Motor buzzing/vibration

[darkzu]

I have a HX X8 scooter thats rebranded as a bunch others (Turboant, Levy, Cecotec etc), it wouldnt turn on anymore so I ordered a new controller/lcd unit combo, its now turning on and functions are correct but the motor is kind of buzzing and vibrating a lot and missing power at start when I stand on the scooter, its less noticeable in the air but you can still hear kind of a weird engine revving.

The controller is identical all wires/connectors are same, but the housing shape was slightly different because they changed the battery connector, so I had to transfer it in my old controller housing.

I took a picture from the old controller PCB because I was trying to fix it while waiting for the new one (ignore the 13V label), and I took a picture from the new one and noticed just now that the connector which goes to the motor has 1 pin labeled as 3V3 vs +5VA on old PCB, would this be the cause of this issue and how could I fix it, there is a unsoldered +5V pad, could I solder the 3V3 wire there?

Issue video: Watch VID_20240422_220852 | Streamable

 

[amberwolf]

If the motor has hall sensors, and they are not 3.3v parts, they probably can't correctly operate at that voltage, so doing what is needed to give them the required voltage would then fix it. Before moving the wire, I'd recommend checking what voltage is actually present there.

Note that many controllers have a diode between the actual 5v supply and the motor hall 5v output, to prevent feedback of motor voltages (induced or shorted) into the controller 5v, that can then damage everything else on that 5v line. If yours has this on any of the 5v outputs I would recommend using that output. (it's unlikely the 3.3v supply has any such diode, as I don't think there's any typical motor halls that would reliably work on the leftover voltage after the diode drop; some of them barely work on 4-5v as it is).


While you're checking things, is rest of the wiring identical from motor to controller on the new one, to the old one? Meaning, do the exact same wires from the same signals/pads on the PCB (regardless of color) go to the exact same pins on the connector that leads to the motor?

If not, then the phase/hall combination to the motor is probably wrong, and rewiring the new controller to match the old would fix that.

 

[darkzu]

I will disassemble later and check the voltages and come back, I have currently a front pic only of the old controller, Ill take a picture of the new one as well.

Do you mean the 3 phase wires? They are each connected separately with a single bullet like plug to the motor by color, cant confirm if they go out of the PCB in same order as old one, will check that too.

I am a bit confused about the old +5VA label would that mean its for analog components?
And also about the pads on the new PCB the empty +5V, maybe thats left so it could be bridged/soldered to get the 3V3 to 5V like the old PCB was, same as +15V in the same row is soldered to something.
There is also this "AD_HALL" pad label would be something like analog/digital hall?

Sorry if my questions are stupid, I am a amateur and trying to learn.

 

[amberwolf]

Do you mean the 3 phase wires? They are each connected separately with a single bullet like plug to the motor by color, cant confirm if they go out of the PCB in same order as old one, will check that too.

I mean *all* the wires, every single one of however many there are.

If there are only 3 wires from motor to controller, then there are no halls, no hall 5v, etc. to connect, so no need to worry about whether there is 3.3v or 5v, etc.

And as I said, color does not matter.

What matters is whether the same wire from the controller same signal source goes to the same pin on the connector to the motor, or if there are no pins then to the same wire on the motor.

If "pad x" on the controller goes to "motor blue wire" on the old controller, then that must be the same on the new one.

If any wiring is different, then the motor is probably being driven wrong, and you would need to either change the wiring to match the way it was on the old one, or you would need to manually find the right phase/hall combination, etc. (assuming that's what's wrong, which sounds likely). Or if the controller has a self-learn feature, use that.



But, as noted, that's all a secondary thing--first make sure what voltage your 3v3 pin really has. If it's 5v already, then you should verify the wiring as previously described.

If it's not 5v, you can try disconnecting the hall 5v wire from the 3v3 pad and move it to a 5v pad that is verified to supply 5v.

I am a bit confused about the old +5VA label would that mean its for analog components?

Voltage is voltage for this purpose, doesn't matter analog or digital parts being supplied with it***, so no reason to label it for that. It probably just means it's a "separate" supply from other 5v pads that are also labelled as +5V (so they may use 5v, 5va, 5vb, 5vc, etc).

***there are design considerations when originally designing power supplies for some digital or analog uses that make a difference to how they're used, but it's not something you'd need to worry about here.

