You went to ride and nothing happened. The display is dark and won't wake, or it lights up fine but the motor gives you nothing — you're just pedaling a heavy bicycle — or worst of all it worked this morning and now it stutters and cuts out every time you hit a bump. On a machine that cost one to three thousand dollars, the mind goes straight to the two most expensive words on it: dead motor, or dead battery.
It is almost never either one. Here's the single idea that reorganizes every one of those symptoms, and that almost no e-bike troubleshooting says plainly:
The chain: a row of "yes" gates, and the motor is at the far end
Think of turning your e-bike on as a series of handshakes, each of which can quietly refuse. In rough order, before a single watt reaches the wheel:
- The battery has to physically connect — seated in its cradle, its contacts clean, and switched on. A pack that's a millimeter proud of its rails, or has a corroded contact, delivers nothing.
- The battery's own brain (the BMS) has to agree to deliver power. A lithium battery is not a fuel tank — it has a management board that can, and will, refuse to put out current (more on this below). A "dead" battery is very often a battery that has decided to say no.
- The key/lock and power button have to be on, and the display has to power up and complete its handshake with the controller.
- Neither brake can be "held." Nearly every e-bike cuts motor power the instant you pull a brake lever — a safety feature. If a lever or its little switch is stuck, the bike believes you're braking forever and will not drive the motor, no matter what else you do.
- A valid drive command has to arrive — either the throttle's sensor or the pedal-assist sensor telling the controller "the rider wants power now."
- Only then does the controller fire the motor.
Notice what this means: the motor sits at the end of a corridor of five doors, and every one of those doors is cheaper and more exposed than the motor itself. Contacts corrode, a BMS goes to sleep, a brake switch jams, a connector wiggles loose, a pedal sensor drifts out of alignment. When someone says "the motor's dead," they've almost always got a door stuck shut — and they're about to spend hundreds replacing the one part at the end of the hall that's still fine.
The fact nobody tells you: the battery is a bouncer, not a fuel tank
Here's the half of the reframe that explains the scariest symptom — the bike that's completely, silently dead. Inside every lithium e-bike battery is a Battery Management System (BMS): a small circuit board whose whole job is to protect the cells by cutting the output. It is designed to refuse. It will shut the pack's output off — hard — when it sees over-discharge (voltage too low), over-current (too much draw), a cell that's too hot or too cold, or cells that have drifted out of balance. When it does, the battery goes electrically silent: no lights, no output, nothing to power the bike. It looks exactly like a dead battery, even though the cells inside can be perfectly healthy.
The most common version is over-discharge sleep. Leave an e-bike sitting for months without charging and the pack slowly self-drains below the BMS's low-voltage cutoff; to protect the cells, the BMS latches off and stops responding. Now the battery "won't turn on and won't charge" — and people junk a $500–$800 pack over it. Manufacturers and battery makers describe the fix directly: often you simply leave the charger connected — sometimes for 20–30 minutes, sometimes several hours or overnight per the maker's instructions — and the small trickle of voltage is enough to wake the BMS back up so normal charging resumes. (The honest caveat: a battery that has sat deeply discharged for a very long time can drop too low for a standard charger to revive safely — in that case the maker has a proper procedure, and you should use it rather than a risky DIY jump.)
How to tell which side is at fault: watch the charger. If the charger's light behaves normally when connected (shows charging, then goes to "full") but the bike still won't run, suspect the connection to the bike or a gate downstream. If the charger goes green instantly or won't start at all, and the pack has been sitting, suspect a sleeping/protecting BMS — and leave it connected to wake it. If the charger itself shows no light into a known-good outlet, you may have a dead charger, not a dead battery — a $30 part people routinely misread as a $600 one. Test the charger and the battery as two separate suspects.
Symptom 1: "It won't turn on at all — the display is dark"
Nothing wakes. No backlight, no beep. Power isn't even reaching the display, which means you're failing one of the first gates — and every one of them is cheap. Work them in this order before you ever think about electronics:
- Battery seating and contacts. Slide the pack out and back in until it clicks fully home; a pack that isn't seated all the way is the number-one "my bike is dead." Look at the metal contacts on both the pack and the cradle — corrosion, grime, or a bent pin breaks the circuit. Clean them (contact cleaner and a gentle scuff), make sure the pack's own key/switch is on, and confirm it's actually charged.
