I’ve lost count of how many times someone has walked into the shop holding a control module in a plastic bag, absolutely convinced it’s toast. Nine times out of ten, we plug it back in, run a few checks on the harness, and find a chafed wire, a crusty ground strap, or a relay that’s been sticking for six months.

That’s the honest truth about ECU failure: the module usually gets blamed for the sins of everything around it.

So let’s talk about what actually kills an ECU, what only looks like a dead ECU, and how to tell the difference before you spend serious money.

First, a quick vocabulary check

You’ll see four names thrown around, and people use them interchangeably:

• ECU — Engine Control Unit. The generic term.
• ECM — Engine Control Module. Same idea, different badge.
• PCM — Powertrain Control Module. Runs the engine and the transmission.
• Control module — the catch-all term for any of the 20–80 computers scattered around a modern car (BCM, TCM, ABS module, gateway module, and so on).

Under the lid, they’re all the same animal: a sealed aluminium or plastic box holding a printed circuit board, a microcontroller, power supply circuitry, and rows of driver transistors that switch injectors, coils, solenoids and relays on and off thousands of times a minute.

That last part matters. An ECU isn’t just a brain — it’s a power switching device. And power switching devices die in very specific, very predictable ways.

The 11 real causes of ECU failure

1. Water damage

Water is the number one killer, full stop. Modules mounted under the cowl, behind the kick panel, under a seat, or in the engine bay near the firewall are all exposed to leaking windscreen seals, blocked scuttle drains, clogged sunroof drains and flooded footwells.

The nasty part? Water damage often doesn’t cause instant death. The car runs fine for weeks, then throws random misfires, phantom sensor codes, or intermittent no-starts on damp mornings. Pull the module, crack the case, and you’ll find white or green crust spreading across the board.

Tell-tale signs: damp carpet, musty smell, faults that come and go with the weather.

2. Corrosion

Corrosion is water damage’s slow cousin. Moisture and road salt creep into the connector, oxidise the pin plating, and quietly raise the resistance of a circuit that’s supposed to be near-zero ohms.

Higher resistance means voltage drop. Voltage drop means the ECU sees a sensor signal that’s slightly wrong — or a ground reference that’s floating. The computer isn’t broken; it’s being lied to.

Corrosion also eats copper traces inside the module once moisture gets past the conformal coating. On boards that have sat wet for months, traces can be completely gone.

3. Overvoltage

This one’s brutal and it’s usually self-inflicted. Common causes:

• A failing alternator with a bad regulator pushing 17–18 volts
• Jump-starting from a 24V truck system
• Arc welding on the chassis without disconnecting the battery
• Load dump — disconnecting a battery terminal while the engine is running

Modern control modules are designed for roughly 9–16 volts. Push a sustained 18V through them and the internal 5V regulator, the driver transistors, or both, will let go. Overvoltage damage is usually permanent and often visible — scorched components, lifted pads, burnt smell.

4. Reverse polarity

Connect the battery backwards, even for a second, and current flows through paths that were never designed to carry it. Protection diodes clamp hard, get red hot, and either sacrifice themselves or take the surrounding circuitry with them.

Reverse polarity commonly happens during battery replacement, DIY audio installs, or when someone hooks up a booster pack in a rush. If a car went dead immediately after a battery job, this is your first suspect — and it usually damages more than one module.

5. Shorted actuators

Here’s the one that catches out even experienced techs. The ECU switches injectors, ignition coils, idle valves, EGR solenoids, turbo actuators, VVT solenoids and fuel pump relays. Every one of those is a coil of wire.

When a coil shorts internally, or its wire rubs through against the block and touches chassis, the ECU’s driver transistor tries to supply far more current than it’s rated for. The driver dies.

And this is the critical bit: if you fit a brand-new module without finding and fixing the shorted actuator, the new module dies too. I’ve seen people go through three ECUs before checking a single injector’s resistance. Test your actuators. Every time.

6. Failed voltage regulators

Inside every module is a small power supply that steps 12V down to a clean 5V (and often 3.3V) for the microcontroller and sensor references. When that regulator fails, symptoms are wild and confusing: the module may boot and immediately reset, communicate intermittently, or throw a dozen unrelated sensor codes at once because the 5V reference has sagged.

