Ebike issues - strange and unseen

Get all your technical information about electric bikes here.
Navarro   10 mW

10 mW
Posts: 31
Joined: May 12 2018 1:55am

Ebike issues - strange and unseen

Post by Navarro » Dec 08 2018 4:27am

Hello everyone.
The background: I have an old Pegas fat bike converted with a typical chinese 48V1000w hub motor kit. Everything went really well for 2 years and about 800km.
After that, I had an issue with the kit: when accelerating with the throttle, the motor would jerk and give an error (googled it - hall sensor issue). So I replaced the Hall sensors with corresponding ones.
Now, when I accelerate, the motor moves 3-4 cm and stops. The display shows me no error, only continuous operation (just like normal behaviour).
I measured the hall sensors and they put out 0-4.9 volts, they remain at 5Volts when the movement stops. Also, the transistors do not seem to be blown, they have 3.99KOhms resistance.
Any help or ideas?
Last edited by Navarro on Dec 08 2018 9:17am, edited 1 time in total.

larsb   100 kW

100 kW
Posts: 1109
Joined: Dec 10 2014 5:12am
Location: Gothenburg, Sweden

Re: Ebike issues - strange and unseen

Post by larsb » Dec 08 2018 4:51am

If all three halls switch ok then it might be the hall and/or phase combination is not correct anymore. Is it possible you changed the connections?
measured the hall sensors and they put out 0-4.9 volts, they remain at 5Volts when the movement stops
This is not normal, they have to switch states when you turn the motor, do they?

Could be:
Change in wiring connections
Flipped sensor

Get yourself a hall tester, it's worth it :D
https://m.ebay.co.uk/itm/40W-24V-36V-48 ... noapp=true

Need more info:
What motor,
what controller,
which sensors did you switch to,
how many did you replace,
how did you know they were faulty
Ride on :D

Navarro   10 mW

10 mW
Posts: 31
Joined: May 12 2018 1:55am

Re: Ebike issues - strange and unseen

Post by Navarro » Dec 08 2018 6:34am

Hi larsb and thanks for your quick reply.
The kit is the most common on ebay:

https://www.ebay.co.uk/itm/Electric-Bik ... qlhVckEByA

The sensors switch voltage from 0-4.9 volts when I. move the motor(even 2.8V but I understood this is quite normal). I replaced all 3 sensors, did not measured them before replacement. Basically I googled my error code and based on it and the jerking movement of the motor when throttling I guessed they might be to blame.

larsb   100 kW

100 kW
Posts: 1109
Joined: Dec 10 2014 5:12am
Location: Gothenburg, Sweden

Re: Ebike issues - strange and unseen

Post by larsb » Dec 08 2018 7:21am

can you program this kit or readout hall signals? If not then i think you should get the tester.

Are you sure all motor to controller phase connections are the same as before?
Are you sure all hall sensor connections are the same as before?
Which sensors did you use and which sensors did you replace (name of sensors)?
Ride on :D

Navarro   10 mW

10 mW
Posts: 31
Joined: May 12 2018 1:55am

Re: Ebike issues - strange and unseen

Post by Navarro » Dec 08 2018 9:14am

]Hi again,
The sensors were the same type like the one inside - I made sure of that. I can open again the motor and see their type and make a snapshot of them. I soldered them exactly like the ones I extracted . I don't know how to reprogram the controller. The phase wires and the Hall sensor wires are as before.
The strange thing is that when I throttle to accelerate, the display shows normal behaviour, the power goes up, the rightside shows "wheel turning" - but the wheel stops after 5cm.
My display:
IMG_20181208_160739.jpg
My motor:
IMG_20181208_160725.jpg
My controller:
IMG_20181208_160711.jpg

larsb   100 kW

100 kW
Posts: 1109
Joined: Dec 10 2014 5:12am
Location: Gothenburg, Sweden

Re: Ebike issues - strange and unseen

Post by larsb » Dec 08 2018 3:58pm

Could you post a short video of the behaviour when you throttle?
Is it the same no matter what the starting position of the wheel is or is there a variation if you try a few times?

i wonder if controller could be busted, hall sensors might not be the issue.

Did it behave the same before you changed the sensors?
Ride on :D

User avatar
amberwolf   100 GW

100 GW
Posts: 27250
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Ebike issues - strange and unseen

Post by amberwolf » Dec 08 2018 11:57pm

If the hall sensors themselves operate correctly, when measuring with controller plugged into motor, then it's probably not a connection or hall problem.

But it could be a phase problem, either connection between motor and controller, or failed FETs inside the controller.

It could also be a mismatch between the order of hall wires (which it doesn't look like you can change around) and the order of phase wires (which looks easy to undo and redo). If the phase wires were disconnected during any of the work done on it, it's possible they may have been reconnected in a different order than before.

If the controller has a self-learn feature, you can engage that (usually plugging some single-wire connector pair together temporarily) and it'll fix it by itself.