And also about the pads on the new PCB the empty +5V, maybe thats left so it could be bridged/soldered to get the 3V3 to 5V like the old PCB was, same as +15V in the same row is soldered to something.

You would never want to short one voltage supply to another--it will damage both supplies and whatever lower voltage stuff is conected to the higher voltage by doing this, turning your controller into a useless box of junk.

There is also this "AD_HALL" pad label would be something like analog/digital hall?

No way of knowing what a manufacturer meant by any specific marking on a board unless they provide this in a manual with it (which pretty much never happens).

Most likely it is a signal input, not a voltage supply, but what it's for I couldn't tell you. If it is wired to something already, then whatever that is for may help you figure out what the label means.

 

[darkzu]

I mean *all* the wires, every single one of however many there are.

If there are only 3 wires from motor to controller, then there are no halls, no hall 5v, etc. to connect, so no need to worry about whether there is 3.3v or 5v, etc.

Oh no no, there is the mentioned 3 seperate phase wires and a 5 pin connector which goes to the motor and would be the halls thats the top left on the PCB which has the voltage label discrepancy, the bottom left 5 pin connector which goes to the lcd unit and theres a 2 pin connector for brake light, they cant be plugged wrong so theres no issue as it wouldnt work at all

If "pad x" on the controller goes to "motor blue wire" on the old controller, then that must be the same on the new one.

If any wiring is different, then the motor is probably being driven wrong, and you would need to either change the wiring to match the way it was on the old one, or you would need to manually find the right phase/hall combination, etc. (assuming that's what's wrong, which sounds likely). Or if the controller has a self-learn feature, use that.

How can I find out if "pad x" on the controller goes to "motor blue wire" they were just connected blue to blue wire, yellow to yellow, and green to green and thats how I plugged them to the new one, I would assume they wouldnt put mismatching colors

Voltage is voltage for this purpose, doesn't matter analog or digital parts being supplied with it***, so no reason to label it for that. It probably just means it's a "separate" supply from other 5v pads that are also labelled as +5V (so they may use 5v, 5va, 5vb, 5vc, etc).

***there are design considerations when originally designing power supplies for some digital or analog uses that make a difference to how they're used, but it's not something you'd need to worry about here.

Got it

You would never want to short one voltage supply to another--it will damage both supplies and whatever lower voltage stuff is conected to the higher voltage by doing this, turning your controller into a useless box of junk.

I didnt mean to short the two voltages supplies, but to connect the 5v unused pad to something (whatever is on that side I havent looked yet) which would make the 3v output as 5v instead like on old controller, I am thinking of it as a optional "switch" sort of if something like that is even possible

 

[amberwolf]

Well, first, as noted, just verify what's at the 3v3 pad, before you worry about anything else.

Other thoughts below:

Oh no no, there is the mentioned 3 seperate phase wires and a 5 pin connector which goes to the motor and would be the halls thats the top left on the PCB which has the voltage label discrepancy, the bottom left 5 pin connector which goes to the lcd unit and theres a 2 pin connector for brake light, they cant be plugged wrong so theres no issue as it wouldnt work at all

Doesn't matter if they can be plugged in wrong or not.

The order of the wires can be different.

If they are different, it doesn't work the same, and you can have the problem you see.

If you don't want to check the wire order, that's ok, but if it's the problem and has to be changed to correct it, you won't be able to fix your scooter.

How can I find out if "pad x" on the controller goes to "motor blue wire"

By checking and writing down these:

What pad on the old board is connected to the *motor* blue phase wire?
What pad on the old board is connected to the *motor* green phase wire?
What pad on the old board is connected to the *motor* yellow phase wire?
What pad on the old board is connected to the *motor*'s first hall signal wire?
What pad on the old board is connected to the *motor*'s second hall signal wire?
What pad on the old board is connected to the *motor*'s third hall signal wire?
What pad on the old board is connected to the *motor* hall power wire?
What pad on the old board is connected to the *motor* hall ground wire?
What pad(s) on the old board are connected to the *motor* (other wires, if any)?


Then check and write down these:

What pad on the new board is connected to the *motor* blue phase wire?
What pad on the new board is connected to the *motor* green phase wire?
What pad on the new board is connected to the *motor* yellow phase wire?
What pad on the new board is connected to the *motor*'s first hall signal wire?
What pad on the new board is connected to the *motor*'s second hall signal wire?
What pad on the new board is connected to the *motor*'s third hall signal wire?
What pad on the new board is connected to the *motor* hall power wire?
What pad on the new board is connected to the *motor* hall ground wire?
What pad(s) on the new board are connected to the *motor* (other wires, if any)?



