How to Test a Crankshaft Sensor: 7 Checks That Stop the Part-Swapping

Let me start with a sentence I repeat every time I see a “replace crankshaft sensor” estimate: a code doesn’t equal a part. In July 2024, a shop called me about a delivery van that still wouldn’t start after two new crank sensors. The original diagnosis was P0335. The real problem was a corroded terminal in the connector. The sensor was never dead. The terminal took about ten minutes to fix after the second sensor replacement took almost an hour.

I coordinate rush parts and urgent repairs for automotive facilities, so I’ve watched this pattern happen more often than I’d like. Low on time, order the part, install the part, then find out it was never the part. That’s why learning how to test a crankshaft sensor is about more than knowing where to put a multimeter. It’s about knowing what to check before you commit to a repair.

When This Checklist Works

Use this checklist when the engine cranks normally but doesn’t start, or when it stalls and the scan tool points to the crankshaft position sensor circuit. It also applies if you’re trying to figure out why there’s no spark and no injector pulse.

If the engine doesn’t crank at all, stop here. Check the battery, starter, engine grounds, and anti-theft system first. That’s a different problem and a different checklist.

How to Test a Crankshaft Sensor: 7 Checks in Order

1. Check Battery and Engine Grounds First

Crank sensor signals are small. A weak battery, a loose engine ground strap, or a bad ground at the cylinder head can make a good sensor look dead. More than once I’ve seen a P0335 code set because the engine ground was loose. The fix was a clean, tight ground, not a sensor.

While you’re at it, watch cranking voltage with a multimeter or the scan tool. If it drops much below 10 volts during cranking, even a healthy Hall-effect sensor may not switch correctly. Charge the battery and recheck before you blame the part.

2. Read Live Data, Not Just the Fault Code

The code tells you which circuit is suspect. It doesn’t tell you which part is bad. Use a scan tool and go to live data while the engine is cranking.

Watch the engine RPM PID. If it stays at zero, you’re in the right area: sensor, wiring, power supply, ground, or tone ring. If RPM shows a value and the engine still won’t start, the crank sensor may be fine. Start looking at fuel pressure, spark, compression, and timing.

Also check the freeze-frame data. Was the code set during a hot restart? During cold cranking? While driving down the road? Those details change your next step.

3. Identify the Sensor Type

The fastest way to sort this out is to count the pins and pull up a wiring diagram. A two-wire sensor is often a magnetic sensor, also called a variable reluctance sensor. It needs no supply voltage. It generates its own small AC voltage as the tone wheel passes.

A three-wire sensor is often a Hall-effect sensor. It gets a reference voltage and ground, then sends a square-wave signal back to the ECU. Some active sensors only have two wires and use current as the signal, but they’re less common. Don’t assume two wires always means magnetic. Verify with the service schematic.

4. Inspect the Wiring and Connector Before Testing

Corrosion, pushed-out pins, oil in the connector, and chafed wires cause a large percentage of crankshaft sensor faults. If the connector terminal is loose, the sensor might test fine on the bench but still fail as soon as the engine rocks on its mounts.

Back-probe carefully or use a terminal test lead. Don’t stab the wire with a test light. A test light can load the circuit and damage electronics. You want to see the signal the way the ECU sees it, not force extra current through the wire.

5. Measure the Signal While Cranking

For a magnetic sensor, disconnect the connector and measure resistance across the two sensor-side pins only if the service information gives you a value. Resistance alone doesn’t prove the sensor works, but an open winding or a short to ground is a quick red flag.

Then reconnect the sensor, back-probe the two signal wires, set the multimeter to AC volts or AC millivolts, and crank the engine for about five to ten seconds. You’re not looking for one universal number. You’re looking for a clean voltage that increases with cranking speed and drops when you stop. If it stays at zero, the sensor isn’t generating a signal.

For a Hall-effect sensor, check the reference voltage first. Back-probe the signal wire, turn the key on, and crank the engine. The signal should switch between low and high—often around 0 to 5 volts. A digital multimeter will usually average that out to roughly 2.5 volts if the sensor is switching. If it stays pegged low or high, you may have no signal, no power, no ground, or a sensor that isn’t sensing the tone wheel.

An oscilloscope is the better tool for this step because it shows the actual waveform shape. A bad sensor or tone ring may still produce a voltage, but the waveform can be flat, distorted, or missing teeth.

6. Verify the Tone Ring, Air Gap, and Bracket

A brand-new sensor will not fix a damaged tone ring. Inspect the reluctor or trigger wheel on the crankshaft, flexplate, flywheel, or pulley. Look for missing teeth, bent teeth, grease buildup, rust chunks, and cracks. If the ring is behind a cover, use a borescope when possible.

In my world, this is also where metal quality matters. Stamped tone rings and sensor brackets are not cosmetic parts. If a stamped ring is out of spec because the tooling wore down, the sensor might never see the right tooth profile. The result is an intermittent no-start that makes no sense on the scan tool.

For adjustable sensors, check the air gap against the vehicle specification. For non-adjustable sensors, make sure the bracket isn’t bent and the mounting surface isn’t damaged.

7. Torque Everything, Especially When the Part Is “Known Good”

Torque specs are not the exciting part of this job, but they cause comebacks. A sensor bracket that’s too loose can shift with engine vibration. A bolt that’s too tight can crack the bracket or distort the sensor housing. Use the service manual value for the crankshaft sensor mounting bolt.

The same lesson applies when a wheel bearing is on the same work order. If you search for Timken wheel bearing torque specs and find one number on a random forum, slow down. Timken doesn’t set the axle nut torque. The vehicle manufacturer does. The Timken contact page can help confirm part fitment and give installation recommendations, but the nut torque for your specific vehicle has to come from the OE service data or an equally reliable source.

I’ve seen more hub damage from guessing torque than from using a cheap bearing. It’s the same logic as a crank sensor bolt: right part, wrong installation torque, early failure.

Two Things That Look Like Crank Sensor Problems But Aren’t

Fuel Delivery Can Fake a No-Start

Some vehicles won’t energize the fuel pump relay until the ECU sees a crankshaft signal. If the crank sensor is dead, the fuel pump may not run even if the pump is fine. On the other hand, a fuel system problem can make you chase the crank sensor if you don’t check fuel pressure early.

If you’re dealing with a known Walbro fuel pump f90000267, for example, don’t assume the pump is bad just because there’s no pressure. Check that the relay is being commanded, check power and ground at the pump while cranking, and confirm the pump itself is the cause before you replace it. A brand-new pump won’t fix no fuel pressure if the ECU never turns the relay on.

Don’t Mix Unrelated Electrical Faults

Here is the boring truth: not every fault is connected. If a customer also has a burned-out Vision car headlight bulb, treat that as a separate circuit. Check the bulb connector, ground, and voltage drop. A headlight issue won’t cause a crankshaft sensor code, but it can distract you from a careful diagnostic process if you let it.

Stay organized. Write down the customer’s complaints in order, test the system that matches each complaint, and don’t let one faulty part convince you that everything else on the car is failing too.

The Bottom Line

The biggest mistake I still see isn’t using the wrong meter setting. It’s replacing the sensor first and unhooking the scan tool before checking live data. When the car comes back on a tow truck, the second attempt is always more expensive.

So test the signal. Test the wiring. Inspect the mechanical side. Then install the new part to the correct torque. If the test points to a crankshaft sensor, replace it with a quality unit and move on. If the test doesn’t point to the sensor, don’t let the code bully you into buying one anyway.