- The battery is asleep. If it's seated, clean, and still nothing, it's the BMS story above — put it on the charger and give it time to wake.
- The main fuse or a loose main connector. Many packs and controllers have an inline fuse or a large main power connector between battery and controller. A blown fuse or a connector that vibrated half-loose kills everything downstream. These are cheap and user-checkable.
- The key/lock and power button. Some bikes won't power up unless the frame lock is turned to the right position or a specific power-on sequence is used. Check the manual's exact power-on step — people miss it after a battery swap.
Symptom 2: "The display is on, but there's no assist and the throttle does nothing"
This is the most misdiagnosed symptom on an e-bike, because the display lighting up proves the battery, the connection, and the controller are alive. You've cleared the first gates. Power got into the system — and then a later gate refused. The motor is almost certainly fine. Check these, roughly in the order they go wrong:
The brake motor-cutoff switch — the single most common culprit
Nearly every e-bike is wired to cut motor power the instant you pull either brake lever. It's a safety interlock: brake and the assist stops. The catch is that this is done with a tiny switch (or a small magnet and sensor) at each lever — and if one gets stuck "engaged," the bike believes a brake is being held all the time and will refuse to drive the motor. No assist, dead throttle, everything else normal. It gets stuck from a drop or a knock that jams the lever, from a magnet that shifted out of alignment, or from the switch simply wearing out or getting wet.
The tell: if both pedal-assist and throttle are dead but the display and lights are fine, a stuck brake cutoff is the first thing to suspect. You can confirm it: wiggle each lever and watch whether assist returns; or, following your wiring, briefly disconnect a brake sensor's little two-pin connector — if the motor comes back to life with that brake unplugged, that lever's switch (a $10–$15 part) is your problem. A multimeter continuity check is the clean version: a healthy two-pin brake switch should only show continuity when the lever is pulled; if it reads continuous with the lever released, it's stuck engaged and telling the controller to keep the motor off.
The pedal-assist (cadence) sensor — assist gone, but the throttle still works
If your throttle works but pedal-assist doesn't, that split points almost exactly at the pedal-assist sensor. Most e-bikes read your pedaling with a magnet disc on the crank spinning past a sensor mounted a few millimeters away. A bump, a crank service, or a knock can shift the magnet ring or the sensor so the gap is wrong — and even a few extra millimeters makes the sensor "blind" to your pedaling. It stops registering rotation, so the controller never gets the "rider is pedaling, give assist" signal. Often there's no error code at all — everything looks normal, the bike just won't help you pedal. The fixes are cheap and mechanical: clean the sensor face, re-seat the magnet disc square on the axle, and set the gap back to spec. And check the dumbest cause first — the assist level set to 0 gives you exactly "pedaling does nothing."
The throttle's sensor — throttle dead, but pedal-assist still works
The mirror image: if pedal-assist works but the throttle is dead, suspect the throttle itself — its internal Hall sensor, its connector, or water that got into it. A throttle is a small, exposed, replaceable part; a dead one is not a dead motor.
Symptom 3: "It cuts out over bumps, or works only if I jiggle the wires" — a connector, not a component
Power that comes and goes with motion — dies over a pothole, returns when you wiggle a cable, drops when you flex the bars — is the signature of a connection, not a failed part. E-bikes are rolling vibration machines with dozens of little waterproof connectors, and the usual offender is one that's half-seated, corroded, or a wire that's chafed where it flexes (at the bars, at the headset, or at the motor cable near the dropout). A component that has actually failed tends to stay failed; a fault that appears and disappears with movement is almost always mechanical. Firmly re-seat the connectors along the harness (battery-to-controller, display, brakes, throttle, motor), look for a pinched or frayed spot at the flex points, and dry out anything wet. This is a five-minute check that resolves a huge share of "intermittent dead" bikes.
Symptom 4: "It cuts out under load or on hills, then comes back" — that's a limit, not a break
Different pattern, different cause. If the bike is fine on the flat but cuts power exactly when you demand the most — a steep hill, hard acceleration, a heavy load — and then recovers after it rests, you're most likely hitting a protective limit, not a broken part:
- BMS over-current or low-voltage cutoff. Under heavy load the pack sags; if it draws more current than the BMS allows, or the voltage dips below its cutoff, the BMS trips to protect the cells and the motor goes dead for a moment, then resets. On an older or cheaper pack this happens sooner — it's a capacity/health signal, and cross-references the reason the battery, not the motor, is the part that really ages on an e-bike.
