Broken ebike..how to fix?

Hello,
I just assembled my first DIY ebike, took it for a ride, and everything was great until the 3 phase hub motor spun out of the bike's dropouts. Now the bike is dead.
All the wires running into the motor look OK. None of them are torn from what I can see. ( I also tested the phase wires. This ebike doesn't need hall wires to be connected, so I didn't test them).
Unfortunately when I tested the hall wires (black and red) running out of the controller, a reading of 0V was returned. Strange..but then this bike doesn't need hall wires anyway.
My LCD3 display shows that the battery is totally drained, even though it's fully charged..
Any help is much appreciated!
Cheers

Sometimes you can short/destroy the controller when you wrap the phase wires around the axle and they touch each other. The tear may be inside the axle. The short may have been momentary, or it could still be in a short. You'd have to open the hub and pull the phase wires a bit for an examination.

Sounds like you didn't install a torque arm to prevent spinout. It's amazing that companies are still selling kits without them.

neptronix said:

Sometimes you can short/destroy the controller when you wrap the phase wires around the axle and they touch each other. The tear may be inside the axle. The short may have been momentary, or it could still be in a short. You'd have to open the hub and pull the phase wires a bit for an examination.

Sounds like you didn't install a torque arm to prevent spinout. It's amazing that companies are still selling kits without them.

Click to expand...

Thanks for your reply
I've opened up the hub motor and all the wires seem fine..For there to be a short, there would have to be two wires touching? I can't see any evidence of this.

Is it possible that simply trying to accelerate while the motor was jammed would fry something in the controller?

I should add that the power supply to the throttle is half a volt, whereas it's apparently supposed to be 5v

Well that sucks.

Telling us what brand of motor, along with pics of how you have it set up might help a bit.
You said your controller doesn't need the halls, but were they hooked up when the motor spun?

A common mode of failure is for there to be a short between the Phase wire and one of the Hall wires. That dumps full battery power back into the 5 volt side of the controller, and lets all the magic smoke out.Other ways this can happen, 1 or more of the Phase can short to the hub body and snuff out the controller that way, or short to each other and send your controller to e-Valhala. Often the shorts won't appear when you look for them. Under stress, the wires are pulled against a sharp edge that cuts the insulation just enough to let metal touch other metal. When that stress is released by unwinding the mess of wires, the hole in the insulation closes up and can't always be found with the naked eye.

it's unlikely that controller blew from the motor being jammed. the Current limiter in the controller would prevent that unless it got so hot it melted. that would take a fair amount of time.

The good news is there are cheap controllers out there. While I'm not sure which kit you have, the controller is often the least expensive of the 3 major components of an ebike, starting around $40 for a sensorless controller.

I'd bet that your 5 volt power supply on the controller blew up to some degree. The hall sensor lines use this 5v power supply rail and i'd bet the hall lines are shorted inside.

Of course this could have only happened if you had the halls plugged in.

neptronix said:

Of course this could have only happened if you had the halls plugged in.

Click to expand...

Without knowing anything further about the controller or its design it's certainly conceivable that he blew the 5v without the halls plugged in while shorting the phase leads. The craziest shit happens with electronics.

Drunkskunk said:

Telling us what brand of motor, along with pics of how you have it set up might help a bit.
You said your controller doesn't need the halls, but were they hooked up when the motor spun?

Click to expand...

I'm not sure of the brand, but here's a link to where I bought the kit:
https://www.alibaba.com/product-detail/High-Quality-48V-1500W-Electric-Bicycle_62035606251.html?spm=a2700.icbuShop.82.9.7beb3749MBBK1C

And yes, the Hall sensors were indeed plugged in. I guess they shorted the 5V power supply which is why the throttle doesn't have power and I can't get the bike to move.

One thing though..wouldn't the PAS be able to get the bike to move?

PAS probably uses the 5v power supply. A lot of things do. The battery indicator might also run off the 5v power supply.

neptronix said:

PAS probably uses the 5v power supply. A lot of things do. The battery indicator might also run off the 5v power supply.

Click to expand...

That makes a lot of sense!

A new controller it is..

Hopefully the Alibaba supplier sends it for free. Not including torque arms with a 1500W motor, or at least saying they're a must..pure madness!

marc123 said:

And yes, the Hall sensors were indeed plugged in. I guess they shorted the 5V power supply which is why the throttle doesn't have power and I can't get the bike to move.

One thing though..wouldn't the PAS be able to get the bike to move?

Click to expand...

Its got about a chance of a snowflake in hell of working, but that's still a chance. If you have the parts, it can't hurt to try.
That 5v powers the CPU chip, the throttle, the halls, But also the LCD, which you said worked enough to tell you the battery was dead. it also powers the PAS. A slim chance of it working is better than none. You can't break it any more than it is, at least.

