EBike LiFePO4 Battery Troubleshooting

This instructable is to help troubleshoot a malfunctioning LiFePO4 (Lithium Iron Phosphate) eBike battery. These batteries are commonly sold online through various sellers. Though they often differ in size, voltage, capacity and cell type, they are electrically similar and the troubleshooting techniques below should hold fast for any multi-cell LiFePO4 battery pack with internal BMS(Battery Management System).

Step 1: LiFePO4 Battery Basics

Lithium Iron Phosphate (AKA - LiFePO4) is increasingly being used for electric vehicles (and as a replacement for Lead-Acid batteries in general) due to it's long lifespan ( >1000 charge cycles), light weight, flat discharge curve and awesome chemical stability. LiFePO4 batteries are far less prone to the fire/shorting issues that Lithium Ion batteries are now famous for. LiFePO4 cells also hold their charge indefinitely. They don't leak off their charge or go dead over time.

Modern eBike batteries use Lithium Iron Phosphate chemistry with multiple cells arranged together in a grid to provide the correct voltage, current and capacity. Batteries packs are described by the number of cells they have in series (S) and in parallel (P). So a 16S4P battery will have 64 cells total (16x4). Battery pack voltage is the sum of single cell battery voltage (about 3.33V-3.35V charged) times the number of cells in series. So our 16S example battery would have a voltage of about 53.5VDC charged. Battery pack amperage is the sum of the individual cell amperage times the number of cells in parallel. So if our example cells were 2000mAh, then out 16S4P pack would have a capacity of 8000mAh.

In addition to the cells there will also be a BMS (Battery Management System) board inside the battery. This board monitors the cell voltages inside the battery and ensures that no individual cell voltage charges to above 3.65V or discharges below 2.0V where they would be damaged. It also balances the cells during charging (ensures each cell receives the same charge) and disconnects the battery from it's load if short-circuit or over-current conditions are detected.

Step 2: Battery Problems

Battery Problems are generally of 2 types-

  • Charging Problems

This is when the battery either won't charge or won't hold a charge. (i.e. after charging all night it's still dead/empty)

  • Discharging Problems

With discharge problems, the main symptom is the bike completely dying under load and not starting again until the battery is disconnected then reconnected to the motor.

The other common discharging issue you may see is that the battery works, but does not last very long. This is indicative of an imbalanced or damaged battery and can be caused by damaged cells or by a broken balance circuit on the BMS not charging all of the series cells equally.

Step 3: Digging In

The video above gives a look inside an ebike battery and describes the basic procedure used to troubleshoot a battery that is cutting out or failing to charge. Also included is a schematic of the BMS circuit. It's not required to troubleshoot, but is there for anyone who is curious...

You will need:

  • a Digital Volt Meter
  • scissors or utility knife to open battery (usually heat shrink)

And the basic troubleshooting steps are-

  1. disconnect battery from eBike.
  2. Open the battery and expose the BMS board
  3. Test DC voltage at the 3 negative cables on the BMS (the cables are the negative terminals for charging discharging and the battery) The difference between each should be 0 VDC (see video)
  4. If charging problems exist repeat 3 with charger attached to battery
  5. Test battery voltage at main discharge connector (should be full battery voltage)
  6. Test battery voltage at main charging connector (should be full battery voltage AND match 5)
  7. Test individual cell voltages at balance connector.(should be between 2.5-3.4 VDC per cell depending on charge. All cells should also have close to the same voltage. Anything under 2.0 or over 3.65 VDC per cell indicates permanent damage to the cell.)
  8. (if still faulting) Bypass Discharge protection and test under load (see video)

These steps will help you determine if you problem lies in your BMS, your cells or your eBike motor and will hopefully help someone rescue their broken battery without having to spend an arm and a leg...

Step 4: Longevity Tips

For your LiFePO4 battery to have a long and happy life-

Charge Slowly

The more wear and tear the internal electrodes of the batter see, the faster the battery will fail. Charging slowly places less stress on the battery. Using that 'Quick Charger' when the bike is going to sit overnight anyway will over time reduce cell lifespan. Charging LiFePO4 batteries quickly also places more wear on the BMS. (My own eBike BMS runs 20 deg. C hotter on a 5A 'quick' charger than it does on a 2.5A 'standard' charger.) Batteries charged slowly balance better too.

Do Not Store Fully Charged

If you are storing your bike for an extended period, run the battery down some first. Storing a dead battery can hose it. But Storing a 100% full battery causes internal battery wear also. (The internal electrodes of the battery wear more quickly when more chemically active. And they are at their most active when fully charged.) So before you put your bike away for winter, run it down to around 2/3 of full.

Check Periodically
When storing you should check the battery every couple months. While LiFePO4 cells do not leak or lose charge, the BMS usually steals some . The Signalab V1 BMS steals juice from only the first 4 cells for it's logc/MOSFET buffering. So theoretically over long term storage those cells can imbalance... (You can see this in the included schematic. The line coming from charging circuit #5 to R36 powers the quad NAND gate IC on the back of the BMS.) Newer versions steal equally from all cells. Either way, checking every once in a while to make sure that the BMS has not run the battery down is prudent.

Do NOT Leave On the Charger

This is probably the most important longevity tip and is for the same reason we don't store it fully charged. Keeping a LiFePO4 cell at ~3.5VDC per cell 'top-off' voltage keeps the internal electrodes at their most active and causes unnecessary battery wear. Once the battery is done charging, unplug the charger.


This is a work in progress and I'm going to keep adding info so feel free to post any questions and I'll answer them if I can.