If they are all the same pads to the same things on the motor, then there's nothing to rewire.

If they are not the same then rewire the new controller to match.

To check them, if they are all visible individual wires, you can trace them by eye.

If they end inside a connector you cannot see which wires go to which pins, then you will need to use your multimeter with it's ohms or continuity function (see it's manual for how to use it for that) to find out which pin goes to which wire/pad.

they were just connected blue to blue wire, yellow to yellow, and green to green and thats how I plugged them to the new one, I would assume they wouldnt put mismatching colors

Color

Does

Not

Matter.




You cannot assume that colors will match.



Sometimes you get lucky; colors *might* match, but they don't have to. I recommend looking up the many posts and threads about finding the right phase / hall wiring combo (combination), to see how this works on various systems. Many times you will find phases wired with, oh, say Y-Y G-B B-G from controller to motor, and the halls in a different pattern, say, Y-B B-G G-Y, to make the motor run smoothly at the right speed at the lowest no-load current.

I've even seen controllers that have all the *same* color wire for everything--one I have has green wires for every single hall and phase wire, including ground and power.

There are even systems that use totally different wire colors on one side of a connector/cable--I have motors that use red, white, and black for phases, but the controllers don't, and the halls may be purple, green, orange, with grey and yellow as power and ground. Etc.

I didnt mean to short the two voltages supplies, but to connect the 5v unused pad to something (whatever is on that side I havent looked yet) which would make the 3v output as 5v instead like on old controller, I am thinking of it as a optional "switch" sort of if something like that is even possible

Unless you cut and change things on the board to modify what is wired to the 3v3 pad, you can't make it output anything different than whatever it is wired to output.

I don't recommend doing that, as you risk breaking other board functions, or destroying parts, unnecessarily.

There isn't a need to switch anything either. You just find what you need to supply, where to get it from, and wire it up. If this was something you intend to experiment with many different motors that have halls with different power supplies (I don't know of any) then there would be a reason for a switch....


Instead, *if* it does not output the same thing your old board did, move the wire from the motor that goes to that pad to a pad that *does* output the same thing your old board did. It's that simple.

 

[darkzu]

You are right the colors dont mean anything, I guess I was just naive since its a 1:1 replacement controller for the same scooter but a newer revision I thought that it would have the same outputs/colors so that people can just plug it in regardless if they had the older revision scooter.

Lets say the phase wire colors arent representing the same as on old one, if I tried Y-Y G-B B-G or other combination/s would that damage anything?
Assuming they are right now plugged wrong and producing this issue does it do any damage to the motor etc?

Anyways I have checked each pad to wire old vs new and they are the same if I understood that correctly (scratching my head), ill plug the battery and check the 3v3 and +5 label voltages next.

Its kind of a pain in the ass to see much from front because of that huge cap and the glue

EDIT: the labeled 3V3 which goes to motor halls measured 5V so I guess thats fine, and it leaves the motor/hall sensors as the cause? Its going to be fun opening it with those allen screws stripping easily.

 

[amberwolf]

If your'e getting 5v from the 3v3 pad, then that's what most hall sensors run from so that should be normal, and is not causing the symptoms.

If all the wires on the new controller match those on the old, for where they go on the motor, so that for instance Hall A on both go to the same motor hall wire, and Phase U on both go to the same motor phase wire, etc., then it's not a wire order problem either.

That means that unless each controller's pads are not driven by the brain (MCU) with the same signals (so that one reads Hall A from the hall A pad, but the ohter one actually reads Hall A from teh Hall B or C pad, or drives Phase U from the Phase V part of the drive system, etc). the phase hall combination should be the same for both, so you shoudl not need to experiment with wire combinations...but....

Looks from the images like these were verified by measuring the actual wire resistance/continuity, so you know the controller wires are good up to the motor connector / wire ends, then it's not a cable/connector issue.

So that leaves either a problem with the motor itself, or a difference in the way each controller operates.

