Abnormally high rpm on overvolted escooter.

[ibra2672]

I have this 36v 350w 8.5inch escooter, which I decided to overvolt. So I bought a 500w 48v controller for it. Just for reference, with no load the scooter with the old controller did 30kmh max (also cuz the display couldn't display more than 30kmh). With the new controller and no load it would do 60kmh, which i thought was normal. All of a sudden it did 86kmh, which kinda scared me. After that it still does 86kmh. I don't know why it did 60kmh then suddenly decided to do 86kmh. Could it be that i messed up the hall sensor? The hub motor has a 5 pin hall sensor while the new 48v controller has a 6 pin. The extra cable from the 6 pin is white, while the rest of the cables match.


This is the new controller and display:
https://a.aliexpress.com/_EzSqZtD


The old controller and display:

 

[amberwolf]

what is teh no load battery current? it should be very low (an amp or two), and just about the same with either controller at any voltage.

if it is not, it is probably the wrong phase/hall combination. wrong combinations can "advance" hall timing, which can cause higher speed, which usually takes more current.


since wiring colors are essentially random for various controllers, matching by color doesn't work, and even positions within the connector for "standard" connectors, for which phase is which or which hall signal is which may be random (becuase the end soldered inside the controller is essentially random).

(note that "random" means that whoever solders things up at the factory might or might not follow a wiring diagram (if they are given one), and even if they always do, controller company x might or might not use the same wiring pattern that motor company y uses, etc. )

the white wire is usually for a speed sensor, somtimes a temperature sensor, rarely a multiplexed combination of both.

 

[ibra2672]

The new controller has also a self learn cable, could that solve the issue? I tried plugging but the only thing it did was make the wheel spin the other way.
Also with the old controller the acceleration to max speed was much slower, like to 2 -3 sec to reach the max speed of 30kmh while with the new controller it doesn't even take 0.3-0.5 sec to reach 86kmh. I think the current draw with no load might be a lot different between the 2 controllers.
I'll try to measure the amperage.

 

[amberwolf]

The new controller has also a self learn cable, could that solve the issue? I tried plugging but the only thing it did was make the wheel spin the other way.

If you've used the self-learn per the specific controller directions (most of them use it the same way), then the combo would be correct for the direction it's spinning in, assuming the self-learn works as it should.

If no-load current is low, and nothing is getting hot either no-load or under load, then the combo is probably correct.

Max no-load speed is generally proportional to battery voltage, assuming full throttle and no speed limiting in the controller. So if you had a 36v battery that only gave 30kmh (no load speed?) at full charge, and a 48v battery that gives 86kmh (no load speed?) at full charge, that's far from proportional (48v/36v= 1.34x higher, thus about that much faster...30kmh x 1.34 = 40kmh).

So either the new speed reading is wrong, the old speed reading is wrong, both are wrong, or something (timing advance, etc) is causing the motor to spin far faster than it should on the new controller, or something is limiting the speed on the old controller to much less than it should be able to go for the raw voltage supplied.

You also stated that the old display couldn't go past 30kmh; if the motor is actually spinning faster than that then to know what speed it is actually going you'd need some other speedometer. If you don't know it's actual speed you can't compare the old controller to the new one.

There are apps for phones that can use the camera as a tachometer, to give you the wheel rpm offground, then you can use online calculators for RPM to kph.

Also with the old controller the acceleration to max speed was much slower, like to 2 -3 sec to reach the max speed of 30kmh while with the new controller it doesn't even take 0.3-0.5 sec to reach 86kmh. I think the current draw with no load might be a lot different between the 2 controllers.

no-load (off-ground) current has just about nothing to do with the current under load (in use), or controller current limits--it's more about the specific motor itself and the way the controller is driving it (and whether there is non-neutral hall timing, if it is a sensored controller and motor).

So it should still be just about the same no-load current for both controllers, for the same motor, etc.


The faster acceleration new vs old can be because of a higher current limit on the new controller, and/or the higher voltage, either or both of which would provide more total power to the motor.

 

[ibra2672]

Ok, here is the result of my test.
This is with full throttle and no load, with the highest level of assistance.

Peak current during acceleration: 5.64A
Constant current once max speed has been reached: 0.76A
Max speed: 86km/h
Voltage: 54.65 V

As you suggested I also tried to measure the RPM of the wheel with an app on mobile called Strobe RPM Tachometer. I don't know how accurate it is but I got an RMP reading of about 1800 RMP, which equals to 73kmh.

And btw, my old display wouldn't show more than 30kmh even if I was going faster than 30kmh. I think that particular model of display is locked at 30kmh by software.

 

[DanGT86]

Your no load current doesn't seem too bad so I don't think its an issue with the incorrect hall and phase combination. If the Hall and Phase wiring is wrong you will draw lots of current and the motor will get hot really quickly.

I suspect what was happening before is that you were actually going faster than 30kph with the old setup.

A 36v nominal system is 37v avg. You are now running 54v so you should be running about 1.5x faster.

A lot of controllers have a 3 speed switch option. They usually default to speed 2 of 3 without the switch hooked up. Its entirely possible that the old controller was not running full speed and your new one is.

The good news is you don't have to ride it that fast. Keep an eye on the current and motor heat and if all seems well I would just ride it how it is and not worry about it.

 

[ibra2672]

My nominal voltage is 48v, which i think is still 1.3x faster than the 37v nominal because 54 volts is like a fully charged 48v battery. A fully charged 37 volt battery is 42v. So 54 divided by 42 is still 1.3x.

 

[amberwolf]

As you suggested I also tried to measure the RPM of the wheel with an app on mobile called Strobe RPM Tachometer. I don't know how accurate it is but I got an RMP reading of about 1800 RMP, which equals to 73kmh.

