BMS reset:your battery still will not charge — work these six checks in order
Disconnecting the negative, holding a reset button, or simply waiting — if the pack still will not charge, it is almost never a failed reset. It is that the condition that tripped the protection has not cleared. An under-voltage lockout only releases once external charge current is detected; a cell that fell below 2.0V puts the BMS into a hard lockout that refuses to self-reset; below the low-temperature threshold the charge path never closes at all; and with a wide cell spread the over-voltage trip fires the moment the first cell reaches the ceiling. The six steps below run in field order, each one stating how to tell and what to do — plus three things you must never do yourself.
verifiedEvery threshold here is taken from engineering and after-sales pages already published on this site (uphill cut-off diagnosis, smart BMS, sub-zero operation, the field diagnostics centre) and can be cross-checked item by item. Read the prohibitions in section 4 before touching the pack.
Pack will not charge: the six-step checklist, in order
Do not skip a step. The first two rule out most misdiagnoses without opening anything, step three decides whether the pack is still recoverable, and steps five and six are about what happens once a charger is connected.
- 01multimeter
First confirm it is BMS protection, not a flat battery
How to tell: the charger fan never starts, the indicator never changes and there is no current at the output. Multimeter the battery output connector — static voltage reads 0V, or clearly below the platform nominal (48V platform: 51.2V nominal / 58.4V full charge; 24S platform: 76.8V nominal / 87.6V full charge). No voltage at all means the BMS has opened the discharge path, which is a different thing from low charge.
Tool: digital multimeter
What to do: read the BMS freeze frame: ERR-02 Peak-OCP (hardware over-current), ERR-05 Cell-UVP (cell under-voltage), ERR-09 MOS-OTP (MOSFET over-temperature). OCP is usually the transient inrush of a loaded pull-away or a speed hump and self-recovers about 5 seconds after the throttle is released; OTP needs roughly 15 minutes parked to cool; UVP takes you to step 3. - 02ac_unit
Cold protection: discharging at −20°C does not mean it will charge
How to tell: ambient at or below 0°C, charge current never rises, or the BMS opens the charge path immediately. This is a hard protection, not a fault: lithium intercalation into the graphite anode slows sharply in the cold, and forcing a high current in plates lithium on the anode surface, growing dendrites that pierce the separator. The BMS refuses charge on purpose.
Low-temperature charge cut-off < 10ms
What to do: move the pack indoors above 10°C and let it equalise before charging. Cold-region models with a self-heating film let you plug straight in: the BMS first directs external power to the heating film, then moves to staged charging once the cell core reaches 10°C; a scheduled self-heating wake-up is also available in vehicle mode. Do not repeatedly plug in and force it, and do not try to "charge it a few more times" until it takes current. - 03battery_alert
Over-discharge: needs a dedicated trickle device — never a customer-side fix
How to tell: once the pack has rested, measure every string individually. All strings above 2.2V sit in the recoverable window. If any string has fallen below 2.0V, the BMS forces the discharge path open and refuses to self-reset — this is the "it only powers up with the charger connected" hard lockout, the highest safety tier, and no reset button releases it.
All strings > 2.2V recoverable / any string < 1.5V scrap
What to do: with all strings above 2.2V, wake the pack with a 0.05C trickle repair charger or the factory charger at low current. Any string below 1.5V has passed the copper-foil dissolution limit and must never be recharged — scrap and replace it. Waking a deep-discharged pack is a professional job that needs per-string data. Do not let the customer short anything or force charge it — the reasons are in the next section. - 04balance
Over-voltage and balancing: once the spread is wide enough, the top cell trips at full charge
How to tell: the pack trips at 80–95% state of charge, or charges for ten minutes, stops, then resumes by itself. Read the highest cell voltage at the instant of the trip: at 3.65V per LiFePO4 string the BMS opens the charging MOS loop immediately to prevent swelling. If the static spread exceeds 100mV, the top string hits the ceiling at the end of every charge while the lowest string never fills — that is not a charger problem.
Charging MOS opens at 3.65V per string
What to do: do not push through it with "a few more charges". Run a balancing cycle: active balancing transfers energy bidirectionally at 1.5–2.0A and can level a 24S pack to within 15mV in 45 minutes; passive resistive drain needs a long rest instead. The factory matching basis is ΔV ≤ 3mV with ΔIR ≤ 0.5 mΩ; a spread that keeps widening means a lagging cell is in the pack, so follow the delta-repair procedure in the field diagnostics centre. - 05power
Charger mismatch: wrong voltage platform, or a lead-acid constant-voltage charger
How to tell: three classic cases — a 60V charger on a 48V pack, a lead-acid constant-voltage charger on lithium, or a plain constant-voltage supply with no communication. The results are no charge, no end-of-charge transition, repeated nuisance trips, and in the worst case a permanently latched BMS. A lead-acid charger carries a de-sulphation high-voltage pulse stage and a float stage whose termination voltage does not match LiFePO4 at all; sustained use stretches the cell spread past 300mV.
