Grumpy Old Geezer Garage Learning Center Banner

Home > Learning Center > Air, Fuel & Combustion > What’s in an Engine Knock?

What’s in an Engine Knock?


Intro

Engine knock isn’t just “some noise.” Old engines can tick, tap, slap, leak, clatter, and rattle like a coffee can full of bad decisions without suffering combustion knock. A lazy lifter, cracked exhaust manifold, loose accessory, worn timing chain, cold piston slap, bad mount, loose tin, or plain old-car nonsense can make a driver stare under the hood like the engine owes him money.

That’s ordinary racket until combustion gets involved. Real knock is a combustion problem, which means the important evidence is not only what the noise sounds like but when it appears, what load the engine is carrying, how hot the engine is, where the timing is, what fuel is in the tank, and what the chamber is doing while the piston is trying to make pressure.

Inside a healthy cylinder, the spark starts a controlled burn and the pressure rise follows the piston through a useful part of the crankshaft rotation. Knock means part of that process has stopped behaving. The burn may still have started normally, but the remaining mixture gets hot and compressed enough to react on its own. Pressure no longer rises as one orderly shove. Sharp pressure waves get added to the event, and the piston, rings, bearings, gasket, valves, plug, chamber, and head all get to absorb the argument.

That distinction matters because knock is not the same thing as simply making high cylinder pressure. A strong engine is supposed to make high pressure. The useful question is whether that pressure develops at the right time and at a rate the parts can use. Plenty of engines make serious pressure without knocking. Trouble starts when heat, pressure, timing, fuel tolerance, mixture behavior, chamber shape, or operating load push part of the charge past its ability to stay under control.

Octane belongs in the discussion because fuel’s resistance to autoignition matters, but octane isn’t the whole story. Better fuel may be exactly what the engine needs. Better fuel may also hide a timing curve that arrives too soon, a hot chamber, a lean cylinder, weak quench, an overheated plug, or an owner lugging the engine like rpm costs extra.

This page is about the mechanism behind the noise: what normal combustion looks like, how detonation differs from pre-ignition, why load and heat expose the problem, what timing and mixture do to the margin, what parts get hurt, and how to diagnose the complaint without throwing fuel, timing, and parts at it like a man trying to put out a grease fire with a parts catalog.

The useful question isn’t, “Did I hear a noise?” The useful question is, “Did combustion stay under control?” Once the answer turns ugly, the cylinder has quit asking politely.

↑ Back to the Top, Before the Cylinder Starts Complaining


Normal Combustion: What’s Supposed to Happen

Before knock makes sense, normal combustion needs a clean baseline. The spark plug fires before top dead center because the mixture does not burn instantly. A small flame kernel starts at the plug gap, grows into a flame front, and moves through the compressed mixture. While that flame is traveling, the piston is still moving, the crankshaft is still turning, and cylinder pressure is changing every degree of crank rotation.

The goal is not to create the earliest possible pressure or the biggest possible bang. The goal is to build pressure so the strongest part of the push arrives after the piston has crossed top dead center and the crank has enough leverage to turn that pressure into torque. Fire too late and useful pressure arrives after the crank has already given away leverage. Fire too early and pressure begins fighting the piston while it is still coming up.

That is why normal combustion is a timed pressure event, not an explosion contest. The flame front burns progressively through the chamber. The already-burned gas expands. The unburned mixture ahead of the flame — the end-gas — gets squeezed and heated by both piston motion and the pressure of the burning charge. In a healthy event, that end-gas stays chemically stable long enough for the flame front to reach it and finish the burn in order.

The end-gas is where the knock argument lives. As compression and combustion raise pressure, the remaining unburned mixture gets hotter. Temperature matters because chemical reactions speed up with heat. Pressure matters because the molecules are being packed tighter while the mixture is already getting hotter. Fuel quality, chamber temperature, mixture strength, residual exhaust gas, and chamber motion all affect how much margin remains before the end-gas decides it has waited long enough.

There is also a time component hiding inside that end-gas. Autoignition is not an on/off temperature switch. The hotter and more compressed the unburned mixture becomes, the shorter its chemical delay before it can react on its own. Normal combustion wins when the flame front reaches that mixture before the delay runs out. Knock begins when pressure, temperature, fuel resistance, and time under those conditions shorten the delay enough that the end-gas lights itself first. That is why a few degrees of timing, a little more load, or a hotter chamber can move an engine from quiet to rattling without changing any hard part.

Normal combustion wins the race when the flame front gets through the chamber before the end-gas autoignites. That is one reason chamber shape, spark-plug location, quench, turbulence, and mixture motion matter. A faster, cleaner burn can require less spark advance and give the end-gas less time to sit hot and compressed. A lazy chamber can force the engine to start the burn earlier just to put pressure in the right place, which can leave the far side of the chamber spending more time under ugly conditions.

Peak cylinder pressure is supposed to occur after top dead center, not on the wrong side of it. The exact crank angle varies with engine design and operating condition, so there is no one magic number to tattoo on every distributor. The teaching point is simpler: pressure should rise progressively, peak where the crank can use it, and then fall as the piston moves down and volume increases.

Worth Knowing

Worth Knowing

High cylinder pressure is not the enemy. Badly timed or violently rising pressure is. A healthy engine can make serious pressure without knock if the burn stays controlled and the pressure arrives where the crank can use it.

The piston, rod, crank, bearings, rings, and gasket are designed around that controlled rise. They still live a hard life, but the force is delivered in a pattern the engine was built to carry. Knock changes the shape of that force. Instead of one useful pressure curve, sharp oscillations are superimposed on it, and the parts start seeing local hammer blows that the average cylinder-pressure number can hide.