Thanks for viewing!



  • Remix Contest

    Remix Contest
  • Build a Tool Contest

    Build a Tool Contest
  • Organization Contest

    Organization Contest

12 Discussions

Alex in NZ

2 years ago

Wow. I was trying to recommission my e-Bike after a long break, and repeated charge/balance cycles weren't giving me all LEDs lit. I'd buzzed out the affected cell and it was 0.2v below the other ones. I could not find any relevant data online, and was about to rip the pack apart to investigate further. Your video is the first place I've found which said "LED indicated overvolt. Does not need to appear for cell to be charged.

Thank you _so_ much for the perfectly timed post!

4 replies
boffinryAlex in NZ

Reply 2 years ago

Glad it helped. My Battery is similar, the first 5 cells power the BMS circuitry so since they are always trickling those cell's LEDs almost never come on. Anyone who messes with RC vehicles or multirotors/UAVs is used to balance chargers and how they work. But eBike battery vendors should provide better documentation. Cheers! ;-)

BTW- While I wouldn't pull the battery apart over a 0.2V differential, if you have easy access to the balance connector I would try disconnecting it from the BMS and connecting a 1S LiFePO4 charger to the low cell only to manually. balance it. (Hobby/RC chargers sold by sites like Hobby King have LiFePO4 settings and work nicely for manually balancing cells in battery packs)

Alex in NZboffinry

Reply 2 years ago

Thanks for that. My BMS is external to the main pack, so isolating it from the cells is easy enough without breaching the wrapper on the pack.

I don't have a 1S charger. I had found comment on an e-bike site that said that 50mA could be used to tickle a cell back into line, so was considering using a 5v PSU with a 100 ohms high-power resistor in series to limit the current. Delighted to accept advice on alternatives.

And totally agree that some documentation would have been nice!

boffinryAlex in NZ

Reply 2 years ago

I've seen that trick but prefer something like this - http://www.hobbyking.com/hobbyking/store/__64456__...

You can just set it for 'LiFe' on a slow charge rate and connect the main out (red, black) to JUST the cell you want to charge (via the balance connector on the battery) and it'll charge it as a 1S.

I think it's a better option as The charge controller in that device is computerized and applies the 'correct' charge curve to the battery. (As opposed to the 16 analog charge circuits that are all ganged-together on most ebike BMS boards.)

The only crucial thing to remember is that the charger MUST be set for LiFE. Charging a LiFe (~3.5V) to Lipo(4.2V) voltage will kill it.

Anyway, now you've inspired me and I'm probably going to add a 'how to manually balance one of your cells' video to this Instructable. So Thanks!!! ;-)


Reply 1 day ago

I bought a 5V power supply 2Amps for a LED strip lighting .When my cells are not quite balanced I attach it to the cell that is down and charge it up. As long as you keep the current under 1C it will not harm the cells and you carefully watch the rising voltage.
To discharge a cell I use some large 5 watt square resistors of about 100 ohmm and connect them across the cell and watch the voltage drop.
To learn more on LiFePo4 join this site


Question 5 months ago on Step 4

I have a 10s4p Li-ion battery pack that is faulty. I thought the problem might be with the BMS so I replaced it, but the pack is still faulty. So I have charged each of the parallel cell groups independently and they each measure 4.15 to 4.18 volts. Measuring from the negative terminal of the first group to the positive terminal of the 10th group gives 41.5 volts. Measuring between each of heavy wires on the BMS gives 0 volts. There is also continuity between each pin of the balance connector and the corresponding cell group. Yet when I measure the voltage of the pack (BMS P- terminal to group 10 positive terminal I get 20.5 volts instead of the 41.5 volts I expect. Do you have any suggestions for what to try next?


1 answer

Answer 1 day ago

Tke the BMS off and try running your gear off the battery direct for a short time to confirm the cells are OK . Then it is either the BMS still faulty or its mis-wired .


2 months ago

Lifepo4 cells absolutely will go dead over time and I know because I have a dead battery in front of me right now that was in storage for a few years. If anything, I think lifepo4 will leak it's current out faster than a lithium cobalt due to its porous molecular structure.

1 reply

Reply 1 day ago

Supposedly at 40%charge they will sit on a shelf unharmed for at least 5 years.
If you have discharged the cells too far and then left them sitting what else would you expect . No technology will take that treatment.


Question 6 months ago

This is the BMS of a 10S4P batt pack.Where should I connect the + and the - 42V? 2 rows of 4P show 0.20 V while the rest of the 8 rows is fully charged.
So, where should I connect the + and the - of the chatger just to test the BMS?
A new one is on its way to me.
Do you think I MUST replace the 2 rows of 4P batteries or I may give a try with them?
Tx in advance for your advice.


1 year ago

good tutorial. insightful. My BMS seems the reverse of this. It has 2 red and 3 black wires. It is a 48 v 15 ah Ebay pack. If I leave balancing plugs in it will drop from

57 to about 49 v overnight. I can't go far w/ BMS plugged , so bike is ridden unplugged. Another issue is that once it is fully charged charger stays green

and doesn't vacillate back to red to balance. Also, there are no LEDS, rather there is potted material that gets hot. Probing w/ multimeter is reverse. I probe the points in the plug insert w/ negative probe and red probe on red of bms board. Probing black on black and it reads .3v. Is it the bms or the battery pack?

Riding voltage is around 45, but I'm not sure the pack is disconnected from

BMS. This battery was not abused. It has about 9 years but has less 100 half cycles maybe 20 full cycles.