There exists a test set for helping to diagnose this sort of problem. Ebike tester
You may not need it, but knowing how to test and troubleshoot your ebike is a good skill to have.
there are a bunch of youtube vids on how to test with them.
[youtube]Rh2nCfksh3U[/youtube]


Also, Torque arms. Never ride without protection

Also i'd recommend torque arms from grin technologies ( ebikes.ca ).
I tried the cheap ones. They're all loosey goosey and typically have 1mm of slop.

...hopefully the rear dropouts survived..

neptronix said:

Also i'd recommend torque arms from grin technologies ( ebikes.ca ).
I tried the cheap ones. They're all loosey goosey and typically have 1mm of slop.

...hopefully the rear dropouts survived..

Click to expand...

One of the dropout is quite bent but it should be fixable, or simply not an issue with good torque arms.

Drunkskunk said:

Its got about a chance of a snowflake in hell of working, but that's still a chance. If you have the parts, it can't hurt to try.

Click to expand...

I tried..it failed

It's odd that the 5v powers the LCD, and yet it actually does still work. Maybe it's one of the few parts that can operated under a reduced voltage. The throttle and PAS do have power being supplied, but nowhere near 5V.

I'm not sure if this means anything, but as soon as I turn on the ebike, there's a small grey resistor on the controller that gets VERY hot.

marc123 said:

Hopefully the Alibaba supplier sends it for free. Not including torque arms with a 1500W motor, or at least saying they're a must..pure madness!

Click to expand...

Yeah - the trick is to come here first. I'd have probably done the same thing if I hadn't.

marc123 said:

All the wires running into the motor look OK. None of them are torn from what I can see.

Click to expand...

Almost certainly the insulation is damaged on them, even if you can't see it, and when they are twisted together pressing against each other, then they short together.

Sometimes this is inside the axle, where you can't see it at all wihtout opening up the motor and pulling the cable inside the motor until the entire twisted section is inside.

That's also the simplest way to repair the cable--pull it in, then cut the damaged section out, and splice the good sections back together. Usually lots easier than cutting it at the axle then trying to force the cable back into it "from scratch".

Unfortunately when I tested the hall wires (black and red) running out of the controller, a reading of 0V was returned.

Click to expand...

That can mean that the 5v power supply inside the controller was destroyed.

Unfortunately, the most likely way for it to be destroyed in an axle spinout is for a phase wire to short to it, which then feeds essentially battery voltage to the 5v line, meaning anything that is powered by it (throttle, motor halls, PAS sensor, controller MCU, 5v regulator, etc) tends to be destroyed.

If that's what happened, then most likely you'll end up needing a new controller.

The LCD has it's own internal power supply running from the battery power supplied by the controller, so it probably wasnt' damaged. It doesn't display speed/etc anymore, because all that info comes from the controller's MCU, so if the controller isn't working it can't tell the LCD what to show. It is possible that it *is* damaged, if the phase power thru the 5v line cooked the MCU and hten also made it's way to the serial data lines to the LCD. But probably not.

I'd recommend a pair of good torque arms from http://ebikes.ca as well.

Moot point, but check your LCD3 wire diagrams. There's no 5 volts in there. It runs off the battery line, generating onboard power from a DC-DC converter.

Edit: Amberwolf just told you the same, so I agree.

Years ago when I took a golf lesson from a retired PGA professional, one of his comments during the lesson was "Sometimes you got to look at the ball, I mean really look at the ball." That was a lesson for a highly skilled golfer with a lot of experience, not for a novice.

What you are getting here is comments from the best in eBikedom. Aka the pros......However, I am not seeing the basics of diagnostics yet.

Ok, lets start from the basics.

Is your battery fully charged? What is your battery voltage?

With your hall sensor connector disconnected and your controller plugged to the battery, do you get a voltage reading from your controller to the Halls? What is it?

Have you gone through this to get a reliable notion on your halls functionality? https://www.ebikes.ca/documents/HallSensorTestingFinal.pdf

Does your motored wheel spin freely with everything disconnected?

Does it spin freely with your phase wires connected?

Answer these questions first before doing anything else.

e-beach said:

Is your battery fully charged? What is your battery voltage?

With your hall sensor connector disconnected and your controller plugged to the battery, do you get a voltage reading from your controller to the Halls? What is it?

Have you gone through this to get a reliable notion on your halls functionality? https://www.ebikes.ca/documents/HallSensorTestingFinal.pdf

Does your motored wheel spin freely with everything disconnected?

Does it spin freely with your phase wires connected?

Click to expand...

The battery reads 54V, so fully charged.

The halls are getting 0V from the controller, and the throttle is getting around half a volt. The PAS is getting just over a volt.

I'll check if the wheel spins freely with/without the phase wires once I'm back from work.