CC-CV curve plus BMS handshake within 50ms
What to do: fit a CC-CV charger matched to the platform (58.4V cut-off on 48V, 87.6V cut-off on 24S) and confirm the charger-to-BMS CAN / RS485 handshake completes, which should take under 50ms. If a lead-acid charger is still in the shop, retire it — it is the single most common cause of a voided warranty. - 06memory
Handshake failure: the pack charges, but the vehicle thinks it is not ready
How to tell: the pack accepts charge current, but the dash shows no SOC, reports a communication fault and the vehicle logic refuses to power up — the real cause is that the charger or vehicle never completed the handshake on the BMS link (CAN / RS485 / UART), so the vehicle believes the pack is not ready. With power off, measure across CAN_H and CAN_L: it should read 60Ω (two 120Ω terminations in parallel). Baud rate must match the cluster (250kbps or 500kbps). After 3 seconds without bus traffic the BMS drops into restricted limp mode or lights a fault lamp.
CAN_H to CAN_L should read 60Ω
What to do: check the termination resistors, the CAN_H / CAN_L order, single-end shield earthing, and whether the comms pair is bundled parallel to the three-phase motor cable (keep more than 10cm — PWM interference under hard acceleration causes dynamic packet loss). Frame and register definitions are in the CAN bus guide and the BMS protocol centre.
Battery fault or vehicle fault: the cross-swap method
Do not guess. Move the pack to another vehicle on the same voltage platform, or fit a known-good pack to the faulty vehicle — the two results together separate the two sides.
More likely the battery
Fit this pack to another vehicle and the symptom follows it- cancelStatic outputNo voltage at the output connector (0V), and no charge current even with the charger connected
- cancelCell dataOne string below 2.2V, or a static spread above 100mV that keeps widening
- cancelInternal hot spotThermal scan shows a local hot spot inside the enclosure (MOSFET bank, busbar or terminal above 60°C)
- cancelCross-checkSymptom travels with the packIf it also fails to charge on another vehicle, the battery side is confirmed
More likely the vehicle
Fit a known-good pack and the symptom stays- check_circleThe pack itselfIt charges normally on another vehicle and its spread is inside the ≤20mV field specification
- check_circleCommunicationThe dash reports a comms fault while the pack still accepts charge current — the problem is the vehicle-side link and protocol alignment
- check_circleHarness and connectorsA hot or softened charge port, Anderson plug or crimped lug; a floating earth can also latch the BMS permanently
- check_circleCross-checkSymptom stays with the vehicleIf the fault returns with a good pack, move to the charger model, cluster protocol and vehicle wiring
What you must never do yourself
Never short BMS P- to B-
Shorting bypasses every semiconductor protection. With the cell-level second line removed, an external short circuit leads straight to fire. Not one of the six steps above requires it. Of all the backyard "battery rescues", this is the one with the worst consequences.
Never open a sealed steel-enclosure pack
The pack is an IP67 sealed steel enclosure with a potted BMS bay and a high-voltage DC busbar inside. Opening it destroys the seal, the insulation and the potting; moisture ingress makes the internal short-circuit risk irreversible, and without factory tooling the seal and bolt torque cannot be restored — the warranty basis goes with it.
Never force a charge with a non-original charger
A charger on the wrong voltage platform, a lead-acid constant-voltage charger, or a constant-voltage supply without a BMS handshake will drive the pack into over-voltage protection or gas the cells. Putting a 60V charger on a 48V pack, or pushing high current in below 0°C, turns a controllable protection trip into an uncontrollable cell failure.
BMS reset and no-charge questions we get most
Everyone says disconnect the negative or hold a reset button. Why does my pack still not charge after that?expand_more
How do I tell whether a 48V pack's BMS is locked out?expand_more
It will not charge in the cold — has the battery failed?expand_more
The charger never moves to the end-of-charge state, or stops every ten minutes. Why?expand_more
Can I wake an over-discharged pack myself with an adjustable bench supply at low current?expand_more
Three pages to read next
Smart BMS battery: buying a battery is buying its BMS
How thresholds are matched to the duty cycle, active versus passive balancing and service life, and the CAN 2.0B frame and Modbus-RTU register lookup.
Open the BMS topicarrow_forwardUphill cut-off diagnosis and BMS current limits
The 192.4A inrush, the bus sagging 64.2V to 49.1V, 18ms MOSFET shut-off, plus the four-step field SOP and freeze-frame lookup.
Open the diagnosis guidearrow_forwardTroubleshooting and field diagnostics centre
A six-symptom fault matrix, cell delta and internal-resistance checks, charging faults and nuisance trips, and the minimum field toolkit.
Enter the diagnostics centrearrow_forwardSend your readings — we will tell you whether to reset, replace or return it
Four items are enough: static output voltage, per-string cell voltages (or the BMS freeze-frame code), charger model and voltage platform, and ambient temperature. The engineering team replies within 24 hours on working days with a judgement and the next action.
- Fleet case studies in the same duty cycle available under NDA
- Volume customers can request live remote access to the low-level registers
- New projects can start with a 3–5 day sample pack before committing to a container
Specification notice:The figures above are engineering reference values. They may differ with production batch, configuration and actual duty cycle, and the percentages for life, efficiency, availability and coverage are engineering estimates and bench references. Contact us for the detailed specification sheet.