Regulators fail from heat, from age, and — very often — from a shorted 5V sensor reference wire in the harness dragging the supply to ground.

7. Damaged MOSFET drivers

MOSFETs and IGBTs are the output switches doing the actual heavy lifting. They fail from overcurrent (see: shorted actuators), from heat, and from voltage spikes generated by inductive loads that weren’t properly clamped.

Classic symptom: one injector or one coil dead, everything else perfect. That’s not a random glitch — that’s one specific driver channel that’s gone open or shorted. On many modules this is a genuinely repairable fault, which is why sending a unit to a specialist beats buying new.

8. Cracked solder joints

Cars vibrate. Engine bays swing from -20°C to +100°C. Solder joints expand, contract, and eventually fatigue — especially on heavy components like connectors, relays, capacitors and transformer coils.

A cracked joint gives you the most maddening fault in automotive electronics: the intermittent. Runs perfect for a week, dies at a roundabout, restarts fine, no stored codes. Tap the module with a screwdriver handle and it cuts out. That’s a cold solder joint, and it’s often a 15-minute fix under a microscope.

9. Broken PCB traces

Traces break from corrosion, from physical impact (dropped modules, collision damage), or from overcurrent literally vaporising the copper. Sometimes the break is hairline and invisible under conformal coating.

A broken trace behaves exactly like an internal open circuit — the module powers up, communicates fine, but one function is simply gone. Repairable with a jumper wire by someone who knows what they’re doing, and completely undiagnosable from the driver’s seat.

10. CAN-bus communication faults

Modern cars are networks. Modules talk to each other over CAN, LIN, FlexRay or Ethernet. When a scan tool says “no communication with ECM,” most people assume the ECM is dead.

Usually it isn’t. The real culprits:

• A shorted CAN-High or CAN-Low wire
• A missing or failed 120-ohm terminating resistor
• One faulty module on the bus dragging the whole network down
• A corroded gateway module or damaged connector pin
• Loss of power or ground to the module — it can’t talk if it isn’t awake

A quick resistance check across CAN-H and CAN-L with the ignition off should read around 60 ohms. If you see 120, you’ve lost a terminating resistor. If you see near zero, you’ve got a short. Neither means your ECU is bad.

11. Faulty connectors and power supplies

The unglamorous one that solves the most cases. Bent pins, backed-out terminals, spread female contacts, cracked connector housings, chewed wiring (rodents love modern soy-based insulation), corroded ground straps, failing relays and blown fuses.

A ground strap with 0.8 volts of drop across it will make a perfectly healthy ECU behave like it’s possessed. Fix the ground, fix the car.

Before you condemn an ECU: the checklist

This is the part I’d tattoo on every workshop wall. Never replace a control module until all of this checks out:

Power supply

• Battery voltage at rest (12.4–12.7V) and while cranking (above 9.6V)
• Charging voltage between 13.5–14.7V, with no spikes
• All module power feeds present at the connector, under load, with the key in the right position
• Fuses tested for voltage on both sides, not just visually inspected

Grounds

• Voltage-drop test each ground circuit — under 0.1V is what you want
• Inspect and clean chassis ground points; they corrode where you can’t see them

Wiring and connectors

• Unplug and inspect every pin for corrosion, spread, push-out and green crust
• Wiggle-test the harness while monitoring live data
• Check for chafing at every point the loom crosses metal or moves with the engine

Actuators

• Measure resistance on every injector, coil and solenoid the module drives
• Check each output circuit for shorts to ground and shorts to power
• Compare readings against spec — not against “feels about right”

Communication

• 60 ohms across CAN-H and CAN-L, key off
• Scope the bus if you can; look for clean, square, mirrored waveforms
• Try communicating with other modules to see whether the network or just one node is down

Only then should the module itself be on trial. And even then — a specialist repair or a tested remanufactured unit is usually cheaper, faster and less hassle than a new dealer module that needs programming and immobiliser matching.

One more thing about programming

Fitting a replacement ECU is rarely plug-and-play anymore. Most modern modules need VIN writing, immobiliser/key synchronisation, variant coding and sometimes online security access from the manufacturer. Budget for that before you order the part, not after.