A running engine is controlled violence with a schedule. Fuel, air, spark, compression, flame travel, chamber shape, piston motion, and crank angle all have to agree closely enough to turn fire into work. When they agree, the engine pulls cleanly. When the end-gas gets hotter, more compressed, or more impatient than the flame front can tolerate, the schedule starts coming apart.

A good burn shoves the crank. Knock adds a hammer to the shove.

Crank angle changes how useful the pressure can be. Near top dead center the rod and crank have poor leverage even when cylinder pressure is enormous. A little farther down the stroke, leverage improves while expanding cylinder volume starts pulling pressure down. Good combustion has to balance those moving targets across a useful part of the power stroke. That is why chasing one peak-pressure number misses the larger job. The engine needs a pressure curve the crankshaft can use, not one heroic instant that arrives in the wrong place.

↑ Back to the Top, Before the Cylinder Starts Complaining


Knock: When the Burn Gets Ugly

Knock happens when combustion quits behaving after the spark has already started the normal burn. The plug fires. The flame kernel forms. The flame front begins moving through the chamber. Meanwhile the remaining end-gas is being squeezed and heated. If that end-gas crosses its autoignition limit before the normal flame front reaches it, parts of the remaining charge react almost at once instead of waiting their turn.

That sudden reaction creates very rapid local pressure rise. The important difference is not that the fuel suddenly became “more powerful.” The engine already had energy in the mixture. The problem is the rate and location at which that energy gets released. A controlled flame front moves through the charge. Detonation adds spontaneous combustion in the end-gas, producing pressure waves that travel through the chamber and reflect from the piston crown, cylinder walls, and head.

Those waves can ring the chamber at frequencies high enough to create the familiar metallic ping or rattle. What the driver hears is the structure responding to the pressure event. The noise is evidence, not the damage mechanism itself. The chamber was already being hammered before the sound made it through the head, block, firewall, carpet, and whatever exhaust system the owner swears is only a little loud.

Severity changes the picture. Light, brief trace knock under an unusual hot pull is not the same event as heavy sustained detonation. The first may leave little or no visible damage. The second can pound ring lands, erode piston crowns, hammer bearings, disturb the head gasket seal, overheat plugs, and create enough additional heat to make the next combustion event even less tolerant.

That feedback is what makes repeated knock dangerous. Detonation raises local heat. More heat reduces knock margin. Reduced margin makes the next event easier to knock. If the driver keeps the throttle down, the engine can move from an occasional complaint to a self-reinforcing beating where every bad cycle makes the following cycle easier to hurt.

Knock often appears only under load because the cylinder has to be filled and compressed hard enough to create the conditions. The engine may idle perfectly. It may rev cleanly in neutral. It may cruise at light throttle without a sound. None of that proves the chamber has enough margin when the throttle opens, volumetric efficiency rises, cylinder pressure climbs, and heat has had time to soak into the metal.

Knock also does not require every cylinder to misbehave equally. One cylinder may run hotter because of coolant flow, mixture distribution, plug condition, chamber deposits, or manifold layout. One chamber may have a little more compression because of carbon or machining variation. One intake runner may feed differently. A knock problem can begin as one-cylinder trouble while the rest of the engine looks innocent.

That matters during diagnosis because the average numbers can lie. Average air/fuel ratio can look safe while one cylinder is lean. Total timing can look reasonable while the advance curve brings too much in too soon. Coolant temperature can look normal while one exhaust-side chamber area is much hotter than the sender location. The cylinder only lives by the conditions inside that cylinder.

Reality Check

Reality Check

The sound is only the warning. The abnormal pressure event happened inside the cylinder first. A quiet engine under no load has not proved anything about what it will do when cylinder filling, heat, and pressure climb together.

An engine may survive occasional light knock, but survival is not a tuning standard. The useful question is still what combination of load, heat, timing, fuel, and chamber behavior pushed the end-gas over the line.

The pressure waves are extremely fast compared with the ordinary flame front. They bounce through the chamber while the main combustion event is still unfolding, so an average pressure curve can hide violent local oscillations riding on top of it. That is how detonation can punish a piston edge, ring land, gasket area, or plug even when the engine is not making some ridiculous average pressure. The local pressure and heat can be much uglier than a single simplified cylinder-pressure number suggests.

↑ Back to the Top, Before the Cylinder Starts Complaining


Detonation vs. Pre-Ignition

Detonation and pre-ignition get thrown together in garage talk because both live in the same bad part of the combustion neighborhood. They are not the same failure, and the distinction matters because the timing of the abnormal event changes what the piston and chamber have to survive.

Detonation happens after the spark has started the intended burn. The plug fires. The flame front begins moving. Pressure and temperature rise. The remaining end-gas gets compressed and heated until some of it autoignites before the normal flame front reaches it. The resulting pressure waves hammer the chamber even though the original spark event started on schedule.

Pre-ignition starts earlier. The mixture lights before the spark plug is supposed to fire. A glowing carbon deposit, overheated plug tip, sharp chamber edge, hot exhaust valve, overheated piece of metal, or another local hot spot becomes an accidental ignition source. Now combustion can begin while the piston is still traveling upward on the compression stroke.

That is a much uglier argument with the crankshaft. The piston is trying to compress the charge while the burning mixture is already expanding and pushing back. Pressure can rise far too early. Heat climbs rapidly. The piston crown and exhaust-side chamber can get brutally hot. A few severe pre-ignition events can do damage that hundreds of light trace-knock events might never manage.

