I picked up a cheap Yuba Mundo cargo bike with a Bionx 48V battery that wouldn’t charge. This is my first foray into any battery related repair and I learned a lot thanks to all the posters here and elsewhere. I didn’t find any thread that detailed my exact issue so I figured I’d share my troubleshooting and resolution.
Low battery warning chirp after attempting to charge overnight
Powers up for a split second when attached to the bike, but then shuts off
I checked the overall voltage and measured it at 43.2V. Other threads here suggest the BMS refuses to charge if the voltage is below 35V or so but I couldn’t find hard documentation from Bionx on this.
Thanks to the teardown post, I got the idea to check the voltage of each balance region. To do so, I followed this video. TL;DR: measure with a multimeter starting with negative on the black negative wire and the positive on the wire to the right of that. For the next region, move negative and positive to the right one wire.
For example, this would check the first balance region:
|-||+|
[0][1][2][3][4][5][6][7][8][9][10][11][12]
And this would check the next one:
|-||+|
[0][1][2][3][4][5][6][7][8][9][10][11][12]
After checking them all, I found one region was reading 0.0V
Adding these up gives me 43.2V - the same as the overall pack voltage and a nice sanity check.
I opened the battery up to see how many cells I need to replace. The balance region had 4 (it's a 13S4P pack) so I decided to just replace all 4. Cells are Samsung ICR18650-22P. I found this datasheet on them, so I decided to hunt for similar cells.
I went to my local recycling center and dug through the electronics bin to find an old laptop battery. I opened it up and was happy to see Samsung cells, specifically ICR18650-26D. Based on the name, I assumed they would be the same but with slightly more capacity. However, the closest ICR18650-26 datasheet I could find suggested the max draw was 5.2C, much less than the alleged 10C of the BionX cells.
However, based on reading some threads here, I began to question the 10C. This recent thread seemed to confirm that the 10C is inaccurate and that 5.2C like the other ICR18650 varieties was more likely.
All that was left was to break the welds and swap in the new cells. I bought a spot welder off Amazon, practiced on the dead cells, and made some passable welds to secure the new ones.
A huge thanks to everyone here for creating and compiling a wealth of information. It allowed me to go from zero ebike knowledge to an eGrin, even if it came more from my successful repair than the paltry 350W motor.
Symptoms
LED ring flashes red when chargingLow battery warning chirp after attempting to charge overnight
Powers up for a split second when attached to the bike, but then shuts off
Diagnosis
This seems to be an excerpt from a Bionx manual, identifying flashing red as the battery being in deep discharge. Based on this I decided to check the voltage of the battery. However, the Battery Management System is inside the case - no voltage goes to the exterior pins if the BMS says the battery is no good. I followed this teardown of the same battery I have to open the well-sealed case and disconnect the battery from the BMS.I checked the overall voltage and measured it at 43.2V. Other threads here suggest the BMS refuses to charge if the voltage is below 35V or so but I couldn’t find hard documentation from Bionx on this.
Thanks to the teardown post, I got the idea to check the voltage of each balance region. To do so, I followed this video. TL;DR: measure with a multimeter starting with negative on the black negative wire and the positive on the wire to the right of that. For the next region, move negative and positive to the right one wire.
For example, this would check the first balance region:
|-||+|
[0][1][2][3][4][5][6][7][8][9][10][11][12]
And this would check the next one:
|-||+|
[0][1][2][3][4][5][6][7][8][9][10][11][12]
After checking them all, I found one region was reading 0.0V
Negative | Positive | Voltage |
Black | Brown | 3.55 |
Brown | Purple/White | 0.00 |
Purple/White | Orange | 3.55 |
Orange | Yellow | 3.57 |
Yellow | Green | 3.59 |
Green | Blue | 3.59 |
Blue | Purple | 3.60 |
Purple | Grey | 3.61 |
Grey | White | 3.62 |
White | Grey/White | 3.62 |
Grey/White | Red/White | 3.62 |
Red/White | Blue/White | 3.63 |
Blue/White | Red | 3.63 |
Adding these up gives me 43.2V - the same as the overall pack voltage and a nice sanity check.
Repair
I was pretty surprised to see a 0.00V reading rather than something super low. It turns out that yes - the Samsung cells have a safety feature where they “pop” if gas inside the cell expands. This cuts off the connection and brings it to 0.00V. There are YouTube videos of people poking holes and releasing the gas to pop the connector back, but this is a Bad Idea™. The safety feature is a one-way street, the cells may have been damaged and are a fire risk even if they still have capacity. Plus you’re now removing a safety feature.I opened the battery up to see how many cells I need to replace. The balance region had 4 (it's a 13S4P pack) so I decided to just replace all 4. Cells are Samsung ICR18650-22P. I found this datasheet on them, so I decided to hunt for similar cells.
I went to my local recycling center and dug through the electronics bin to find an old laptop battery. I opened it up and was happy to see Samsung cells, specifically ICR18650-26D. Based on the name, I assumed they would be the same but with slightly more capacity. However, the closest ICR18650-26 datasheet I could find suggested the max draw was 5.2C, much less than the alleged 10C of the BionX cells.
However, based on reading some threads here, I began to question the 10C. This recent thread seemed to confirm that the 10C is inaccurate and that 5.2C like the other ICR18650 varieties was more likely.
All that was left was to break the welds and swap in the new cells. I bought a spot welder off Amazon, practiced on the dead cells, and made some passable welds to secure the new ones.
A huge thanks to everyone here for creating and compiling a wealth of information. It allowed me to go from zero ebike knowledge to an eGrin, even if it came more from my successful repair than the paltry 350W motor.