Controlling two motors - AWD Lawn mower project

[cimpok]

Hello, I am new here looking for technical advice from Germany.

I am planning to mount an additional electric front wheel drive on my petrol lawn mower (ride-on) tractor.
The tractor has a rear wheel drive which sometimes does not have enough grip to climb on the hilly terrain I have to mow, there are more expensive all-wheel drive models that solve this problem, but I plan another solution.

My idea is to buy 2 pcs of so-called electric wheelbarrow kits, these are geared high-torque BLDC hub motors, I am looking at those with 15 inch tyres mounted, These have enough torque (50-70Nm per wheel, so up to 140Nm for the whole front drive) at starting/low speed that would perfectly solve my problem if I mount these instead of the current (free-wheeling) front wheels. I plan to go for a 48V / 500W version per wheel, and a 48V 15-20Ah battery pack. (The battery capacity is not an issue, maybe can be even lower, I will need the extra grip of the electric front drive only for some minutes on occasions , not the whole time mowing)

My questions:
Is it OK I use the electric wheels at undriven state (either all power cut-off or with the controller input in neutral), would they free-wheel without a significant drag? is it Ok for the controller's electronics that they rotate? Or perhaps a future upgrade: Could I use these wheels to regeneratively charging the battery?

I will have separate controller per wheel, but need to exactly synchronize their rotation speed. I plan to use the same analog input line split parallel into the two controllers (left + right wheel). The analog control would come from my electronic board with a microcontroller driving a DAC outputting the 0-10V signal, the microcontroller would sense the either the rear wheel's rotation with an encoder or the throttle position, have not decided yet, will try perhaps both. So I will not have to manually adjust the front wheels speed, will have only a RWD/AWD main switch.

- Do I get there very same rotation when the analog input is at the same voltage level?, are these controllers reliable in this way?
- Are there controllers with digital input (RS232/UART/Can bus/etc) that would accept a Speed/sine wave Frequency value as input, instead of the inherently less precise analog control?
- I plan to sense the steering wheel position so I could calculate turning radius of the vehicle and apply the needed speed difference between the inner and outer wheel in curves (an electronic / software version of a differential) What do you think, does it worth the hassle, or is the skid negligible on a grassy/soft terrain anyway?

Thank you all who can help with any clues/ideas.

 

[csailor]

That's requires a lot to make work. Skype just to start

 

[CMD130]

This is f-ing brilliant! I do hope you get it to work. I suspect you will. I love doing projects like this

 

[slomobile]

would they free-wheel without a significant drag?

No. You will have the inertia drag of the motor multiplied by gear ratio, plus friction at each stage, multiplied by the remaining gear ratio to the ground.

is it Ok for the controller's electronics that they rotate?

Probably ok if only driven slowly, depending on the controllers chosen. If the controller has any LEDs, you might see them light up from regen current. Definitely stop if those LEDs get very bright. Going too fast (unpowered, PM motor) can generate higher regen voltage than the absolute max withstand voltage of caps, transistors, or other crucial parts in the controller.

future upgrade: Could I use these wheels to regeneratively charging the battery?

Not future. Regen in a spinning PM motor happens whether you want it or not. When powered, it is called back EMF. It always has to go somewhere. Braking torque, vibration, heating coils or laminations or resistors, ionizing the air, breaking down semiconductor junctions, charging caps, or charging the battery. Choice of controller topology and programming determines how much goes toward each.

I will have separate controller per wheel, but need to exactly synchronize their rotation speed. I plan to use the same analog input line split parallel into the two controllers (left + right wheel).

- I plan to sense the steering wheel position so I could calculate turning radius of the vehicle and apply the needed speed difference between the inner and outer wheel in curves (an electronic / software version of a differential)

Those 2 different ideas are at odds. You cannot send exactly the same signal all the time, and also a differential signal sometimes. Splitting the cable is a first step just to make sure all your hardware functions at a basic level. Then immediately after a straight line test you need to reconfigure and work on your programming. Sending the same signal when you need to drive straight and differential signal when you need to turn.

- Do I get there very same rotation when the analog input is at the same voltage level?, are these controllers reliable in this way?

No. If EVERYTHING is equal between the 2 wheels, their speeds will be equal. Even with no load, everything is not equal. A slightly tighter wind on one motor can make it spin faster with the same voltage. But it doesn't really matter because the dominating force affecting the motor RPM will be ground friction. That force is only overcome when the current in the motor is substantial, so there is a continuous balance happening between motor torgue(current) and ground friction that equal RPM. Sometimes one contributes more to RPM, sometimes the other. That gives you a rather large fudge factor that is forgiving to your torque vector programming, especially because you have a steering wheel to provide additional steering force. A naive turning circle to RPM calculation will get you close enough.

I know this is a necropost. Sorry. But it was one of the only on topic posts when I searched "differential steering" after joining moments ago. It encapsulated so many of the questions I had when starting out, that I wanted to answer them for posterity. For the next guy searching "differential steering".