Pre-ignition and detonation can also feed each other. A detonation problem can raise local metal temperature and damage a spark plug or piston edge until that hot part becomes a pre-ignition source. Pre-ignition can create such extreme pressure and temperature that the remaining charge detonates afterward. By the time the engine is torn down, the evidence may show both problems and make people argue about which one arrived first.

The damage pattern can overlap too. Broken ring lands, peppered piston crowns, melted edges, damaged plugs, hammered bearings, and hurt head gaskets do not carry little labels saying “detonation did this” or “pre-ignition did that.” The operating history matters. Was it rattling under load for weeks? Did one hard pull suddenly kill power? Was a plug overheated? Did the piston crown begin melting near a hot edge? The parts give clues, but the event sequence has to make mechanical sense.

Spark timing changes detonation risk because it changes when the intended pressure rise begins. Spark timing does not directly command pre-ignition, because pre-ignition is being started by some hot source other than the plug. Retarding the distributor may reduce detonation and still do nothing to remove a glowing deposit or overheated plug that is lighting the mixture on its own.

Fuel octane follows the same logic. More knock-resistant fuel can increase resistance to end-gas autoignition and help a detonation problem. It cannot cool a glowing metal edge by decree. If the chamber contains a genuine pre-ignition source, higher octane may not remove the cause, and the owner can burn expensive fuel while the hot spot keeps doing freelance ignition work.

The simplest way to keep the terms straight is to ask who lit the mixture and when. With normal combustion, the plug starts the event on schedule. With detonation, the plug starts normally but the end-gas quits waiting. With pre-ignition, something else starts the burn before the plug gets its turn.

Warning

Warning

Detonation pounds after the spark. Pre-ignition starts the burn before the spark was invited. Pre-ignition can create destructive pressure and heat while the piston is still coming up, so treating both as harmless “pinging” misses the dangerous part of the distinction.

Loose conversation may drag all three under the word “knock.” Diagnosis cannot afford to be that lazy.

There is another timing clue in the failure. Detonation normally needs the intended burn to get far enough along that the end-gas has been squeezed and heated toward autoignition. Pre-ignition does not wait for that sequence. The rogue hot spot starts the fire on its own schedule. Severe pre-ignition can therefore begin punishing the piston before the crank reaches the point where normal combustion pressure would be useful. The engine is not merely burning badly; it is burning at the wrong time for the hardware underneath it.

↑ Back to the Top, Before the Cylinder Starts Complaining


What Knock Sounds and Feels Like

Knock doesn’t always sound the same. Light spark knock may sound like faint pinging under acceleration. A harder case may sound like marbles rattling in a can. Heavy detonation can sound sharper and more metallic, especially when the engine is loaded hard at lower rpm. The cabin, exhaust, insulation, engine mount design, and plain old hearing loss can all change what reaches the driver.

The operating pattern is usually more useful than the exact adjective. Combustion knock tends to follow cylinder load. Climbing a hill, accelerating in high gear, towing, running hot, opening the throttle at low rpm, or making a hard pull after heat soak are classic places for a marginal combination to tell on itself. Lift the throttle and the noise often disappears because cylinder filling and pressure fall immediately.

That load sensitivity helps separate combustion knock from many mechanical noises. A loose rocker or lifter tick may follow engine speed whether the engine is pulling or not. Piston slap is often strongest cold and may fade as the piston warms and clearance changes. An exhaust leak can tick sharply and may be loudest cold. A rod-bearing knock can change with rpm and load too, but it usually does not behave like a light metallic rattle that appears only when the throttle is opened against a heavy gear.

None of those patterns is a magic verdict. Engines are talented liars when two problems happen at once. A loose flexplate, cracked manifold, worn wrist pin, accessory drive, or valvetrain problem can fool a good ear. The point is to use load, temperature, rpm, throttle position, and timing together instead of deciding that every metallic sound is detonation.

The way the engine feels adds evidence. A knocking engine may flatten out under load, feel rough, or lose the crisp pull it had at lighter throttle. Severe abnormal combustion can make the engine feel like it is fighting itself because some pressure is arriving too early or too violently. On a computer-controlled engine, the driver may never hear much at all because the knock sensor hears the event and the computer removes timing before the cabin gets the full concert.

That creates a modern diagnostic trap. An engine can be living on the edge and feel merely soft because the control system keeps protecting it. The absence of audible ping does not prove the chamber had comfortable margin. Scan data showing repeated knock retard or timing being pulled under the same load can be the electronic version of the old metallic rattle.

Older engines can have inaudible knock too. Exhaust noise, insulation, engine construction, frequency, hearing loss, and a short-duration event can keep light detonation from sounding dramatic in the cabin. The engine does not require the driver’s permission before the pressure waves count. A plug that begins showing peppering, a piston that develops suspicious speckling or broken lands, or a problem that consistently improves with less load, less timing, cooler charge, or more knock-resistant fuel can be evidence even when nobody heard the classic coffee-can rattle.

Recognition comes from pattern, not panic. Ask what changes the noise. More throttle? More load? More heat? Lower rpm in a tall gear? Different fuel? Timing change? If the complaint tracks combustion conditions instead of simple engine speed, the cylinder deserves attention before the parts have to make the diagnosis permanent.

