How citations work on this page: Every superscript number (e.g., 15) links to the Primary Source Directory at the bottom of this page, where you'll find the direct URL to the official NHTSA technical service bulletin, manufacturer datasheet, or engineering standard behind the claim.
How a Coil and a Plug Work as One Circuit
An ignition coil is an autotransformer — a device that uses a collapsing magnetic field to convert your battery's 12 volts into the tens of thousands of volts needed to fire a spark plug.17 Inside, a primary winding of roughly 200 to 400 turns of heavy copper wire wraps around a soft-iron core, and a secondary winding of 10,000 to 20,000 turns of hair-thin wire wraps around that. When the engine computer cuts current to the primary winding, the magnetic field around the core collapses instantly, and that collapse induces a massive voltage spike in the secondary winding — routed straight down to the spark plug.17
Here is the detail that matters for this entire question: the coil doesn't fire at one fixed voltage every time.It fires on demand, generating exactly as much voltage as it takes to force a spark across whatever gap the plug presents at that moment. A plug with a fresh, narrow gap in a low-compression idle might only need 8,000 to 15,000 volts. The same coil, faced with a worn plug's wider gap under load, might have to produce 25,000 volts or more just to bridge it.14 The coil and the plug are not two separate parts doing two separate jobs — they are opposite ends of a single electrical circuit, and the condition of one directly sets the workload of the other.
Why a Worn Plug Quietly Destroys a Coil
Every time a spark plug fires, the high-voltage arc vaporizes a microscopic amount of metal off the electrode tips — a process called evaporation erosion. Over tens of thousands of miles, that erosion physically widens the gap between the center and ground electrodes.14A wider gap needs more voltage to jump, by the same physics (Paschen's Law) that governs any electrical breakdown across a gas-filled gap: breakdown voltage rises with gap distance and pressure.18 So as the plug ages, the coil is forced to work harder on every single firing event, with no warning light telling you it's happening.
That rising voltage translates directly into heat inside the coil. Pushing peak voltage through the secondary winding on every cycle — instead of the modest voltage a fresh plug would require — generates sustained internal heat that NGK's own engineering literature describes as a “heatwave” effect eating away at the coil's internal insulation.12Given enough time, the epoxy potting compound that insulates the coil's ultra-fine secondary windings cracks and delaminates, the internal enamel coating on the wire melts, and the coil starts arcing internally instead of pushing voltage down to the plug where it belongs.11 That internal short is what actually kills most coils — and a badly worn plug is very often what put it there.
The trap: if a worn plug is what killed your old coil, bolting in a brand-new coil without touching the plug does not fix the underlying problem — it just restarts the same clock. The new coil inherits the exact same oversized voltage demand that destroyed its predecessor, and it is now racing toward the same failure.
If your plugs are approaching or past their manufacturer-recommended interval, that context matters here as much as anything in this article — see our companion guide on how often spark plugs actually need to be replaced by electrode metal and engine type before deciding whether a lone coil swap makes sense.
The Two Questions That Decide the Answer
Strip away the brand-specific paperwork and every OEM protocol in this article reduces to the same two checks, performed in this order:
| Question | If Yes | If No |
|---|---|---|
| 1. Is there any carbon tracking visible on the plug's ceramic insulator? | Both parts must be replaced. No exception, regardless of manufacturer. | Continue to question 2. |
| 2. Does testing confirm the coil failed on its own (bad resistance reading, swapping it to another cylinder moves the misfire with it)? | The coil can be replaced alone if the plug is otherwise within spec and gap. | The plug is the likely root cause. Replace both, or the new coil is at risk. |
Question 1 is a five-second visual check with no tools required, and it overrides everything else — which is why it comes first. Question 2 requires actual diagnostic testing, covered in detail further down this page.
Carbon Tracking: The Failure That Forces Both
A spark plug's ceramic insulator is built with visible ribs — small corrugations that lengthen the surface path an electrical arc would have to travel to escape down the outside of the plug rather than jumping the electrode gap inside the cylinder.23 Normally, that design works. But if the internal gap becomes too wide, or thin high-altitude air outside the cylinder offers less natural insulation than expected, the spark can find it easier to arc down the exterior of the ceramic instead of crossing the gap.23
When that happens, the arc burns straight through the rubber ignition boot and bakes a thin layer of conductive carbon dust directly into the microscopic pores of the ceramic ribs — leaving a permanent black hairline scar running down the white insulator.23 Carbon conducts electricity, and once that track exists, every future spark takes the same low-resistance path down the outside of the plug instead of firing across the gap, producing a hard, repeatable misfire under load.
