Intermittent Misfire Under Load on a Coil-On-Plug V8: A Flawed Swap Test

The swap was methodologically sloppy but, in this particular case, still informative - because the misfire did not move.

The logic:

• The classic rule is move one part at a time. When you move two components simultaneously, a misfire that stays on the original cylinder is still a valid exclusion of both parts (neither one caused it). A misfire that moves would be ambiguous - you would not know whether the coil or the plug was responsible.
• Here the misfire stayed on cylinder 7 at 380 counts, while cylinders 3 and 5 - now carrying plug 7 and coil 7 respectively - both read 0. So coil 7 and plug 7 are both functionally good, and the misfire stayed with the location, not the parts.
• That shifts suspicion to anything still physically attached to cylinder 7: the injector, the injector driver circuit, the intake port seal at that runner, valve train, or a mechanical fault.
• The black, fuel-wet plug is a consequence, not a cause. A cylinder that misfires under load gets fuel but no clean burn, so the plug soots. Its gap is correct, which further removes it from suspicion.

Lesson to state plainly: if the misfire had moved, this test would have been worthless and the technician would have had to start over. One component at a time, one road test at a time, one cylinder's counters at a time.

Fuel trims near zero (STFT -3, LTFT +2) largely exclude a grossly rich or lean mixture as the cause of this misfire. A leaking or plugged injector on one cylinder would typically show up as a bank-level trim shift; +2/-3 percent says the bank is metering correctly overall.

Load dependence is the strongest clue. A misfire that appears at 68 percent calculated load and 2,850 rpm but not at idle is a demand problem: high cylinder pressure makes it much harder to jump the gap and much less forgiving of a weak injector, a poor connection, or a marginal spark path.

Test sequence:

1. Move injector 7 to a different cylinder (one component, one road test). If the misfire moves, the injector is at fault. If it stays, the injector is excluded.
2. Check the injector circuit - resistance compared to a known cylinder, and a current-ramp or noid-light pattern while the harness is flexed. A chafed harness that opens under engine torque/load is exactly this symptom and will not show up at idle in the bay.
3. Perform an injector balance/power-balance test to quantify each cylinder's contribution rather than relying on counters alone.
4. Inspect the intake runner and port gasket at cylinder 7 for a leak that only opens as the manifold flexes under load - this can create a lean misfire on one cylinder without moving bank-level trims enough to notice.
5. Recheck ignition secondary on cylinder 7 with the coil reinstalled, looking for carbon tracking at the boot or a corroded spring, since the coil tested good in a different hole but the interface stayed behind.

Do not sell eight coils. The data excludes coil failure.

The self-clearing lamp is not evidence of repair - it is evidence that the fault is intermittent.

Explain the mechanism:

• OBD-II misfire logic runs continuously. A misfire that exceeds the catalyst-damaging threshold sets a flashing MIL in real time. When the misfire stops exceeding that threshold, the lamp reverts to steady, and after a number of warm-up cycles without a recurrence, the ECM clears the code on its own.
• So the sequence the customer experienced - flash, then off - maps precisely onto an intermittent load-dependent fault. The condition that triggered it (68 percent load at 2,850 rpm during a highway merge) simply has not recurred yet.

What the freeze frame adds:

• It captures the exact operating window where the fault lives: 2,850 rpm, 47 mph, 68 percent load, engine at full operating temperature. That is a reproducible road-test target. Duplicating that condition is how you prove a repair, not idling in the bay.
• Coolant at 197°F rules out a cold-start or open-loop contributor, and steady trims rule out a mixture problem, sharpening the diagnosis to a demand-related ignition or injection issue on one cylinder.

Recommendation: the vehicle should be diagnosed now, not later. Each flashing event pushes unburned charge into the converter, and a converter is far more expensive than the injector or harness repair likely needed here. Document the misfire counts in writing so the customer understands the fault is still present, and re-run the load test after repair at the same rpm and load to confirm counts stay at zero.