↑ Back to the Top, Before the Cylinder Starts Complaining


Main Causes of Knock

Knock usually starts where several weak margins pile into the same cylinder event. Fuel, spark timing, compression, chamber temperature, mixture quality, chamber shape, residual exhaust, and operating load all get a vote. The engine does not grade them one at a time like a polite schoolteacher. It adds them together while the piston is coming up and the flame front is trying to finish the job.

Fuel can run out of knock resistance. Timing can start the burn too early for the fuel and pressure being used. Compression can raise the pressure and temperature the mixture reaches before and during combustion. Heat can move the chamber closer to trouble before the plug ever fires. Mixture problems can create a hot or slow-burning cylinder. Weak quench or poor chamber motion can leave the end-gas sitting around under pressure longer than it should.

Deposits can change more than one variable at once. Carbon on the piston and chamber can slightly reduce clearance volume, which raises effective compression. Deposits can retain heat. A sharp or glowing deposit can become a local hot spot. Heavy deposits can also disturb the intended chamber shape and quench area. One dirty chamber can therefore have less knock margin than the compression-ratio number on the build sheet suggests.

Residual exhaust gas and mixture distribution belong in the same pile. Hot leftover gas can raise the starting temperature of the next charge, while uneven distribution can leave one cylinder leaner or hotter than the average mixture reading suggests. The engine may therefore have seven cylinders with comfortable margin and one cylinder living on borrowed time. Knock begins where the worst local combination crosses the line, not where the average engine looks good on paper.

Cylinder filling matters too. A mild engine at light throttle may have plenty of margin because the cylinder never traps enough air to create serious pressure. Open the throttle, hit the engine at its torque peak, improve volumetric efficiency, add boost, tow a load, or give the engine a cooler denser intake charge and the same static compression ratio can suddenly produce much higher pressure. The ratio did not change. The amount of mixture trapped and squeezed did.

That is why one engine can tolerate a combination another engine hates. The first may have a compact chamber, strong quench, sensible spark curve, even mixture distribution, good cooling, the right plug, and gearing that keeps it out of heavy low-rpm load. The second may have the same advertised compression ratio but a lazy chamber, hot intake air, too much timing too soon, weak fuel distribution, and a tall gear asking the engine to pull like a tractor.

The suspect lineup is therefore broader than “bad gas” or “too much timing.” Fuel tolerance, spark curve, compression, chamber motion, charge temperature, metal temperature, mixture distribution, plug heat range, deposits, cooling, gearing, vehicle load, and driver habit can all move the same margin.

None of them has to be guilty alone, and none of them gets cleared just because one paper number looks reasonable. Knock is where weak margins meet.

Related Pages:

↑ Back to the Top, Before the Cylinder Starts Complaining


Why Knock Usually Shows Up Under Load

Knock usually shows up under load because load is where the cylinder has to prove the combination. A free-rev in the driveway is a weak test. The throttle is barely open, cylinder filling is limited, and the engine is not being asked to produce much torque. It can sound perfect while the same engine rattles the first time the car hits a grade in high gear.

Open the throttle and the intake restriction falls. More air and fuel enter the cylinder, so the trapped charge rises. Compression now starts with more mixture in the same basic clearance volume. Pressure and temperature climb, and when the plug fires there is more chemical energy available in that firing cycle. That is exactly what makes torque — and exactly what makes a marginal knock combination quit hiding.

The useful distinction is between engine speed and cylinder load. RPM tells how often the engine gets another firing event. Load tells how much work each event is being asked to do. A lightly loaded engine at 2,000 rpm may have plenty of knock margin because the throttle is only partly open and the cylinders are not heavily filled. The same 2,000 rpm with the throttle buried can be a completely different combustion problem.

Manifold pressure is one clue. With the throttle mostly closed, a gasoline engine usually has substantial intake vacuum and the cylinder starts with less trapped charge. Open the throttle and manifold pressure moves closer to atmospheric pressure, cylinder filling improves, and the pressure after compression rises with it. On a boosted engine the change can be even larger because the cylinder starts with charge pressure above atmospheric before the piston ever begins squeezing it.

RPM changes the test because it changes both time and engine demand. Lugging the engine at low rpm with a lot of throttle can be ugly because the cylinder fills well, pressure is high, and the piston spends relatively more time through the crank-angle region where the end-gas is being compressed and heated. The flame still has to cross the chamber, and the engine is asking one firing event to make a large share of the torque needed to move the vehicle.

Downshifting changes that bargain. The engine turns more revolutions for the same road speed, the gear multiplication improves, and each individual firing event does not have to carry as much torque demand. The throttle may not need to be open as far. The engine can make the required road power with more cycles per minute instead of leaning so hard on each cycle. That is why a rattle at 1,800 rpm in high gear can disappear immediately at 2,700 rpm one gear lower even though the hill, fuel, coolant temperature, and road speed barely changed.

Vehicle weight and gearing decide how hard that load arrives. Tall axle gears, overdrive, a tight converter, heavy towing, steep grades, or a heavy car can all keep the engine in a high-load, low-rpm part of the map. A lighter car with shorter gearing may use the same engine and fuel without ever spending much time there. The engine did not suddenly become better. The vehicle stopped asking the same cylinder to do as much work per firing event.

Worth Knowing

Worth Knowing

Low rpm is not automatically bad, and high rpm is not automatically safe. The important question is how much cylinder filling and torque demand each firing event is carrying. A lower gear can raise rpm while reducing the load each individual combustion event has to shoulder.

Heat stacks on top of the load. A hill after ten minutes of traffic is a tougher test than the same hill first thing on a cool morning. Intake air is hotter, coolant and oil are warmer, the chamber has soaked up heat, and the exhaust-valve area may already be carrying a thermal load. The knock margin that looked comfortable cold can disappear once everything is hot and the throttle opens.