Try unplugging the motor halls and measure the 5v again. Repeat with throttle and PAS. One of those could be shorted and pulling the 5v down. If you don't get 5v with all of those unplugged, then the controller is bad for sure.

marc123 said:

The halls are getting 0V from the controller,

Click to expand...

Where are you metering on your Hall sensor connector? Got a picture you can show us?
Is your battery plugged in and the system turned on while testing the Hall connector voltage?
Did you attempt the bad hall test I linked to you?
If you are metering the wrong wires and the hall you are testing is latched to ground, you will get a 0 voltage reading.

and the throttle is getting around half a volt.

Click to expand...

Where are you metering on your throttle connector? Got a picture you can show us?
Is your battery plugged in and the system turned on while testing the throttle voltage?
What color wires are you metering to?
If you are metering the ground and return wire without applying the throttle you will get a reading of just under 1v.

Either describe to us where on the connectors you are metering or give us a picture so we can be sure you are doing what we think you are doing. Detail is better then no detail.

Have you opened your controller to see if some of the wires have been pulled from the PC board?

e-beach said:

Where are you metering on your Hall sensor connector? Got a picture you can show us?
Is your battery plugged in and the system turned on while testing the Hall connector voltage?
Did you attempt the bad hall test I linked to you?
If you are metering the wrong wires and the hall you are testing is latched to ground, you will get a 0 voltage reading.

Click to expand...

1) Yes everything is on.
2) The "bad hall test"?
3)As per the PDF you linked me, I'm metering the hall between the black and red wires of the hall 5 wire bundle exiting the controller. This is where I'm getting a 0v reading. That said, between either the black or red wires and each of the three other wires, I do get readings (6 different ones depending on the wire combo), all between 1.5-2.25v. Odd..
This is with the hall bundle exiting the motor connected and disconnected (same results either way)
4) The voltage did not oscillate between 0v and 5v like it was supposed to as per the test.

e-beach said:

Where are you metering on your throttle connector? Got a picture you can show us?
Is your battery plugged in and the system turned on while testing the throttle voltage?
What color wires are you metering to?
If you are metering the ground and return wire without applying the throttle you will get a reading of just under 1v.

Either describe to us where on the connectors you are metering or give us a picture so we can be sure you are doing what we think you are doing. Detail is better then no detail.

Have you opened your controller to see if some of the wires have been pulled from the PC board?

Click to expand...

1) I tested all possible combinations of the three wires exiting the controller and going to the throttle. Red to black: 0v, Red to blue: .46v, and Black to blue: .49v. This is with the throttle attached and the bike on. The position of the throttle didn't change the readings.
2) I did the same for the PAS; I tested all possible combinations of the three wires exiting the controller and going to the PAS. Red to black: 1.69v, Red to yellow: .66v and finally Black to yellow: 1v
3. I did open the controller, and everything is alright. Nothing torn or out of place. I can even see the 78M05HP 5V regulator, but I can't tell if it's shorted.

e-beach said:

Does it spin freely with your phase wires connected?

Click to expand...

Yes, the wheel spines freely although with a bit of resistance, both with or without the phase wires connected.

Well, if you can see the 7805, why not check the output. Left side is Vin. Ctr is Gnd. Right side is %V.

marc123 said:

Yes, the wheel spines freely although with a bit of resistance, both with or without the phase wires connected.

Click to expand...

Sorry for your troubles.

Ok, as others have stated, it does indeed sound like the controller is bad. The question is why did it go bad? If no wires were pulled out by the indecent, then the only other option I can see is a short in your motor cable. As others have also stated. Which means if you purchase a new controller and have shorted wires, you could fry your new controller as well.

Here is some tips on testing the motor cable.

The phase wires:
Check the three phase wires with your digital multi meter on the diode setting.


On the controller side of the motor cable check your connectors with the ohm meter. When you test Blue to Green, Blue to Yellow, Yellow to Green (or what ever colors you have) you should get an open reading --- No connection between any of them.

Check the phase wires against the axle. Connect one probe to the axle, make sure you have a really good connection, and probe one of your phase wires to see if you have continuity. If you do, you have a phase grounded internally. If no continuity, that is good. Test the other phase wires in this way.

Checking the Hall Sensors with a power source:

Do you have a 5v power source? Connect the 5v supply to your Hall sensor positive and negative. Make sure you do not short anything!!! Then go through the process from the pdf I linked on testing Hall sensors. (this one: https://www.ebikes.ca/documents/HallSensorTestingFinal.pdf)

In reality most halls can take more then 5v, but since we don't know your type of Halls, we should stay as close to 5v as we can. Do you have a 5v phone charging cable that you can sacrifice for the job? Or, a depleted 9v battery could work. If the 9v battery is under 8v you should be safe. However it need to be more then 4.5v or the Halls may not work.

If anybody else has other testing ideas or better ways to test please speak up :wink:

Any marc123, do these tests and let us know the results.