E-bike battery health test: four checks you can do at home, and when to re-cell or replace
BatteryAny e-bike
Four independent measurements that anyone can make without opening the pack: a range test, a voltage sag test, a charge time test and the app or dealer state-of-health reading. Together they tell you whether the pack is fine, evenly worn, or has a bad cell group, and whether a cell replacement is worth it.
Warning: Measure only at the connectors the pack already exposes; never open the case or probe inside it.
What you need
- the original charger and a clock
- a repeatable test route of 10 to 20 km with a hill, ridden in one fixed assist level
- the bike's trip meter or a phone with a ride app
- a multimeter (optional, for the sag test on packs with an accessible connector)
- the brand app where available (Bosch eBike Flow, Shimano E-TUBE, Specialized Mission Control, Giant RideControl)
- the pack's rated capacity in Wh from its label
Step by step
1. Establish the baseline
Read the capacity from the label (for example 36 V 13.4 Ah = 482 Wh). Find the range you got when the pack was new on a comparable ride, or use the manufacturer's range table for your assist level and halve it for real-world riding. Write both down.
If OK: You have a Wh figure and a reference range: go to step 2.
If not: No idea of the original range: use the Wh figure and the rule of thumb of 10 to 15 Wh per km in a middle assist level on flat ground for a 500 Wh pack, then compare later tests with each other rather than with new.
Read the capacity from the label (for example 36 V 13.4 Ah = 482 Wh). Find the range you got when the pack was new on a comparable ride, or use the manufacturer's range table for your assist level and halve it for real-world riding. Write both down.
If OK: You have a Wh figure and a reference range: go to step 2.
If not: No idea of the original range: use the Wh figure and the rule of thumb of 10 to 15 Wh per km in a middle assist level on flat ground for a 500 Wh pack, then compare later tests with each other rather than with new.
2. Charge time test
Ride the pack to cut-out or to the display's empty warning. Note the time, connect the charger, and note when the charger LED turns green. Divide the pack's Wh by the charger's output (for example 42 V x 4 A = 168 W). A healthy pack takes roughly Wh / W hours plus 20 to 30% for the slow final phase.
If OK: Charge time within 20% of the calculation: the pack is taking a full charge. Go to step 3.
If not: Much shorter than calculated: the pack is storing less than its label, either evenly worn or ending early because one cell group is high (imbalance). Much longer: charger or a high-resistance charge port. Note it and continue.
Ride the pack to cut-out or to the display's empty warning. Note the time, connect the charger, and note when the charger LED turns green. Divide the pack's Wh by the charger's output (for example 42 V x 4 A = 168 W). A healthy pack takes roughly Wh / W hours plus 20 to 30% for the slow final phase.
If OK: Charge time within 20% of the calculation: the pack is taking a full charge. Go to step 3.
If not: Much shorter than calculated: the pack is storing less than its label, either evenly worn or ending early because one cell group is high (imbalance). Much longer: charger or a high-resistance charge port. Note it and continue.
3. Full-charge LED and rest check
After the charger goes green, leave it 1 hour, unplug, rest 30 minutes, press the button. All LEDs should light. On the bike, the display should read 100%.
If OK: All LEDs and 100%: the top of the pack is fine. Go to step 4.
If not: One LED short every time, or the display never reaches 100%: cell imbalance. Do two more full cycles; if it persists, that is a weak cell group.
After the charger goes green, leave it 1 hour, unplug, rest 30 minutes, press the button. All LEDs should light. On the bike, the display should read 100%.
If OK: All LEDs and 100%: the top of the pack is fine. Go to step 4.
If not: One LED short every time, or the display never reaches 100%: cell imbalance. Do two more full cycles; if it persists, that is a weak cell group.
4. Range test
Reset the trip meter and ride the test route in the chosen assist level from full until cut-out, or note the percentage at the end of the route and scale up. Repeat on a warm day for the fairest result. Record the km and any sudden jump or cut-out.
If OK: 70% or more of the reference range, no sudden cut-out: the cells are in usable condition. Go to step 5 to confirm with sag.