That is why towing, steep grades, repeated hard pulls, and low-rpm overdrive are useful diagnostic conditions. They combine cylinder filling, temperature, timing demand, and vehicle load in the place where the engine is most likely to expose a weak margin. A no-load rev only proves the engine can rev without doing much work. The loaded pull is where the combustion combination has to tell the truth.

↑ Back to the Top, Before the Cylinder Starts Complaining


Timing: Helpful Until It Gets Stupid

Spark advance is necessary because the mixture takes time to burn. The spark has to fire before top dead center so the flame has time to grow and useful pressure can arrive after the piston crosses the top. The problem begins when the spark starts the pressure rise earlier than the particular fuel, chamber, compression, temperature, and load can tolerate.

Think of timing as moving the whole pressure-building process around the crankshaft. Advance the spark and combustion begins earlier. More mixture burns while the piston is still coming up. Pressure rises sooner. That can improve torque when the previous timing was too late, because the pressure peak moves toward a more useful crank angle. Keep advancing and eventually the pressure rise begins leaning on the piston too early and squeezing the unburned end-gas harder while it is already getting hotter.

That is the knock connection. Earlier pressure means the end-gas spends more of the late compression stroke and early combustion period under high pressure and temperature. If the fuel and chamber have enough margin, no problem. If they do not, the end-gas can autoignite before the flame front gets there. The distributor did not directly “cause an explosion.” It moved the normal burn early enough to erase the remaining tolerance.

This is why an engine can gain torque as timing is advanced and then begin losing the argument before the driver notices a dramatic change in power. The useful timing window has an upper edge. Past that edge the engine may rattle, the computer may pull timing, or the extra negative work before top dead center may simply stop producing gains. More advance is not more horsepower by definition.

Older distributor engines split the curve into pieces. Initial timing sets the starting point. Mechanical advance adds timing as rpm rises. Vacuum advance adds timing under light-load conditions where cylinder pressure is lower and the slower, diluted mixture often needs more lead. Those systems overlap, so a single “total timing” number can hide a curve that is wrong where the car actually knocks.

Mechanical advance that comes in too quickly can give the engine too much timing at low or mid rpm under heavy throttle even if the final number looks sane at 3,500 rpm. Vacuum advance usually drops away as manifold vacuum falls under heavy load, but a wrong canister, wrong port, sticking mechanism, or odd combination can still make diagnosis interesting. Computer systems replace springs and weights with tables and sensors, but the job is the same: put spark where the engine needs it for that speed and load.

Chamber design changes timing demand because it changes how far and how fast the flame has to travel. Good mixture motion, compact chambers, sensible plug location, and effective quench can help the burn finish with less advance. A lazy chamber may need more lead just to put the pressure peak in the right place, which can make the end-gas spend more time hot and compressed.

Backing timing off may stop knock, but that does not automatically prove the original setting was the only fault. Retard can create margin that masks weak fuel, hot intake air, poor cooling, heavy deposits, or a bad chamber. The engine may now be safe but lazy. Diagnosis asks why the original combination ran out of margin instead of congratulating the wrench for making the noise disappear.

Geezer Says

Geezer Says

Timing is a curve, not a trophy number. The engine can want more timing at light load, less under heavy load, and a different answer again when the chamber is heat-soaked. One distributor setting does not describe the whole event.

The goal is not coward timing. The goal is enough advance to place pressure where the crank can use it without making the end-gas survive conditions it cannot tolerate.

Related Pages:

RPM changes timing demand because crankshaft speed changes how many crank degrees pass while the flame is traveling. At higher rpm the mixture has less real time to burn during each cycle even though the chamber did not get smaller. That is why engines generally need more advance as rpm rises until airflow, burn rate, and cylinder filling reach the part of the curve where additional lead stops helping. Mechanical advance is following a changing time relationship, not simply rewarding higher rpm with a bigger number.

↑ Back to the Top, Before the Cylinder Starts Complaining


Heat: The Knock Multiplier

Heat makes every knock problem meaner, but “heat” is not one single temperature. Intake air, coolant, chamber metal, piston crown, exhaust valve, spark plug, residual exhaust gas, oil, and underhood air can all be at different temperatures and affect the knock margin in different ways.

Hot intake air starts the mixture closer to trouble before compression begins. Compressing a hotter charge produces a hotter charge near top dead center than compressing the same mixture from a cooler starting point. That is one reason an engine can tolerate more timing on a cool morning and rattle after sitting in summer traffic with the intake manifold and air cleaner baking over a hot engine.

Coolant temperature tells part of the story because the head and chamber surfaces transfer heat into the cooling system. If coolant temperature rises, chamber metal temperature usually rises with it. But the gauge can still hide local trouble. The sender may sit in one passage while the exhaust-valve bridge, plug boss, or rear cylinder runs hotter because of local coolant flow, deposits, casting shape, or exhaust heat.

The exhaust valve lives in an especially ugly neighborhood. It is exposed to hot combustion gas and then has only limited time on the seat to transfer heat into the head. Wrong lash, poor seat contact, a lean cylinder, restricted cooling, wrong valve material, or heavy load can raise valve temperature. A very hot valve edge can become part of a pre-ignition problem even while the dashboard gauge looks respectable.

Spark-plug heat range matters for the same reason. The plug has to run hot enough to resist fouling and cool enough to keep its firing end from becoming an unwanted ignition source. A plug that is too hot for the engine, compression, load, or power level can hold too much heat in the tip. The spark is supposed to start combustion when commanded, not let the plug sit there glowing and freelancing between ignition events.

