[ConnorGames]

I was thinking about using a charge pump for the high-side fets. I also planned to use pwm, which is what amkes things tricky.\

  And I am considering 2xl298 channel per motor, but I would like to build a more powerful/efficient bridge if possible(for other projects and learning experience)
Thanks for the advice -- Connor

[hlavac]

Or make a freaking boost converter controlled by the micro! Thats just two pins - switch driver and feedback pin... one NPN transistor for a switch, about 3 resistors, big 330uH inductor, diode and storage cap, maybe extra cap and resistor for a snubber to dampen the turn-off transient/ringing...

[Psi]

Or you could use high current (transistor/fet/igbt) digital push/pull opto-isolators for the bridge, that way you don't require higher voltages to turn them on/off.
The switching is all leds which you can have to ground reference.

They will be a little bit more expensive than a standard fet but still not too bad.
One 3A push-pull opto might even be able to replace both the high and low fet. So you'd only need two for a full H bridge.

Maybe something like http://www.digikey.com/product-detail/en/FOD3182S/FOD3182S-ND/2804329 (SMD)
or
http://www.digikey.com/product-detail/en/HCPL-3120-060E/516-2197-5-ND/1236185 (DIP)

(Note: i've not actually tried to use optos as an H bridge myself, but it should work. Unless anyone can see a problem ? )



Only potential issue i see is checking the opto can handle the continuous current of your motor.
High current optos are often used as gate drivers for fets so the datasheet is often only concirned about the peak current and not continuous.
However the datasheet for the DIP chip above says +/-2A in the 'recommended operating conditions' section so it should be able to handle it and if your motor is 2A stall then it shouldn't draw that while running normally.

[ConnorGames]

Thinking about going with the boost converter+BJT. Sounds good. I wanted an excuse to play with power FETS, but I agree the BJT is the most practical and economical option. Thanks - Connor