If it's a motor issue, it would probably be a hall sensor problem. That you can test for; there are a number of ways; I'd start with ebikes.ca page for learn-troubleshoot testing hall sensors. The motor doesn't have to be opened to test.

https://ebikes.ca/documents/HallSensorTestingFinal.pdf

If it's a difference in controller operation, it might not be an actual "problem", just that they work differently and don't produce the same results. Unless the new one is user-programmable, or has a display with user-alterable settings, there's not much you can do about that. For instance, if the new one has a different startup method, or if it doesn't use the hall sensors (runs sensorless) even though they're hooked up (I have one like that), etc., it might run poorly for startup under load (but work fine offground), etc.

But if it's a difference in the way the new one is "internally wired" (either in software or from the MCU to the pads on the board) so the signals to/from the motor just don't match up like they did on the old one, then you can do the phase/hall combination testing. There are safe ways to do it, listed in various posts and threads; I can look up and link you to one if you don't find one.

If this *is* the problem, then the system is probably using more current than it should, which will heat up the motor and controller more, and reduce your range. Does the wheel spin faster offground than it used to? (would probably be a lot faster, enough to notice) Does the motor or controller get hotter now than they used to? Or is range noticeably less (or whatever watts or amps display the system has, if any, read higher than before)?

If none of those is happening, and all the motor halls test good, then most likely there's no real "problem" to fix, and it just is going to run differently with this controller than it did with the other one.

 

[darkzu]

I will test the hall sensors next with the method you linked.

The controller has limited 5 "P settings" through the LCD/throttle buttons, kph/mph, cruise control, kick start, wheel size, and max speed 20/25/32kph, nothing really useful in this case.

Are you referring to this post? How to Determine the Wiring for a Brushless Motor
This will be a bit tough to test since I dont have a power supply other than the battery, and the battery connects directly to the controller with a weird 7 pin D-sub connector so not sure how would I put a fuse or measure battery current.
But first hall sensors test.

The wheel spins about the same as before maybe even slower at the 2 slower speed modes offground, on ground with load it sounds like a gas scooter and the slower modes are really slow and only the third fastest mode was able to push when I drove very short distances in yard and nothing really got hot there, display has just 5 bars for battery which just measures battery voltage I guess.

 

[darkzu]

I did the hall sensor test as described, scooter turned on, black and red hall wires have 5V, then probing each hall wire and moving the wheel a bit backward or forward they flip and stand either at 2.9V or at 0V on each sensor, this would mean they are bad and need to be replaced right?

Video: Watch VID_20240424_202045 | Streamable

 

[amberwolf]

As long as the voltages are toggling back and forth on every sensor as you manually rotate the wheel, then the sensors are working, turning on and grounding the hall signal.

The actual voltage you read for the off state (not grounded) is provided by the controller internally with pullup resistors to a voltage source--usually that's 5v, but yours probably uses a 3.3v MCU with 3.3v inputs, and isn't buffering the voltage to those inputs, so it uses a 3.3v pullup (that drops to the voltage you see at the connector).

It's still possible that it is the reason there's an issue, since there can be induced noise on the signal lines from the phase currents that are in wires right next to the halls, and the much lower voltage pullups means that the signal to noise ratio is lower, so it's easier for that induced noise to cause problems with the controller reading the actual signals in there...which can cause timing faults sending the phase currents to the motor, so it doesn't drive properly.

That's not really something you can do a lot about, though.

 

[amberwolf]

Are you referring to this post? How to Determine the Wiring for a Brushless Motor
This will be a bit tough to test since I dont have a power supply other than the battery, and the battery connects directly to the controller with a weird 7 pin D-sub connector so not sure how would I put a fuse or measure battery current.

That makes it tougher to test, but still possible. You would have to unsolder or cut one of the battery-to-controller wires from that D-sub connector and insert your meter between the battery and controller sides of that wire. It doesn't matter if it's the red or the black wire, but you do have to check your meter's instructions on how to measure current in A with it (not the mA range), as it connects the meter leads differently than V or ohms and you can blow up the meter (or it's internal fuse if it has one) if you connect it wrong.


Is still worth doing this test to be sure, but if I had to guess, with the available info at this point, it's more likely that the low pullup voltage on the halls or just a different controller design/functionality is causing the issue.

 

[darkzu]

Well guess I am sol then and can try sell the scooter for parts.
Unless I try to fix the old controller, is there any guides here I can follow to do some testing on it.