Deposits hold heat too. Carbon can create rough surfaces and local hot spots. Oil contamination can make deposits worse. A chamber with sharp edges or an exposed gasket edge can concentrate heat in places the original design never intended. Those are local temperatures, and local temperatures are what the nearby mixture actually experiences.

Heat soak explains why a car can be fine on the first pull and knock on the third. Repeated load adds heat faster than the cooling system and surrounding air can remove it. Shut the engine off and some components can get hotter for a while because coolant stops circulating while exhaust and chamber heat migrate into the head and intake. Restarting a heat-soaked engine can produce a very different knock margin than the same engine had twenty minutes earlier.

Heat does not need to be the first cause. It only needs to erase what little tolerance was left.

That is why the temperature gauge is evidence, not an acquittal. The sender reports one location in the coolant; it does not directly report piston-crown temperature, exhaust-valve temperature, plug-tip temperature, residual-gas temperature, or the end-gas temperature during compression. An engine can carry a perfectly ordinary coolant number while one chamber has lost margin because mixture distribution, deposits, airflow, or local cooling made that cylinder hotter than its neighbors. Knock is a local combustion event before it becomes a whole-engine complaint.

Related Pages:

↑ Back to the Top, Before the Cylinder Starts Complaining


Mixture, Carbon, and Chamber Shape

Mixture and chamber behavior decide how quickly and evenly the burn can move. A cylinder with good mixture preparation, useful motion, a sensible chamber, and enough knock-resistant fuel gives the flame front a fair chance to finish the job before the end-gas gets impatient. A poor mixture or lazy chamber stretches the event and leaves more opportunity for hot, compressed pockets to autoignite.

Air/fuel ratio matters, but the usual garage slogan that “lean is hot” is too crude to carry the whole lesson. A mixture that is too lean to burn well can misfire and make less heat. The dangerous region is a loaded cylinder running lean enough to reduce charge cooling and increase combustion temperature while still making substantial pressure. That cylinder may be hotter and more knock-prone than the average exhaust reading suggests.

Distribution is the bigger trap on many older engines. A carburetor feeds a manifold, and the manifold does not always hand every cylinder the same mixture. Runner shape, fuel dropout, plenum behavior, throttle angle, manifold temperature, and carburetor signal can make one cylinder leaner than another. An oxygen sensor in one collector may report an average that looks fine while one cylinder is already living on the edge.

Fuel evaporation affects temperature too. Turning liquid fuel into vapor absorbs heat. Good atomization and evaporation can help mixture quality and provide charge cooling. Poor distribution can send one runner a wetter mixture and another a drier, leaner mixture. That is one reason a single carburetor jet change can improve the average number without fixing the cylinder that was actually complaining.

Chamber shape is the metal-side version of the same argument. A compact chamber with the plug located near the active mixture can shorten flame travel. Useful quench can force mixture out of the narrow piston-to-head area as the piston approaches top dead center, creating motion and reducing stagnant pockets. Faster, more uniform burning can reduce the amount of spark advance needed and shorten the time the end-gas spends under high pressure.

Quench is not a magic word stamped on a piston. The clearance has to be close enough to create useful motion and far enough to avoid mechanical contact at speed, temperature, rod stretch, and piston rock. A wide quench area that never actually squeezes the mixture does not provide the same effect as a properly matched piston, deck height, gasket, and chamber.

Bore size affects flame travel too. A larger bore can put some mixture farther from the plug and increase the distance the flame has to cover. Plug location, valve arrangement, chamber shape, and piston crown determine whether that extra distance becomes a problem. That is why two engines with the same compression ratio can want different timing and fuel even before camshaft and vehicle load enter the discussion.

Piston shape has to cooperate with the chamber. A big dome can raise compression while blocking or dividing the flame path. A dish can lower compression while either preserving useful quench pads or throwing away chamber motion depending on its shape. The brochure may call all of them the same compression ratio. The flame front does not read brochures.

Reality Check

Reality Check

Compression ratio is only one number in the knock story. Chamber shape, quench clearance, piston crown, plug location, bore, mixture motion, fuel distribution, and temperature decide how difficult that compression ratio is to live with.

Carbon deposits complicate everything. They can reduce clearance volume and nudge effective compression upward. They can retain heat. They can create sharp or glowing edges. They can disturb quench and chamber motion. Heavy oil consumption or poor fuel quality can make deposits build unevenly, which means one cylinder may quietly lose knock margin before the rest of the engine gives any obvious warning.

Residual exhaust gas has an interesting two-sided role. Controlled exhaust-gas dilution can slow combustion and lower peak temperature in some operating conditions, which is one reason EGR can reduce knock in modern systems. Uncontrolled residuals caused by poor scavenging or odd cam/port interaction can make mixture quality and burn behavior less predictable. The important point is that the fresh charge does not always enter an empty, cool cylinder.

The burn happens in the real chamber, not in the compression-ratio line on the spec sheet.

Related Pages:

↑ Back to the Top, Before the Cylinder Starts Complaining


Why Knock Breaks Parts

Knock breaks parts because the pressure event becomes violent and uneven. An engine is built to survive very high pressure when that pressure rises in the expected shape and at the expected time. Detonation adds rapid pressure oscillations on top of the normal combustion load, so local areas of the piston, rings, gasket, plug, chamber, rod, and bearings can see a much nastier event than the average cylinder-pressure number suggests.