If not, you can try the few possbile combinations of the order of the phase wires. At least one of them will run it correctly forward, and at least one will run it backwards.

Navarro   10 mW

10 mW
Posts: 31
Joined: May 12 2018 1:55am

Re: Ebike issues - strange and unseen

Post by Navarro » Dec 09 2018 9:57am

I just made a video of the behaviour, tell me if you guys can understand what is going on - basically that is what happens when I accelerate.
Later, I will move around the phase wires, maybe I'm lucky and it's only this.
https://youtu.be/1l8_l-kozC0

larsb   100 kW

100 kW
Posts: 1109
Joined: Dec 10 2014 5:12am
Location: Gothenburg, Sweden

Re: Ebike issues - strange and unseen

Post by larsb » Dec 09 2018 12:36pm

That's not looking like hall issues. Could you post video with wheel lifted and just slowly increase throttle from zero to max?

Keep phases, halls and throttle connected but If you have any other connections like PAS and brake switch then try and disconnect them before.

What's battery voltage vs undervoltage setting?
Ride on :D

Navarro   10 mW

10 mW
Posts: 31
Joined: May 12 2018 1:55am

Re: Ebike issues - strange and unseen

Post by Navarro » Dec 09 2018 2:38pm

Battery is full, 52Volts I believe.
Tried all combinations of phases - either wheel moves 1cm in one direction then 1cm in the other direction, either just stops there.
I made a small video - throttling from 0 to max and 03-Info popped up.

https://youtu.be/psgXQgtNvPY

After replacing the Hall sensors I checked the wires - they are ok, nothing shot between the sensors and the plug. However... Still there is this.

larsb   100 kW

100 kW
Posts: 1109
Joined: Dec 10 2014 5:12am
Location: Gothenburg, Sweden

Re: Ebike issues - strange and unseen

Post by larsb » Dec 09 2018 3:43pm

OK, so you get hall error again if i read my old manual for the LCD-3..
S-LCD3-User-Manual.pdf
(1.52 MiB) Downloaded 22 times
I'd get the hall tester for christmas, seriously. You can do the same measurements with a DMM for sure but it get's so much easier and you get the full picture from the all the halls at the same time, you can test the throttle etc.

This is what a normal 120 degree commutation logic looks like. If the halls states don't follow the table on the right either top to down or down to top then the hall signals are not correct.
hall commutation scheme.JPG
hall commutation scheme.JPG (28.32 KiB) Viewed 1236 times
Just check it and write what combinations you get when you slowly rotate the wheel forward. Write down (and post here :D ) each hall ON as a 1 and a hall off as a 0. there will normally be the same six states that are recurring when you do this.
Ride on :D

Navarro   10 mW

10 mW
Posts: 31
Joined: May 12 2018 1:55am

Re: Ebike issues - strange and unseen

Post by Navarro » Dec 09 2018 4:03pm

I will get the above tester and post the results.
I was thinking in buying a sensorless controller - but this means rendering the LCD display useless, is this true? I did not saw any display plug to this controller:
48V-84V 1000W 45Amax BLDC motor controller:
https://s.click.aliexpress.com/e/mLbRIpS

User avatar
amberwolf   100 GW

100 GW
Posts: 27250
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Ebike issues - strange and unseen

Post by amberwolf » Dec 09 2018 9:30pm

If the controller is the problem (something in it's hall sensor reading ability), and if it is a KT Kunteng Kun Teng controller, you may be able to use the open-source firmware that Casainho / Stancecoke et.al are creating. They also have some for the LCDs for those controller. Just look up Casainho's posts and you'll find the threads.

larsb   100 kW

100 kW
Posts: 1109
Joined: Dec 10 2014 5:12am
Location: Gothenburg, Sweden

Re: Ebike issues - strange and unseen

Post by larsb » Dec 10 2018 1:38am

The LCD is definitely from KT, controller might be this one?

http://www.szktdz.com/en/news_show.php?article_id=503
Ride on :D

Navarro   10 mW

10 mW
Posts: 31
Joined: May 12 2018 1:55am

Re: Ebike issues - strange and unseen

Post by Navarro » Dec 10 2018 10:11am

Controller looks very similar with mine - mine is a bit shorter. I can count the number of transistors if that helps.

User avatar
amberwolf   100 GW

100 GW
Posts: 27250
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Ebike issues - strange and unseen

Post by amberwolf » Dec 10 2018 9:29pm

To use the firmware, the thing that is important is the manufacturer of the controller and the MCU/etc they used. To find out if yours is compatible, look up Casainho's posts to find the threads about the new firmware.