If not: Under 70%: significant wear. Sudden death at a mid percentage: bad cell group, not general wear. Continue to step 5.
Reset the trip meter and ride the test route in the chosen assist level from full until cut-out, or note the percentage at the end of the route and scale up. Repeat on a warm day for the fairest result. Record the km and any sudden jump or cut-out.
If OK: 70% or more of the reference range, no sudden cut-out: the cells are in usable condition. Go to step 5 to confirm with sag.
If not: Under 70%: significant wear. Sudden death at a mid percentage: bad cell group, not general wear. Continue to step 5.
5. Voltage sag test (optional)
At about 50% charge, measure the rest voltage at the discharge connector. Then, with a helper reading the meter, hold the rear brake and apply the walk assist or throttle for 3 seconds so the motor draws current, and read the voltage under load. Subtract. Do it at room temperature.
If OK: Sag under 2 V (36V), 3 V (48/52V): internal resistance is healthy; any range loss is simple capacity fade.
If not: Sag over 4 V, or the BMS cuts out during the test: high internal resistance or a collapsing cell group. The pack will feel weak on hills and the percentage will jump; it is near end of life regardless of the range figure.
At about 50% charge, measure the rest voltage at the discharge connector. Then, with a helper reading the meter, hold the rear brake and apply the walk assist or throttle for 3 seconds so the motor draws current, and read the voltage under load. Subtract. Do it at room temperature.
If OK: Sag under 2 V (36V), 3 V (48/52V): internal resistance is healthy; any range loss is simple capacity fade.
If not: Sag over 4 V, or the BMS cuts out during the test: high internal resistance or a collapsing cell group. The pack will feel weak on hills and the percentage will jump; it is near end of life regardless of the range figure.
6. Percentage bounce
After the range test, note the percentage at the bottom of a hill, at the top, and again after a 2-minute rest.
If OK: Percentage drops steadily and does not recover after the rest: gauge and cells consistent.
If not: Percentage drops 15 points or more on the climb and climbs back after the rest: the gauge is reading sag; confirms high internal resistance from step 5.
After the range test, note the percentage at the bottom of a hill, at the top, and again after a 2-minute rest.
If OK: Percentage drops steadily and does not recover after the rest: gauge and cells consistent.
If not: Percentage drops 15 points or more on the climb and climbs back after the rest: the gauge is reading sag; confirms high internal resistance from step 5.
7. App or dealer state of health
Bosch Smart System: eBike Flow app, battery section, state of health. Bosch BES2: dealer DiagnosticTool report, shows cycles and health. Shimano: E-TUBE Project app, battery, charge count and health. Specialized: Mission Control, Diagnose, battery health percent and cycles. Giant EnergyPak Smart: RideControl app. Yamaha and generic packs: no app; rely on the ride and sag tests.
If OK: App health agrees with the range test within 10 points: trust both. Write the value down with the date.
If not: App health much higher than the range test: check tyres, brake drag and contacts before blaming the pack; or a single weak cell group that the app averages away. App much lower: the BMS has logged something the ride did not show, ask the dealer for the full report.
Bosch Smart System: eBike Flow app, battery section, state of health. Bosch BES2: dealer DiagnosticTool report, shows cycles and health. Shimano: E-TUBE Project app, battery, charge count and health. Specialized: Mission Control, Diagnose, battery health percent and cycles. Giant EnergyPak Smart: RideControl app. Yamaha and generic packs: no app; rely on the ride and sag tests.
If OK: App health agrees with the range test within 10 points: trust both. Write the value down with the date.
If not: App health much higher than the range test: check tyres, brake drag and contacts before blaming the pack; or a single weak cell group that the app averages away. App much lower: the BMS has logged something the ride did not show, ask the dealer for the full report.