The piston crown sees the combustion event first. Repeated detonation can pound the crown and concentrate stress around the top ring land and ring grooves. Thin or brittle lands can crack. The ring can lose support, beat the groove, or overheat until its end gap closes and the ends butt together. Once the ring package starts losing seal, hot combustion gas gets where it does not belong, and the failure can accelerate from a combustion problem into a piston problem.

That local loading is why damage can begin around edges, ring lands, and other weak geometry instead of appearing as one neat dent in the middle of the piston. Pressure waves reflect through the chamber, and temperature is not uniform either. A piston edge near a hot exhaust side, a thin land above the top ring, or a sharp feature can be living a harder life than the rest of the crown.

The connecting rod and bearings carry the pressure into the crankshaft. Normal combustion already puts enormous force through the rod, but the load follows a smoother curve that the oil film and bearing area were designed to carry. Sharp pressure spikes raise instantaneous load and can squeeze the oil film harder. One light knock event is not likely to wipe out a healthy bearing, but sustained heavy detonation repeatedly asks the bearing to absorb load that arrives faster and less smoothly than intended.

The head gasket has a different problem. It has to keep combustion sealed while the head and block expand, the fasteners stretch, and pressure tries to lift the head away from the deck. Repeated knock adds sharp local loading to that clamp joint. If the deck, gasket, fastener load, or surface finish is already marginal, the pressure oscillations can help combustion work into a weak spot until a fire ring leaks, coolant gets involved, or the gasket finally quits pretending everything is fine.

The spark plug can be both witness and victim. Overheating can damage the insulator or electrodes. Speckling, abnormal deposits, or overheated coloring can provide clues, although plug reading is not a magic courtroom exhibit by itself. A damaged or overheated plug can then become part of a pre-ignition problem, which is how a detonation problem can create the hot source for something even nastier.

Pre-ignition usually shifts the damage toward heat destruction because the burn starts too early and may be building pressure while the piston is still coming up. Piston crowns can soften, erode, or melt. Exhaust-side edges can overheat. Plugs can become dangerously hot. The early burn also forces the piston, rod, bearings, and crankshaft to fight pressure at a crank angle where the engine has poor leverage to use it.

Warning

Warning

Knock damage is rarely one clean hammer mark. Pressure oscillation, heat, ring seal, oil-film load, gasket sealing, and hot spots can start feeding each other. Once that chain begins, the original combustion problem may no longer be the only thing hurting parts.

The ugly part is that mechanical damage and thermal damage can start feeding each other. A hammered ring land hurts sealing. Poor sealing raises blow-by and piston temperature. A damaged plug or sharp melted edge can become a new hot spot. A leaking gasket can disturb cooling. Once the engine crosses that line, the original knock problem may no longer be the only failure running the show.

Engines often survive small mistakes for a while, and that is how people learn the wrong lesson. It pinged last summer and did not blow up. It rattled on hills and made it home. It liked two more degrees of timing right up until the ring land finally had enough. Survival only proves the parts had reserve. It does not prove the pressure event was harmless.

↑ Back to the Top, Before the Cylinder Starts Complaining


Old Engines vs. Modern Engines

Older carbureted and distributor engines usually make knock more obvious because there is less electronic intervention between the cylinder and the driver. If the fuel is weak, the timing curve is aggressive, the engine is hot, or the driver lugs it in high gear, the complaint often arrives as audible ping, lost power, rising temperature, or some combination of the three.

That directness can be useful during diagnosis. Change fuel and the complaint changes. Slow the mechanical advance and the rattle moves. Downshift and the engine cleans up. Cool the intake charge and the margin returns. The old engine may be crude about reporting the problem, but the driver is hearing and feeling the combustion margin more directly.

The price of that honesty is that the old engine usually has very little ability to protect itself. A distributor cannot hear knock and pull two degrees from one cylinder. A carburetor cannot decide that one hot cylinder needs a different correction on the next firing event. Once the mechanical timing curve, mixture distribution, cooling system, and fuel are set, the engine mostly lives with the combination the owner gave it.

Modern engines changed the conversation with knock sensors, electronic spark control, fuel correction, variable cam timing, electronic throttle, boost control, and temperature models. A knock sensor listens to block vibration in the frequency range where knock tends to show up, and the controller compares that signal with engine speed, load, and its learned idea of normal background noise.

When the controller believes knock is occurring, it can reduce spark advance before the driver hears anything. Depending on the system, it may correct one cylinder, one bank, or the whole engine. A boosted engine may reduce boost. Other strategies may alter cam timing, throttle, or commanded mixture. The driver may notice only that the engine feels softer than it did a moment earlier.

That is the important teaching difference: the modern system can spend performance to buy combustion margin. The old engine exposes the problem and makes the driver back out of the throttle. The modern engine may quietly give away spark advance, boost, or torque to keep the same marginal condition from becoming a damaged piston. The symptom can disappear while the cause is still standing there.

That protection also makes diagnosis more data-dependent. An old engine may tell on itself with a rattle during the same hill every time. A modern engine may show repeated knock retard, lower commanded spark, reduced boost, or learned fuel-quality correction without ever making an obvious metallic sound. Looking only for audible ping can miss the entire complaint.

Knock sensors are not magic ears. Valvetrain activity, piston noise, transmission events, rough roads, and normal engine vibration all travel through the block. The controller has to decide which vibration belongs to combustion knock, and manufacturers use frequency windows, crank-angle windows, load information, and learned background noise to make that judgment. The system is far better than a driver guessing through loud exhaust, but it is still interpreting evidence rather than reading a label attached to the combustion event.

