P0387: evidence-led diagnostic file for crankshaft position sensor B low input
P0387 is diagnosed without jumping to one part by validating circuit behaviour, live data, operating conditions and the physical system response together.
Advanced diagnostics and measurement6 sourcesUpdated 2026-07-31
Use the code as a measurement starting point, not a failed-part label. Preserve event data and compare symptoms, likely causes, live data and physical measurements under the same operating condition.
P0387 · DTC
P0387: evidence-led diagnostic file for crankshaft position sensor B low input
IDENTITY
Bind P0387 to the correct control module
The reporting module and failure subtype matter as much as the text.
P0387 base definition: crank sensor B signal amplitude/voltage below expected range; record any OEM subtype separately.
For crankshaft position sensor B low input, record the reporting module, software identity and current/history status together.
The first evidence set in freeze frame is cranking speed, battery voltage, synchronization, engine temperature and stall event.
Before clearing P0387, preserve companion codes and monitor status or the failure context is lost.
MEASUREMENT
Measure the crankshaft position sensor B low input circuit by command and response
A static voltage alone is not a verdict.
First electrical check: sensor supply/ground or, for inductive sensors, resistance, air gap and shielding.
In live data, observe RPM, crank/cam sync, tooth error counter and stall history on the same time base.
Wiggle testing and loaded voltage drop are more useful than unloaded continuity when P0387 is intermittent.
For P0387, when a scope is needed, capture waveform, frequency/duty and reference ground together; a screenshot alone does not condemn a part.
DIFFERENTIAL DIAGNOSIS
Separate look-alike causes for P0387
Electrical, mechanical and software paths require separate evidence.
Priority cause groups: low cranking speed, excessive air gap, damaged trigger wheel, shielding, sensor or ECU input.
Differential test: capture crank waveform with cam reference during crank and hot stall.
Common misdiagnosis: replacing the sensor for no RPM without excluding battery/cranking speed.
When P0387 appears with P0335/P0336, cam-correlation and low-voltage codes, trace the data chain rather than diagnosing from one code.
VERIFICATION
Prove the repair effect on P0387
A cleared code is not proof of repair.
After repair, remeasure cranking waveform amplitude, sync time and hot restart and compare with the pre-repair record.
Do not close P0387 until the same temperature, load and speed conditions are reproduced.
For P0387, record pending/permanent status, readiness monitors and drive-cycle outcome.
For P0387, add part number, software action, measured values and final road test to the service record.
FIELD WORKFLOW
Evidence-preserving diagnostic sequence
Record P0387 and all companion codes before clearing.
Confirm vehicle identity, module software and OEM subtype.
Recreate the freeze-frame conditions: cranking speed, battery voltage, synchronization, engine temperature and stall event.
Perform the loaded circuit check: sensor supply/ground or, for inductive sensors, resistance, air gap and shielding.
Use capture crank waveform with cam reference during crank and hot stall to separate electrical and physical causes.
Confirm normalization of cranking waveform amplitude, sync time and hot restart under the same operating condition.
DIFFERENTIAL DECISION MATRIX
Connect the symptom to evidence, not a guessed part
Do not treat the code label as a parts diagnosis. Narrow root cause through freeze-frame, simultaneous module codes, power/ground, live data and active testing.
Capture freeze-frame, first/last occurrence, load, rpm, temperature, vehicle speed and system voltage.
2. Eliminate shared causes
Rule out battery/charging, power, ground, fuses, network communication and shared-reference faults before replacing parts.
3. Compare command and result
Compare commanded values with real sensor/actuator response under the same operating condition.
4. Prove the repair
Clearing codes is not enough; recreate the monitor condition and verify that code/symptom does not return.
Decision tree
This page is prioritized for deeper evidence-led coverage: reproduce symptom → capture event data → eliminate shared electrical/network causes → perform system-specific measurements → verify under the same condition.
Evidence-led DTC checklist
Capture freeze-frame/event data and a full-module scan before clearing the code.
Separate DTC status bits: active, pending and history/permanent do not carry the same diagnostic weight.
Eliminate supply, grounds, 5 V reference and network health as shared causes.
Graph ECU command against actual sensor/actuator feedback.
Load-test wiring and validate the system physically before replacing parts.
After repair, check readiness/DTC return under the same operating condition.
Verify the code against vehicle identity and system context
The same DTC can lead to different root causes across manufacturers, controllers and operating conditions. Cross-check model, engine/transmission and the measurement chain.
This powertrain context reference is intentionally kept manufacturer-aware. Confirm the exact definition for the vehicle, model year and controller before replacing parts.
1 · Capture the event
Save DTC status, freeze-frame/event data, operating state and companion codes before clearing memory.
2 · Verify identity and power
Confirm the reporting module, supply, grounds, fuses and connector condition before judging a sensor or actuator.
3 · Compare command and feedback
Use live data to compare commanded state with actual feedback under the same operating condition.
4 · Measure the circuit or system
Use the OEM procedure for pin locations and exact thresholds; separate electrical/network faults from mechanical, hydraulic, pneumatic or flow faults.
5 · Reproduce and verify
After repair, reproduce the original load and operating condition and confirm that the code and related symptoms do not return.
This code does not automatically condemn a component; it describes fault behaviour observed by the controller in a monitored system. Preserve event data and eliminate shared causes before narrowing the root cause with system-specific measurement.
Code classPowertrain
Family-based first diagnostic chain
Capture freeze-frame and operating conditions.
Eliminate charging, power, ground and shared-reference faults.
Compare commanded value with sensor/actuator response.
Test wiring/connectors separately from mechanics.
Prove the repair by recreating the condition rather than only clearing the code.