Navarro   10 mW

10 mW
Posts: 31
Joined: May 12 2018 1:55am

Re: Ebike issues - strange and unseen

Post by Navarro » Jan 10 2019 6:51pm

Hi guys,
I bought the tester and when testing the controller (so hall sensor connected output and all 3phases clipped to the controller) I get the below thing. Basically the lights flash as can be seen - then they stop, even if the throttle is opened fully.
Confirmation that I need a new controller?

https://youtu.be/UnmEngg7pqM

larsb   100 kW

100 kW
Posts: 1109
Joined: Dec 10 2014 5:12am
Location: Gothenburg, Sweden

Re: Ebike issues - strange and unseen

Post by larsb » Jan 11 2019 1:52am

I never tried to use the controller connections on the tester but I don't think that's normal. Too many lights on at the same time, should be only two revolving LEDs and the middle.
UsingTheLyenEbikeTester.pdf
(178.55 KiB) Downloaded 17 times
Though controller responds to halls, if these are bad then the rotation would be shown incorrect. But i don't think so, controller is busted.

Just for certainty:
The hall states doesn't show on the lights on the left side, do they do this if you disconnect controller phases and turn the motor?

Do they follow the table?
Ride on :D

Navarro   10 mW

10 mW
Posts: 31
Joined: May 12 2018 1:55am

Re: Ebike issues - strange and unseen

Post by Navarro » Jan 11 2019 3:41am

Hi larsb,
I tested the halls in the motor, they flicker normally. If you move slow, they flicker in the correct order, slowly. Fast - same, but faster.
I will post all the tests later today, I am at work now. Nonethenless, many thanks, I think when spring will come I will be able to ride my ebike again.

Navarro   10 mW

10 mW
Posts: 31
Joined: May 12 2018 1:55am

Re: Ebike issues - strange and unseen

Post by Navarro » Feb 22 2019 12:43pm

Hi everyone!
Buyed another controller and all is not ok.
So when I throttle from 0 the motor starts-shudders-starts again and goes ok. What can be the issue of this behaviour- a shuddered start.
Maybe wrong phase allignment?

So basically a tester is gold if you are a self-doer.

larsb   100 kW

100 kW
Posts: 1109
Joined: Dec 10 2014 5:12am
Location: Gothenburg, Sweden

Re: Ebike issues - strange and unseen

Post by larsb » Feb 24 2019 4:19am

Sounds like not perfect alignment, if you can get it to run when it's up to some speed.

Basicly each current injection timing is off a bit and doesn't give enough push and motion to securely sync the next.

A video would probably show it clearly, it's a guess.
Ride on :D

Navarro   10 mW

10 mW
Posts: 31
Joined: May 12 2018 1:55am

Re: Ebike issues - strange and unseen

Post by Navarro » Feb 25 2019 6:08am

https://youtu.be/yabXR4kINgs

First two starts - all ok. Then the 3rd and fifth you can see the shuddered start.

larsb   100 kW

100 kW
Posts: 1109
Joined: Dec 10 2014 5:12am
Location: Gothenburg, Sweden

Re: Ebike issues - strange and unseen

Post by larsb » Feb 25 2019 2:06pm

Interesting, hard to tell what it is since it's such a short driving. Can you spin it up from zero to max a few times and post a video?
Does it have torque?

Phase current should be lowered before you do it if you can program your controller, if phase angle or connections is off then you could burn the motor otherwise. Or if you have a three speed switch then set the low or mid setting.
Ride on :D

Navarro   10 mW

10 mW
Posts: 31
Joined: May 12 2018 1:55am

Re: Ebike issues - strange and unseen

Post by Navarro » Feb 25 2019 2:19pm

I will try to do this, hopefully I will not destroy anything...
Did some more troubleshooting, all seems ok, except for the shuddered start once in 4-5 starts. Also, the power indicator in the display is always quite low (I am always expecting 300-400w on the power meter but it does not jump more than 20-30w - really really low even if the wheel spins very fast).
Will keep you posted.


LATER EDIT:
1.With the wheel on the groud - no torque, the controller just shuts off and I get "03 info" error.
2.Ok...Check Halls sensors again - the result is in the video below.
https://youtu.be/qcUj9YmWGas

3.Controller is new. Pic is below.
4.With the wheel up, shuddered start once in 4-5 starts, power meter shows 0w even when spedometer shows 10-14 kmph.
5.Checked throttle... Video below.
https://youtu.be/114aDO4f-rw
6.Checked controller phase... Video below.
https://youtu.be/p-VTDAYktW4

Mindblown right now. How can the second controller see Hall issues (like the first one - which the tester showed it as faulty) when the tester shows ok? And why do I get the strange behaviour with the wheel up?
Starting to believe a Hall-less controller is the way to go.
Last edited by Navarro on Feb 25 2019 4:06pm, edited 3 times in total.

Navarro   10 mW

10 mW
Posts: 31
Joined: May 12 2018 1:55am

Re: Ebike issues - strange and unseen

Post by Navarro » Feb 25 2019 3:34pm

Controller tag
IMG_20190225_223354.jpg

Post Reply