Basically here is what happened, the battery is removable and in the stem with not the greatest holding mechanism and not the greatest connector (7W2), so from the vibrations and bumps the scooter started to randomly shut off here and there and after one ride I noticed that one pin was sticking out on the controller side and was hot, on battery side the plastic around that pin was melted a bit and not the cleanest.
So I tried to push the pin down on the controller and accidentaly shorted/discharged it, I didnt know at that time that there was inside a huge cap and voltage, it didnt turn on anymore at that point. I took it out unscrewed the connector and the negative was unsoldered, I removed both connectors completely and connected temporary with wago clamps and still was not turning on.

Visually nothing stands out to have burned
Tested the terminals for short - nothing
Tested the mosfets for short - nothing
I measured the 3 capacitors in circuit and they are around the rating - not verified
I measured the pad labeled 15V of the connector which goes to the lcd unit, when trying to turn it on it was 13V.

 

[amberwolf]

Since you get 13v at the 15v, it means the controler is actually turning on, in that it's power supply (LVPS) is getting battery voltage and creating that voltage, which is used to run other things in the controller.

It's unusual for the controller display to run off anything other than battery voltage, so this is a custom-designed system rather than a typical one, for it to run on 15v. The display turns on with the new controller? Does it turn on with the old one?

What does the 5v of the old controller read? If it's zero, then the problem is probably only in the LVPS.

What did the battery input pin on the controller short to when pushed down? That would give a lead to where to start checking.

 

[darkzu]

I havent tried the old display with new controller or vice versa, I bought the new controller + new display together, since they have to match I guess because they may be differently programmed with different batches/revisions., so I am bit afraid to even try the new controller + old display or vice versa to avoid damaging more stuff.

It was 2 weeks ago so dont know exactly if I measured that but I think it was 5v if I did, I have to solder a new connector to that pcb and take out the new stuff again for that.

Hard to say to what it shorted, it was the negative pin i was pushing, the 5 middle pins had 2 pins shorted (purple) which is for activating the battery otherwise the battery doesnt output voltage, there was 3 other wires soldered which led to the purple circle on pcb I guess it was used for programming it. It could very well also shorted to the positive pin as I was using bent nose pliers.

 

[darkzu]

Okay so today finally the weather was bit better, so i did a 6km test ride, at max speed it was almost ok and not buzzing much, its mostly at start and accelerating up to max speed, but then halfway it suddenly became normal, no buzz/no vibration, completely smooth from start to max speed, but it was like that for only about 200m and then the issue was back again till home, there was maybe one more very short normal operation.
Nothing was really hot, just very slightly warm motor which is normal. I am so confused what could be causing this.

Buzz/vibraiton: Watch VID_20240425_222548 | Streamable
Smooth ride: Watch normal for a a moment | Streamable

 

[AussieRider]

G'day darkzu.
You said the housings are slightly different. Could be that's not the only difference. Maybe the old controller was a sine wave (quieter) & the new one is square wave (noisier).

AussieRider

 

[darkzu]

Hello
Correct, the housings were very slightly different on one side but only because they ship the scooters now with a 3 pin blade connector to battery instead of the D-sub 7W2 which caused problems, I would agree to possible different wave but how do you explain that it suddenly run smooth "sine wave" for about 200m and then again was noisier "square wave" on the same new controller?

 

[AussieRider]

Harmonics.

 

[darkzu]

Harmonics.

I dont know whats that supposed to mean and how does it help me.

Anyways did a 20km ride and it has smooth/rough periods I guess I will ride it like that until something breaks.
Thanks @amberwolf for your time and the help/info.

 

[E-HP]

Do you have a pic of the new controller label? Maybe your hall sensors are fried or have a poor connection, and the controller is dual mode and switching to sensorless.

 

[darkzu]

Yes, I did also took apart the motor after the hall sensor test in post #10 above to take a look inside and there was nothing suspicious/loose and I checked each wire for continuity with no problems. The controller has a Goodix GMF03x MCU as far as I could see.

Apparently it has some fault checks and error codes, one being "Motor Hall fault or connection failure", I dont get any errors.

HX Electric Scooter TroubleShooting - HX Electric Scooter Online Store

How to fix X7 X8 PRO electric scooter. X7 electric scooter troubleshooting, tips & guide on common X8 X9 scooter motor or battery fault

hxescooter.com