This damage cannot be cleaned or salvaged.The carbon is burned into the ceramic itself, not sitting on its surface. General Motors' own technical bulletins for carbon-tracking failures on the LTG 2.0L and related engines instruct technicians that if tracking is found on any single spark plug, all spark plugs and all ignition coils in the engine must be replaced together.1,2 Reusing a new coil boot over an old carbon-tracked plug just gives the arc a fresh path to scorch — destroying the new coil within miles.
GM traced part of this failure pattern to something outside the plug entirely: older ignition coil boots were dusted internally with talc powder to keep the rubber from sticking to the hot ceramic, but talc has far lower insulating strength than proper dielectric grease — which let arcing start in the first place on otherwise healthy components.1 A separate GM service update goes further, tying unresolved carbon-tracking misfires to secondary engine damage severe enough to require piston and cylinder-wall replacement if the condition runs long enough before repair.3
What the Automakers Themselves Require
Beyond the universal carbon-tracking rule, individual manufacturers have issued their own technical service bulletins covering specific failure patterns on specific engines. If your vehicle matches one of these, it settles the question for you.
| Automaker / Condition | Mandated Action |
|---|---|
| General Motors — isolated misfire, no carbon tracking present | Swap the suspect coil to a different cylinder first. If the misfire follows the coil, replace that coil alone.4 |
| Honda — water intrusion via hood scoop into the No. 4 spark plug tube (P0304) | Replace the coil, replace the corroded plug, and install a redesigned cylinder head cover that diverts water away from the tube.5 |
| Honda — piston ring rotation causing oil-fouled plugs (P0301–P0304) | Remove the coils, replace the fouled spark plugs, and update the powertrain control module software.6 |
| Toyota / Lexus hybrids — carbon-buildup misfires (P0301–P0306) | Remove coil and plug, borescope the cylinder for carbon, run a Top Engine Cleaner treatment, flash the ECM, then replace the coil.7 |
| BMW N51/N52/N52K engines — first coil failure at initial service visit | Replace all coils with an updated supplier part across the engine. Spark plugs are evaluated separately and only replaced if worn or due by mileage.8 |
| Volkswagen — specific spark plug production-date batches | Check the part number and date code stamped on the plug. If it matches the recalled batch, replace all four and torque to 22 Nm; if not, the original plugs can be reinstalled.9 |
Notice the BMW case in particular: it's the clearest documented example of an automaker treating coil and plug service as genuinely independent decisions, evaluated on separate criteria rather than automatically bundled together.8 That is the exception, not the rule — most of the other bulletins bundle the two components together precisely because the underlying failure (water, fouling, carbon buildup) damaged both parts simultaneously.
Dielectric Grease: Why the Wrong Kind Ruins the Repair
Whenever a coil boot is reinstalled — whether you're replacing the coil alone or both parts together — a thin coat of dielectric grease inside the rubber boot is what actually prevents the next round of arcing and moisture intrusion. Dielectric grease is a non-conductive, silicone-based compound engineered to seal the boot against moisture and stop it from vulcanizing (fusing) to the hot ceramic insulator over time.11
GM specifically mandates Molykote G-5008, a silicone-and-PTFE grease with a dielectric strength rated at 350 volts per mil and a volume resistivity holding steady even at 177°C engine-bay temperatures — properties verified on its own manufacturer datasheet.10GM's bulletin is explicit that technicians must apply this specific grease to a depth of about 15mm inside the boot, and it separately warns against using clear or white dielectric greases in its place.1
Two mistakes worth knowing before you touch a boot:silicone-based grease can swell and soften certain silicone-rubber boot materials on some current coil-on-plug designs — check your specific vehicle's service documentation before assuming a universal dielectric grease is safe.21 And dielectric grease is an insulator by design — it belongs only on the inner rubber and ceramic surfaces, never smeared onto the metal terminal contacts themselves, where it would block the very current path it's supposed to protect.
How a Technician Confirms Which Part Failed
Beyond the visual carbon-tracking check, ASE's Engine Performance certification standards define two concrete tests that separate an organically failed coil from one that was destroyed by a bad plug.15
Resistance Testing
A digital multimeter measures the coil's primary winding resistance (typically 0.2 to 5.0 ohms) and its secondary winding resistance (typically 5,000 to 25,000 ohms).19A reading of infinite resistance (“OL” on the display) means the fine internal wire has physically broken — consistent with the coil failing on its own from age or a manufacturing defect. A reading well below spec points to an internal short from melted insulation, which is exactly the failure mode that a plug forcing sustained peak voltage would cause.11
Oscilloscope Waveform Analysis
A more definitive test connects a digital oscilloscope to the secondary ignition circuit and graphs voltage over the course of a single spark event.19The height of the initial spike — the “firing line” — shows exactly how much voltage was needed to bridge the plug's gap at that instant. A normal firing line at idle runs roughly 7,000 to 12,000 volts; a firing line spiking to 25,000 volts or higher points directly at excessive resistance in the plug itself, most often a widened gap.20Because the coil's total stored energy is fixed, an abnormally high firing voltage also shortens the remaining spark duration — a second waveform signature confirming the plug, not the coil, is the root cause.20
Put simply: if the waveform shows a normal firing voltage but the coil's own resistance test fails, the coil failed on its own and can be replaced alone. If the waveform shows an abnormally high firing voltage, the plug is the problem no matter what a new coil's resistance test shows — because that new coil is about to be subjected to the same excessive load.
