Help!

I've ordered some hall sensors. I'm going to try that first. I expect to fix my starting torque problem. If the motor is too weak after that, I'll try going with a 3000w controller (and more batteries)

Aging Wheels

Consensus from my group chat was try the Hall effect sensor first and if that doesn’t work then the stall torque on that alternator is just too low, the 1500 watt system you’re putting in is roughly half the power of the gas engine you took out so you could add a higher gear reduction in exchange for top speed or ideally some kind of cvt system from a small scooter or something similar.

Sensorless motor drivers have notoriously low torq at low rpms because they cannot tell the position of the rotor without the rotor spinning (they use the 3rd winding to look for the magnets passing to determine the position of the rotor), also normal sensorless drivers usually get confused because the magnetic caracteristics of an alternator are quite different than a regular bldc motor (they have a lot more iron in the cores to make them more efficient at being a generator) The best option for you in my opinion is to put hall effect sensors on the motor to help the controller at low rpms. Austiwawa has a great video that explains most of the pitfalls of doing it, but the single most important fact is that you need to place the hall sensors in the middle of the coils not on the edge

József Krisztián Kovács

For your entertainment: https://training.ti.com/sensorless-and-sensorless-techniques-driving-three-phase-brushless-dc-motors

The rule with three phases if it's going the wrong way you swap any two phases. I missed the last minute of the video just saw it now you already identified that if you push it it will start going. If you have more motor questions like this my first job out of college was working at a permanent magnet motor place this is a common problem we always had to have hall effect sensors on motors that had to run below a certain RPM to get them going and then we switched in the back EMF mode and they worked great. I'm not sure if that motor controller has that capability.

Usually those motor controllers use back EMF in place of a hall effect sensor and when the motor is not turning very fast there isn't enough back EMF for it to calibrate and it wobbles like that. Usually on three phase permanent magnet motor setups like that when they need to run it low and high speed they will use a hall effect sensor at low speed and back EMF at high speed. I suspect if you give it a good rolling start it would start going just fine. For a go-kart that would be not terribly useful. I suspect the reason it works when you lift it up off the ground is the initial push is enough for it to get the wheels turning and then the back EMF kicks in and it works fine.

I’m thinking that the alternator, because it’s not designed to be a motor, isn’t designed with a reasonable stall current in mind. When most people make these alternator go kart they use *huge* lithium ion packs or just any old lead acid battery. I think your tiny ryobi batteries can’t supply the current needed to get going. In your place I’d either use a deep cycle car battery or like an ebike battery or something. Or switch to a treadmill battery.

Is it possible to lower the start frequency?

Thomas Eriksen

I will preface this by saying I've never tried to use an alternator as a motor. But I have done some work with smaller 3 phase motors. The thing you'd like to measure is whether the motor/alternator is receiving its stall current. Because this isn't intended as a motor, you probably don't have a spec sheet that will tell you that number, so it could be tough to tell, even if you have a clamp ammeter. For an alternator, the stall current could be very high, maybe over 100A due to low DC resistance. If it isn't receiving the stall current, then the motor could work harder with changes on the electronic drive side (which might include the Hall effect senor others have suggested). If the stall current is already present, then you have a mechanical problem: too much load from the drive train for the motor to handle.

Do you know how much current is being provided to the winding with the 12 v supply? From what I'm reading you need like, 2 or 3 amps. Also did you remove the diode pack from the alternator?

Eric Greenberg

I tried cranking up the power with the bench supply, but nothing changed

Aging Wheels

I don't, unfortunately. I'd like to hook up some big chungus lead acid batteries just to see, but I no has.

Aging Wheels

Are you supplying enough power to the [electromagnet] windings?

Robert Paulsen

Do you have another battery you could hook up, just to test, to see if it will start on it's own with a higher voltage input? The fact that it moves when you give it a little push definitely makes it seem like it doesn't have enough power to start itself under load. I don't think the hall effect sensor (or lack there of) has anything to do with it.

Eric Greenberg

Oh Lord. I'm happy to find another person the same issue. Shame you didn't find a solution. I would've loved to have copied your fix! I think what I'm going to try is adding hall effect sensors to the motor so the controller knows what's going on. It'll be a fair amount of fiddling, but it'll be worth it if it works. If it doesn't work, the gas engine is going back on

Aging Wheels

I don't remember where, but I recall seeing some current bleed on the trailing edges of the brushes at low speeds. At low speeds under load it acted just like this. I think maybe even with the factory (somewhat wide) brushes, a car's idle is high enough that it doesn't have this problem. I don't know if narrower brushes would actually solve this, but that's what I'm thinking could be the issue. I also recently had an issue with a "pure sine wave inverter" powering a motor. Turns out that it was not a pure sine wave and more like a stepped square wave, which caused the motor to have a difficult time starting. Once I spun the motor for it to start, it was fine. But the waveform was trash and causing problems. I never did actually fix the issue; I just got 300 feet of extension cord and plugged it into my house. Hopefully someone else has some more direct and meaningful answers for you instead of ramblings...

Garrett Rabenold

I tried doing a nearly identical setup with an alternator, a really similar 1500W controller, and 4 sealed lead acid batteries in series but was never able to overcome that very issue :( In the end, I ended up buying a treadmill on craigslist and using its motor with a simple DC controller, as it's not 3-phase and it ran fine on the 48V. That worked a lot better for me in the end, but I'd love to hear if you find a fix for it! I figured that the alternator just had too little impedence and was drawing too much at the start for the little controller so that's why I didn't bother much further.