And btw, my old display wouldn't show more than 30kmh even if I was going faster than 30kmh. I think that particular model of display is locked at 30kmh by software.

To solve this problem, you'd have to do the app RPM test on the old controller as well. Otherwise you don't have a valid comparison of speed from the old controller to the new one to know if it is actually abnormally high now or not.

 

[ibra2672]

Ok. So i tried riding it. Big problem. It doesn't even go forward and it does this strange sound like a mechanical screech, like it's spinning inside.

 

[DanGT86]

That would lead me to believe that your hall and phase wires are not matched properly. You probably found one of the combinations that is so advanced it spins super fast but uses too much current. It's likely overloading trying to move you.

Not a big deal but you will need to find the correct combination by swapping the wires around.

 

[ibra2672]

I think this might have something to do with what happened before I wanted to overvolt the escooter.
So basically the original battery of the escooter was a 36v 222wh, which wasn't enough so i put a 36v 1200wh battery. I went for a ride to test it on some hills and it still didn't perform well . So i thought that the limiting factor wasn't the battery but the controller. Near the end of my ride i felt the motor had little power compared to the start of the ride. At the end of my ride, on the last incline, the motor suddenly stopped. When I tried pressing the throttle it would go, stop, go, stop and so forth until it came to a full stop. The motor would still spin with no load but with load it would just do the same go stop thing. It's been like that ever since. I thought that perhaps I had fried some component inside the controller that caused the problem. And so that brings us to the current situation. Perhaps it's the motor that has a problem. I might have broken something inside.

 

[DanGT86]

I would suggest inspecting the wires all the way to the motor. If you have a volt meter you can check the resistance between the phase wires to make sure they all match and are not shorted. You can also use the resistance setting on the meter to check the signal from the hall sensors to make sure you didn't damage a hall sensor.

If you are lucky its just a bad connection in your wiring somewhere and not something bad inside the motor.

 

[ibra2672]

Thx for the tip. I'll check.

 

[eee291]

Maybe it overheated the magnets, and that's why it's also spinning very fast.

 

[fechter]

Maybe it overheated the magnets, and that's why it's also spinning very fast.

I agree this sounds like a definite possibility. If the motor magnets get above a certain temperature (usually around 100C), the magnets will be permanently weakened. It's hard to test without having a new motor to compare with. One test is to disconnect the controller and short together the phase wires going to the motor. Then try spinning the motor by hand. With good magnets, it will be very hard to turn quickly. With weak magnets, it will turn with much less resistance.

 

[amberwolf]

Maybe it overheated the magnets, and that's why it's also spinning very fast.

then it would be doing this with the other controller now, too.

 

[eee291]

Maybe it overheated the magnets, and that's why it's also spinning very fast.

then it would be doing this with the other controller now, too.

The other controller might be rpm limited for street legality, I wouldn't write this off just yet.

The only way to increase a motor's KV without rewinding it is to weaken the magnetic field, there are two ways of achieving this.
It's either field weakening or cooked magnets, clocking the hall sensors does not change speed, but does increase amperage drawn.

 

[ibra2672]

So basically i ruined the magnets in the hub motor...Wow. Never thought that would be possible. It never performed that well, even with a big battery. Even my 8 year old e-bike hub motor (throttle only) 36v 240w nominal power outperformed this 36v 350w escooter hub motor. The escooter was relatively new too. I've had it for less than a month and I rarely used it. Although the escooter was the cheapest i could find (279€), so maybe the hub wasn't the best quality.
If the magnets are ruined, is it worth it trying to fix it or should I just buy another motor?

 

[eee291]

The magnets and epoxy are probably almost as expensive as a new hub motor, if you can find the right ones.
Maybe you can modify the scooter to fit a slightly larger motor?

Maybe something like this:
https://a.aliexpress.com/_mOFZ87U
https://a.aliexpress.com/_mtxYP0i

 

[ibra2672]

Update. By shorting phase wires there's residence. Does that mean the magnets are still ok?

 

[fechter]

If there is really strong resistance, then the magnets are still OK. It's hard to quantify just based on how it feels though. If you had a known good one to compare to, it would be a much better test.

 

[ibra2672]

If the magnets are ok, and the hall and phase wires connections are ok, than what could be the problem? Could it be something else inside the motor?

 

[fechter]

Really seems like a controller problem.
Is the new controller programmable?

 

[ibra2672]

My new controller has a self learn wire. Even if I mess up the phase and hall connection, the self learn adjusts everything. I don't think it's a controller related issue because with both controllers the motor doesn't spin from a standstill. It handles very small loads, like if you are already moving and give a bit of throttle. If you give too much throttle or if you're not on a decline, the motor just stops. My guess is that with the old controller, the stop-go thing was because an error kept popping up on the original display of the escooter. The error is "E1". I looked it up and it means that the motor isn't spinning. So with the original setup the escooter cut the power when it sensed that the motor wasn't spinning. Meanwhile on the new controller and display, it doesn't cut power when the hub isn't spinning. I can press the throttle all i want but the only thing heard is an unpleasant mechanical screech. Like something is spinning inside.
Could the problem be the hall sensor in the hub? How do you test the hall sensor with a multimeter? The cables are RED, BLACK, GREEN,YELLOW, BLUE.

 

[fechter]

Yes, to test the hall sensors, measure voltage from the black wire to the red wire and verify around 5v. Then measure from black to each of the yellow, green, blue while slowly turning the motor by hand. Each signal wire should toggle back and forth between near zero and near 5v as the motor turns. A bad usually just stays at one voltage.

The screeching sound could possibly mean all the magnets became unglued and are spinning against the back iron. Could also be a bad hall signal.

You may want to try opening the motor to take a look inside.