- Controller thermal cutback. Push a motor hard up a long hill and the controller (or motor) heats up; many will dial power back or cut out to protect themselves, then return once cooled. Annoying, but it's the electronics protecting themselves — not a failure.
- A Hall sensor in the motor. If it stutters, jerks, or cuts specifically from a standing start or under load, a motor Hall sensor (or its wet/loose wire) can be feeding the controller bad position data. That's a genuine motor-side issue — but it's usually a sensor or a connector, and often repairable, not a whole new motor.
The error code is the bike telling you which gate failed — read it
If your display shows a code (an E number, or a "P" code on some systems) when it fails, don't guess — the bike is naming the failed gate for you. Most controller systems publish an error-code table, and the codes cluster around exactly the gates in this article: communication between display and controller, brake signal, throttle fault, a motor Hall-sensor or speed-sensor fault, over/under-voltage. Specific numbers vary by controller and brand — one maker's E21 is not necessarily another's — so look up your system's table rather than trusting a generic list. A code is a gift: it turns "no power" into "the throttle line faulted" or "the speed sensor isn't reading," which is a five-minute fix instead of a teardown.
Read the symptom to the fix: the whole table
| Symptom | What it usually is | The fix | Dead motor? |
|---|---|---|---|
| Totally dead — display won't light at all | Power not reaching the system — pack unseated / dirty contacts / off / a blown main fuse | Re-seat & switch on the pack; clean contacts; check the inline fuse & main connector (~$0–$15) | No |
| Dead, and the battery won't charge or has sat unused for months | BMS asleep / over-discharge protection (a healthy pack refusing) | Leave it on the charger to wake the BMS; if it stays dead, use the maker's revive procedure | No |
| Display on, but no assist and no throttle | A gate says "no" — usually a stuck brake motor-cutoff switch | Free the lever / check the brake switch; unplug a brake sensor to confirm; replace the switch (~$10–$15) | No |
| Throttle works, but pedal-assist doesn't | Pedal-assist (cadence) sensor misaligned/dirty — or assist level set to 0 | Set assist level up; clean sensor; re-seat the magnet disc & reset the gap (free–cheap) | No |
| Pedal-assist works, but the throttle doesn't | Throttle sensor / connector (or water in the throttle) | Dry & re-seat the throttle connector; replace the throttle if faulty (cheap part) | No |
| Cuts out over bumps / returns when you jiggle a cable | A connection — half-seated or corroded connector, chafed wire at a flex point | Re-seat every connector along the harness; find & fix the pinched/frayed spot; dry wet plugs (free) | No |
| Cuts out under load / on hills, then recovers after a rest | A protective limit — BMS over-current/low-voltage or controller thermal cutback | Ease the load; if it trips early & often, the pack is aging (capacity signal, not a break) | No |
| Stutters / jerks / cuts specifically from a standstill or under power | Motor Hall sensor or its wet/loose wire feeding bad data | The genuine motor-side check: often a sensor or connector repair, rarely a whole new motor | Maybe — get it diagnosed first |
| Display shows an error / "E" or "P" code | The system naming the failed gate (comms, brake, throttle, sensor, voltage) | Look up your controller's code table; fix the named gate — don't guess | Usually no |
Why this is a repair problem, not a replace problem
Step back and look at the entire "no power" failure surface of an e-bike: battery contacts, a main fuse, a BMS that can be woken, two brake cutoff switches, a pedal-assist magnet and sensor, a throttle, a fistful of connectors, and — behind all of them — the pack and the motor. The parts that actually strand people are the cheap, exposed ones at the front of that list. The battery and the motor, the two things everyone fears, are the most robust and the last to fail — and when they do, a pack is often rebuildable and a motor fault is often a sensor. There is no mystery here that a clean contact, a woken BMS, and a $12 brake switch can't out-diagnose most of the time.
