Subaru P0302 Appeared Only Warm at Idle: Valve Clearance Beat Three New Parts
P0302 Only Warm At Idle? Check Clearance
The Subaru’s computer counted cylinder-two misfires, but the driver could barely feel one. The warning appeared only with the engine fully warm at low road speed or idle. Tailpipe readings were clean, the engine pulled normally, and raising speed by only about 150 rpm stopped the counter. Another shop had already fitted a spark plug, ignition coil and direct injector to cylinder two. Three plausible parts had failed to answer one mechanical question: was that cylinder breathing and sealing like the others at low speed?
A removed-plug comparison can still be a sensible early branch when ignition evidence is incomplete. The AUTOOL SPT360 can compare several plugs across an adjustable simulated-speed range. SPT360 can compare up to five removed spark plugs across an adjustable simulated-speed range, but a bench spark result cannot reproduce cylinder pressure, prove coil or injector control, measure compression, or identify zero valve clearance. In this case, the new plug history and the condition-specific evidence meant the investigation needed to move beyond the ignition bench.
What makes this case valuable is the disagreement between human senses, emissions and the PCM’s crank-speed calculation. Instead of dismissing the counter, the technician reproduced its exact warm-idle window and used relative compression plus intake pulsation to find a valve that was only marginally failing to seat.
Quick answer: A warm-idle P0302 survived a plug, coil and injector. Relative compression, intake pulsation and a zero-clearance inlet valve made the mechanical cause visible.

In this guide
- The PCM saw something the driver could not feel
- Freeze frame defined the narrow test window
- Do not reward three failed replacements with a fourth
- Use a spark bench only for the question it can answer
- Let relative compression reopen the mechanical lane
- Follow pulsation to the intake stroke
- Measure clearance only after the evidence earns teardown
- Return to the same warm idle for verification
The PCM saw something the driver could not feel
Freeze-frame data put numbers around the complaint: coolant near 88°C, speed about 1 mph, engine speed around 865 rpm and 44 counts on cylinder two. In the bay, the count returned near the normal 650-rpm idle but fell to zero with a small speed increase. That narrow window is more valuable than the P0302 label alone. If a shop tests the car cold or holds it at 2,000 rpm, it may truthfully report no fault while missing the only state that matters.
Freeze frame defined the narrow test window
The four-gas result made the case stranger rather than closing it. Hydrocarbon, carbon monoxide, oxygen and lambda readings remained normal while the data list continued to count misses. A cylinder cut test produced a similar speed drop on every cylinder. These results argued against a large, continuous combustion failure, but they did not disprove a small crank-speed irregularity at idle. The PCM detects misfire indirectly by watching crankshaft acceleration, so it can notice a marginal mechanical difference before a driver or tailpipe test does.
Do not reward three failed replacements with a fourth
Cylinder numbering also had to be proven because available information disagreed about which location Subaru called number two. Commanded cylinder cut combined with non-intrusive ignition and injector pickup confirmed that the already-serviced left-front cylinder was the one the PCM identified. That prevented an embarrassing teardown or another parts swap on the wrong hole. When diagrams and databases disagree, synchronize the commanded event to the physical cylinder rather than choosing the source you prefer.
Use a spark bench only for the question it can answer
A non-intrusive relative-compression test finally showed one weak contribution—close to a 50-percent relative reduction at cranking speed. Synchronizing the result tied it to cylinder two. The size looked dramatic compared with the mild running symptom, but cranking speed was only about 200 rpm, exactly where the marginal seal had the greatest effect. At higher speed, gas momentum and pressure could disguise the leak enough for the misfire count to stop.
Let relative compression reopen the mechanical lane
The next question was where compression was being lost. Exhaust pulsations did not show an excessive event, and crankcase pulses stayed even, reducing concern about the exhaust valve, piston and rings. Intake-manifold pulsations told a different story: every fourth intake event was weak and aligned with cylinder two. In-cylinder comparison then showed lower peak compression—about 9.77 bar versus 12.75 bar on a reference cylinder—and altered intake turbulence. The evidence now justified mechanical access.
| Finding | What it ruled in or out |
|---|---|
| Count only at warm idle | Defined the only valid verification window |
| Good emissions and equal cylinder-cut response | No large continuous combustion failure shown |
| Weak relative compression on cylinder two | Reopened the mechanical path |
| Weak intake pulse, even crankcase/exhaust patterns | Focused attention on the inlet-valve event |
| 0.03 mm and 0.00 mm inlet clearance | Confirmed the mechanical cause |
Follow pulsation to the intake stroke
With the engine removed and opened as required for this layout, the two inlet clearances on cylinder two measured about 0.03 mm and 0.00 mm. Near-zero clearance kept the valve marginally off its seat as the engine warmed. This explained why a new coil, plug and injector could never fix the code and why the problem appeared at warm idle. The case did not infer clearance from mileage or a forum list; it measured the mechanical condition on the cylinder already isolated by several non-intrusive tests.
Measure clearance only after the evidence earns teardown
The engine was reassembled with inlet clearance set to 0.12 mm in the documented repair. Verification returned to the same warm idle that had reliably increased the counter. Cylinder-two counts stayed at zero, P0302 did not return and the engine retained normal behavior away from idle. A cold start alone would not have been sufficient because cold operation had never been the complaint.
Return to the same warm idle for verification
This does not make valve clearance the universal answer for Subaru P0302. Ignition, fueling, air leakage, compression and control faults remain legitimate possibilities. It does show when to change lanes: if the correct cylinder has already received verified ignition and fueling attention, emissions are clean, the counter is confined to low warm speed and relative compression differs, do not keep buying combustion parts. Ask for cylinder identification, a mechanical comparison and a same-condition retest.
The most human lesson is simple: a car does not have to shake violently for a misfire code to contain useful evidence. This Subaru was warning about a small rotational inequality in a very narrow state. By respecting that state—and refusing to let three new parts become proof—the repair reached a valve clearance that no generic code description could name. Verify the fix only after the engine is fully warm in the original idle window.








Comments 0
Questions, fixes, and real-world diagnostic notes from readers.
Be the first to share a diagnostic result or ask a follow-up question.