24v H-bridge for wheelchair motors

Hello everyone,

I have been trying to create a h bridge rated for 24v. Everything i have tried i have broke i have fried my arduino and multiple mosfets. Has anyone created an h bridge for wheelchair motors that has worked and is somewhat basic that i could myself? I have friends at digikey and can get the parts cheap just cannot manage to hook them up correctly. I dont have the money right now to buy the sabertooth controller either and heard they burn out :s… When i say basic h bridge i mean something like this

I am unsure if i need to ground the arduino to the same ground as the battery. Also do i trigger the p channel mosfet with the ground drain from the n channel mosfet :s basically if someone could draw this out for me as basic as they can and what parts i need i would appreciate it greatly.

Thank You

Everything i have tried i have broke i have fried my arduino and multiple mosfets.

So - how much money have you lost in burned up parts so far?

I dont have the money right now to buy the sabertooth controller either and heard they burn out

Had you not gone down the road of "build it myself", would you be able to afford a pre-built h-bridge?

First off - you're learning the hard way that, to control large and current-hungry motors, such as ones for a wheelchair - takes a lot of money, one way or the other. Your way is "the learning way" - although I tend to wonder what, if anything, you have learned. You must have learned something from the experience though, if nothing else than "not to build something that way".

Ultimately, you need to match your controller to the stall current of the motor (plus 20 percent or so); in other words, once you know your stall current, add 20 percent more current, and build the controller to that spec.

Choose your MOSFETs such that the current they are spec'ed is about double what you current needs are; you'll likely need to parallel your MOSFETs, if they are rated at 20A each, and your motor has a stall current of 100A, parallel 15-20 (and/or get higher amperage MOSFETs). Be sure to bolt them all to a hefty heatsink, with heatsink grease and such (be careful about insulating the tab from the heatsink, if they share the tab with a pin - it all depends on your design). Make sure you use the proper resistors to make the "turn off" happen rapid enough.

Even with all of that; and I am no expert - someone will probably eviscerate the above (honestly, for anything like a wheelchair motor, you are much better to purchase than to build, unless your goal is to become an expert on h-bridge design and you have a fat wallet) - you'll probably still blow and burn things.

The sabertooth controller isn't bad, from what I have heard - provided your motor fits its output. Better options (though much more expensive) are the Vantec motor controllers, as well as Roboteq motor controllers (hella expensive). But again, anything can burn if you don't match the controller to the motor's needs.

Have you tried building a hybrid h-bridge?

Basically - use relays designed for inductive loads with huge current capability (some automotive relays will work great for this - such as "reversing" relays for winches used on 4x4 vehicles) - but put one (or more - depending on current needs) n-channel MOSFET between the low-side and ground. Switch the relay (for direction control), then PWM the low side with the MOSFET for speed control (just don't switch the relays before shutting off the PWM, or you'll weld the relay contacts).

Money I have blown depends... I mean i tried hacking into the wheelchair controller which was probably worth 150.00 on ebay and that went up in smoke. Plus arduino and various things probably close to 200 bucks :~ Maybe I will try one more h-bridge and if that doesnt work wait to purchase the sabertooth. Do you have any links to those other controllers you mentioned? I think my mosfets are big enough I just dont think I am wiring them up correctly. But even when I wire them up just one mosfet to the motor it fries. I hook battery ground to the source and the drain to motor. Before i even hook the 24v battery + to the gate the motor starts. So I dont know if maybe the mosfet is bad or what?

I'm about to start building a motor control for a amp flow motor and at 285amps they are a beast! I was considering using a few mosfets…

Perhaps they would work for you. They are only a $1 each!

I have actually built a system like this. 24V wheelchair with the motors controlled by an arduino. The way we did it was to find a PWM motor controller that fit our build. found 2 of them on ebay for under $50 total. They have been running for 4-5 years now, and still going strong.

When dealing with high-current motors the "Cheap" way of doing things is almost always more expensive in the long run. maybe instead of asking for the cheap way to control these motors.... you should look for the RIGHT way.

Do a simple ebay/amazon search for "PWM motor controller" and go from there. I could be wrong, but i think we used "Victor 884" controllers.

The first thing you need to know is the stall current of your wheelchair motors. If you don't know it, measure the DC resistance (take several readings, rotating the shaft a little between readings, and select the lowest) and divide that into the supply voltage (which you implied is 24V). Then search eBay, Pololu etc. for suitable motor H-bridge motor drivers. Possibly something like http://www.ebay.co.uk/itm/BTS7960B-43A-Motor-Driver-H-Bridge-module-PWM-Control-smart-car-For-Arduino-/271280531552?pt=UK_BOI_Industrial_Automation_Control_ET&hash=item3f29946860 or http://www.ebay.co.uk/itm/240W-High-power-H-bridge-PWM-Motor-Driver-Sodule-smart-Car-Driver-For-Arduino/121058925165?rt=nc&_trksid=p2047675.m1851&_trkparms=aid%3D222002%26algo%3DSIC.FIT%26ao%3D1%26asc%3D17235%26meid%3D1429921756862394611%26pid%3D100005%26prg%3D8045%26rk%3D2%26rkt%3D5%26sd%3D271280531552%26. I advise against building your own high-current H-bridge unless you really know what you are doing.

I found this link and was thinking of try to get the same mosfets and resistors and try to build more of a basic h bridge. I just dont know how to hook it to the arduino? Is the only thing coming from the arduino the pwm or do you have to hook the arduino ground to the battery ground so that it actually gets the pwm signal. :~

Probably best to get one of the 43A h-bridges like below instead of trying DIY. The bottom discussion is worth a read if considering DIY.


The first thing to realise about high voltage high current H-bridges is that protection circuitry is all-important,
you have little/no chance of success if you don't do these things:

  • detect gross over-current and shutdown in a few microseconds. Shutdown means shutdown, not
    temporarily disable bridge-drive and switch it back on again when the current falls. (Or use massively
    overrated devices (for instance 100V 200A 500W) and great care)

  • ensure zero chance of static shoot-through currents in each half of the bridge - get this wrong with 24V
    of lead-acid battery as the supply and MOSFETs explode (many kilowatts of dissipation) - wear eye protection.

  • snub out high-voltage spikes from inductive load of the motor.

  • prevent dynamic shoot-through currents - these cause lots of heating and inductive interference on
    nearby circuitry.

  • Have strong decoupling on the supply rail right next to the power devices, and pay attention
    to stray inductance and high-frequency design - you want to prevent supply rail voltage spiking
    as much as possible.

  • protect gate-source circuit from over-voltage in both directions - stray inductance in the source
    circuit adds voltage spikes to the gate-source circuit, these must be tamed.

  • drive the gates properly with high-current MOSFET drivers so the gate-drain capacitance doesn't
    overwhelm the gate drive - you cannot get away with 20mA of gate drive if the load is switching 50A
    at dozens of volts, the gate-drain capacitance can be pushing higher currents into the gate.

  • handle undervoltage of supplies by full shutdown - any state where the gate drive isn't fully-on or
    fully-off will dump hundreds of watts into the MOSFETs through higher-than-normal Rds(on).

  • use an oscilloscope and dummy loads of various sorts during commisioning - start
    with low current loads and low supply voltage, build up to higher voltage, higher current all the
    while monitoring switching waveforms and looking for issues.