Which is exactly why resale sites and curbs fill up with "not working / for parts" e-bikes that needed a re-seated battery and a brake switch — given up on by someone who read a dark display as a dead machine. A used e-bike that powers up, charges, and rolls under its own motor at all is often a real buy: confirm the pack takes and holds a charge (the expensive unknown), that the motor drives, and that the display throws no persistent code, and you've likely got a sound bike someone abandoned early. There's a real give-back reason to bother, too: an e-bike battery is several pounds of lithium and cobalt — a genuine fire and e-waste hazard in a landfill, and a pile of materials worth recovering — bolted to an aluminum frame that will outlast all of us. Reading the symptom keeps the whole machine rolling, and keeps a lithium pack out of the trash.
If you want to understand why the battery is the part worth protecting and diagnosing first — and how to judge a pack's real remaining life — that's the thesis of "the battery is the bike; everything else is a bicycle." And if you're sizing a replacement pack or a whole bike, the range-by-motor-and-battery calculator is the honest place to start.
The bottom line
An e-bike does one electrically honest thing: it passes power to the wheel only after a row of small safety checks all agree. When it won't turn on, has no power, or cuts out, don't ask "is the motor dead?" — ask which gate said no. Is power not reaching the system (a pack unseated, a contact corroded, a fuse blown), has the battery's own brain gone to sleep (put it on the charger and wait), is a gate downstream refusing (a stuck brake switch, a blind pedal sensor, a dead throttle), or is it a connection that only fails when the bike moves (re-seat the harness)? Nine times out of ten the answer is a part you can clean, wake, or swap for a few dollars in an afternoon — not the big-ticket component at the end of the hall.
And when it's running again, the cheapest habits keep it that way: keep the battery contacts clean, never leave a lithium pack sitting dead for months (top it up periodically so the BMS never has to sleep), and re-seat a connector at the first sign of an intermittent cutout. Do that and the bike will out-serve almost everything in the garage. The rest of our e-bike coverage — range, battery life, and the class-1-vs-3 decision — is here when you want it.
Every mechanism here is from e-bike manufacturers' and battery makers' own support docs or standard e-bike electronics practice, cited below. Specifics — error-code numbers, exact charge-to-wake times, brake-switch pin behavior — are representative and vary by brand, controller, and model; verify against your own manual and your system's error-code table. Nothing here is sponsored or affiliated. This is educational troubleshooting: do electrical checks with the battery removed where possible, respect that a deeply discharged lithium pack can be unsafe to force-charge, and take a genuinely damaged or swollen battery to a professional rather than reviving it yourself.
Sources
- Letrigo — "Solving E-Bike Brake Cutoff Switch Problems" and the Endless Sphere DIY EV forum thread on diagnosing a brake cutoff: how a stuck or misaligned brake motor-inhibitor switch makes the bike behave as if a brake is always held (no assist, dead throttle), and the two-pin continuity test — a healthy switch shows continuity only when the lever is pulled.
- Upway — "Why Your E-Bike Motor Isn't Working or Keeps Cutting Out" and Velotric — "Electric Bike Troubleshooting Guide": the ordered list of real causes (brake cutoff, connectors, sensors, settings) behind "no power" before the motor itself, and the throttle-vs-pedal-assist split as a diagnostic.
- Troxus Mobility — "Consumer Advisory for Sleep Mode on Batteries" and Aniioki — "Why Your E-Bike Battery Won't Charge": the BMS entering low-voltage/over-discharge sleep after long storage, the "leave the charger connected to wake the BMS" revive, and the warning that a very deeply discharged pack may need the maker's procedure rather than a standard charger.
- Macfox — "Cadence Sensor Troubleshooting" and Aventon support — "Pedal Assist Not Functioning & No Error Code Present": how a pedal-assist (cadence) sensor's magnet-to-sensor gap going out of alignment kills assist — often with no error code — while the throttle still works, and the clean/re-seat/re-gap fix.
- Letrigo — "Common E-bike Sensor Errors" and "Electric Bike Error Codes List": motor Hall sensors reading rotor position (bad/loose/wet data → stutter, cut-out, or no drive), speed-sensor faults disabling assist and throttle, and the reminder that error-code numbers vary by controller/brand — read your own system's table.
- E-bike drivetrain first principles: power reaches the wheel only after battery presence, BMS permission, key/display handshake, brake-cutoff clearance, and a valid throttle/PAS command — the basis for the "corridor of gates, motor last" framing and the symptom-to-gate fingerprinting above.