What Happens If You Guess Wrong
Replacing a coil alone when a worn plug was the actual cause doesn't just risk a repeat visit to the parts counter — it risks a much more expensive repair further down the exhaust stream. When a cylinder misfires, its raw, unburned air-fuel mixture gets pumped straight into the exhaust rather than being consumed in the combustion chamber. That fuel ignites inside the catalytic converter, spiking internal temperatures well past its normal 1,200–1,600°F operating range and melting the precious-metal-coated honeycomb substrate designed to scrub exhaust emissions.
A destroyed catalytic converter is a four-figure repair on many vehicles — and it will also fail an emissions inspection outright, a topic our guide to passing emissions with the check engine light on covers in more depth. A sustained misfire also has a second, purely mechanical consequence: unburned fuel condensing on the cylinder wall washes away the thin film of engine oil that protects the piston rings, letting metal scrape directly against metal and accelerating irreversible cylinder wear.
If a repeat misfire is already showing up as a rough idle or a rhythmic shake under load, that symptom pattern is covered separately in our guide to why your car shakes when you drive. The point of diagnosing the root cause correctly the first time isn't just about avoiding a second parts purchase — it's about avoiding a five-minute plug problem turning into a converter or engine-internal repair.
Quick Reference: Coil Only vs. Both
| Situation | Recommended Action |
|---|---|
| Visible carbon tracking on the ceramic insulator | Replace both, always. Not salvageable — the carbon is burned into the ceramic itself.1,2 |
| Coil resistance test fails; plug is new or well within its interval, no gap or fouling issues | Replace the coil alone. The plug is not the cause. |
| Coil failed; plug is near or past its manufacturer-recommended replacement interval | Replace both. A worn plug's rising voltage demand is the most common reason a coil fails prematurely. |
| Oscilloscope shows an abnormally high firing voltage (25kV+) at idle | Replace both. High firing voltage means the plug's gap or resistance is the root problem, not the coil.20 |
| Known TSB condition (water intrusion, oil fouling, recalled plug batch) | Follow the specific manufacturer bulletin for your engine — see the table above. |
| Coil swapped to another cylinder and the misfire follows it, no carbon tracking present | Replace the coil alone — this is GM's own documented isolated-failure test.4 |
Frequently Asked Questions
Will a new ignition coil fix a misfire if the spark plug is bad?
No. If the spark plug's widened gap or fouled insulator is what's forcing the excessive voltage demand, a new coil will briefly restore normal firing and then begin the same accelerated wear cycle that destroyed the previous coil — often failing again within a few thousand miles.
Can I just replace one ignition coil instead of all of them?
Yes, if diagnostic testing (a resistance test or a cylinder-to-cylinder coil swap) confirms the failure is isolated to that one coil and there's no carbon tracking on any plug. If carbon tracking is present on even one cylinder, GM's own bulletins call for replacing every coil and every plug in the engine together, since the underlying arcing condition likely affects the whole set.1
How do I know if carbon tracking is present without a professional inspection?
Pull the plug and look for a black or dark gray hairline mark running lengthwise down the white ceramic insulator, typically starting near the top where the boot seats. A light tan or brown haze near the metal shell is normal “corona stain” from dust and oil attracted by the high voltage field and does not require replacement — the distinction is that corona stain is a diffuse discoloration, while carbon tracking is a distinct, continuous line.
Does using dielectric grease matter if I only replace the coil?
Yes. A fresh coil boot placed over a dry, aged ceramic insulator without dielectric grease is more likely to stick (vulcanize) to the ceramic over time and to let moisture creep in around the seal — both of which can eventually trigger the same arcing and carbon-tracking failure this article describes, even on an otherwise good plug.11
Is it normal for a new ignition coil to fail again quickly?
It is not normal, but it is a well-documented pattern when the original root cause — most often a worn or carbon-tracked spark plug — was never addressed. A coil failing again within a few thousand miles of replacement is a strong signal to inspect the spark plug on that cylinder before buying a third coil.
Do all four (or six, or eight) coils need replacing at once?
Not automatically. Outside of a documented carbon-tracking or supplier-wide TSB condition, most manufacturers treat each coil as an independent component and only require replacing the ones that actually test bad or show physical damage.