8. Decision
Combine: range 70% or more and sag normal: keep riding. Range 50 to 70% and sag normal: evenly worn, decide by budget and need. Range under 50%, sag high, sudden cut-outs, or one LED short: bad cell group, replace or re-cell now. Then price it: a re-cell keeps the case and BMS and costs roughly half of a brand pack, but is only worth it on packs where the BMS is a standard type (most generic and many older brand packs) and where a service offers a warranty; Bosch Smart System, newer Specialized and EnergyPak Smart packs may refuse a rebuilt pack.
If OK: You have a verdict and a route. See verdicts for the next step.
If not: Results contradict each other: ride it for another two weeks, repeat the range test, and take the log to a dealer; two consistent tests beat one.
Combine: range 70% or more and sag normal: keep riding. Range 50 to 70% and sag normal: evenly worn, decide by budget and need. Range under 50%, sag high, sudden cut-outs, or one LED short: bad cell group, replace or re-cell now. Then price it: a re-cell keeps the case and BMS and costs roughly half of a brand pack, but is only worth it on packs where the BMS is a standard type (most generic and many older brand packs) and where a service offers a warranty; Bosch Smart System, newer Specialized and EnergyPak Smart packs may refuse a rebuilt pack.
If OK: You have a verdict and a route. See verdicts for the next step.
If not: Results contradict each other: ride it for another two weeks, repeat the range test, and take the log to a dealer; two consistent tests beat one.
What you now know for sure
- Healthy: Range 70% or more of reference, normal sag, all LEDs at full, app health above 70% where available. → Nothing to do. Repeat the range test once a year, store at 40 to 60% when not in use.
- Evenly worn, usable: Range 50 to 70%, normal sag, no sudden cut-out, charge time proportionally shorter. → Plan a replacement when the range no longer covers your rides. A re-cell is an option on generic and older brand packs; a second pack for long rides is another.
- Bad cell group: Sudden death at a mid percentage, one LED short after full charge, LEDs flashing during charging, or an app health that drops sharply between checks. → Retire the pack now; the weak group is being over-stressed every ride and is the failure mode that leads to swelling. Re-cell (whole set, never single cells) or replacement.
- High internal resistance: Sag over 4 V, percentage bounces on hills, motor feels weak on climbs, range may still look acceptable on the flat. → End of useful life for a bike with hills. Replacement or re-cell; a cell service will confirm with a capacity and IR test on the bench.
- Not the battery: App health good, sag normal, charge time normal, but range is low. → Tyres, brakes, bearing drag, contacts, a new heavier assist setting or cold weather. Fix those and repeat the range test.
- Re-cell worth it: Case and BMS intact, a standard BMS type, a service with a warranty, and the pack is a generic or older brand pack whose replacement costs at least twice the re-cell. → Send it to a re-cell service with the charger. Ask for matched new cells of the same or higher capacity, a printed capacity test on return and at least a 1-year warranty.
Manuals and video
Videos found on YouTube and checked automatically; they are other people's work, not ours.
- How to Test if Your eBike Battery is Dead Tower E-Bikes How To
- How to check your E-bike battery and signs that your battery is damaged. Bicycling
- How to test the Wallke H6 ST ebike battery voltage? Wallke Ebike
Frequently asked
How many cycles does an e-bike battery last?
Quality packs from Bosch, Shimano, Yamaha or Specialized are designed for around 500 to 1000 full cycles to 70 to 80% of original capacity; generic packs vary from 300 to 800. A full cycle is 100% of the capacity used, so two rides from 100 to 50% count as one cycle.
Is the state of health in the app reliable?
Reasonably, for an evenly ageing pack. It is calculated by the BMS from the charge it can store and is corrected for temperature. It tends to hide a single weak cell group, which is why this flow always pairs it with a range test.
Is a cell replacement service safe?
A good one is: it replaces the whole cell set with matched new cells, keeps or replaces the BMS, tests capacity on the bench and gives a warranty. Avoid anyone who offers to replace only a few cells, or who cannot show you a capacity test result; mixed old and new cells are how re-celled packs fail.
Still stuck? Take a photo of the display and show it to a local e-bike workshop. Bring the bike's system name (e.g. Bosch Performance Line CX) and the exact code; that saves diagnostic time and money.