Geezer Says

Geezer Says

A quiet modern engine is not always a happy modern engine. If the computer keeps pulling timing, boost, or torque to stay out of knock, the protection system is doing its job — and spending performance to cover a combustion-margin problem.

Modern control also does not repeal the same physical limits that governed an old carbureted engine. Hot intake air still heats the charge. Too much cylinder pressure still squeezes the end-gas. Poor fuel still reduces knock margin. A bad chamber still burns badly. The difference is that electronics can react faster and hide more of the complaint by sacrificing performance before the driver notices.

So the useful comparison is not old engines knock and modern engines do not. Old engines usually expose the margin directly; modern engines often manage the margin for you. If the controller has to keep pulling power to keep the chamber out of trouble, the engine is protected, but the combination still deserves attention.

↑ Back to the Top, Before the Cylinder Starts Complaining


Fixing Knock Without Guessing

Fixing knock starts with refusing to treat every metallic noise and every detonation problem like the same complaint. Diagnosis should reproduce the condition, identify what changes it, and then restore the missing combustion margin without accidentally creating a different problem somewhere else.

First decide whether the noise behaves like combustion knock. Does it appear under throttle and load? Does it become worse when the engine is hot? Does downshifting or reducing throttle make it disappear immediately? Does the noise occur only in a certain rpm and load range? A steady valvetrain tick, cold piston slap, exhaust leak, loose accessory, or bearing noise may follow a different pattern even if the sound description is similar.

Then record the conditions instead of relying on “it does it sometimes.” Engine temperature, ambient temperature, intake-air temperature if available, rpm, gear, throttle position, vehicle speed, fuel used, timing setting, manifold vacuum or load, and whether the engine was heat-soaked can all matter. A repeatable complaint is much easier to diagnose than a memory of a rattle three weeks ago.

Build the diagnosis in one direction and keep it there: symptom → operating condition → likely physical event → evidence. A metallic rattle only under heavy low-rpm throttle points toward a load-sensitive combustion event. If a downshift removes it, the cylinder-load clue gets stronger. If a small timing reduction or known higher-octane fuel removes it under the same test, the end-gas-margin case gets stronger again. If the noise ignores throttle and follows engine speed cold or hot, the story starts wandering away from combustion knock. The point is not to win with one test. It is to make every test support the same mechanical chain.

Fuel is an easy controlled test when the engine is otherwise healthy. If a known higher-octane fuel consistently reduces or removes the complaint under the same conditions, knock resistance was part of the margin. That does not automatically prove the engine simply “needs premium.” It proves the previous fuel and the current combination were too close to the edge. Timing, heat, compression, deposits, and mixture still deserve a look if the required octane seems unreasonable for the build.

Timing should be checked as a curve, not just at idle. Verify initial timing, mechanical advance, the rpm where the advance comes in, total mechanical timing, and vacuum-advance behavior where applicable. Make sure the timing mark and balancer are telling the truth. A slipped outer ring or wrong pointer can turn a careful timing session into precision guesswork.

If reducing timing removes knock, note how much change was required and where. An engine that needs two degrees less under a brutal hot pull may simply have been tuned at the edge. An engine that needs ten degrees removed everywhere has probably given you a larger clue. The goal is not to leave it retarded and declare victory if something else is forcing the timing far away from what the engine should reasonably tolerate.

Heat comes next because a hot engine can make every other answer look worse. Check cooling-system condition, coolant level, thermostat operation, fan and shroud effectiveness, radiator airflow, water-pump operation, belt condition, and whether the engine is actually running at the temperature the gauge claims. Look for restricted passages, trapped air, blocked fins, hot intake-air routing, and obvious heat-soak problems.

Quick Test

Quick Test

Change one variable at a time under the same load. Different fuel, a small timing change, cooler operating temperature, or a lower gear can tell you which margin is disappearing. Five changes at once may stop the noise and teach you absolutely nothing.

Mixture needs cylinder-level suspicion even when only average measurements are available. Read the plugs. Look for one cylinder that is unusually clean, hot, oily, or deposited. Check for vacuum leaks, carburetor distribution problems, clogged injectors on later engines, fuel-pressure issues, and anything that could make one cylinder leaner than the rest. An average wideband number cannot guarantee that eight cylinders are sharing equally.

Compression and deposits deserve attention when the engine’s required octane or timing suddenly stopped making sense. Verify the actual compression ratio if the build history is uncertain. Heavy milling, thin gaskets, replacement pistons, incorrect deck height assumptions, or a head swap can move the ratio away from the number everyone has been repeating for years. A borescope can reveal heavy carbon or a suspiciously clean cylinder without pulling the heads immediately.

Spark plugs should have the correct reach and heat range, and their condition should make sense for the engine. Too much exposed thread, an overheated projected tip, or the wrong plug can create local trouble. A plug that is clearly hotter in one cylinder may be reporting a mixture, cooling, or chamber problem rather than volunteering to be replaced and forgotten.

Gearing and driver habit belong in the diagnosis because the engine only knows load. A tall overdrive and low rpm can put a heavy car or truck in a high-cylinder-pressure condition that the same engine handles cleanly one gear lower. If downshifting removes the complaint without changing coolant temperature, fuel, or base timing, that is useful evidence. The repair may be calibration, gearing, converter behavior, or simply using the gear the engine actually needs.

For modern engines, scan data can shorten the argument. Look at commanded and actual spark, knock retard, fuel trims, intake-air temperature, coolant temperature, load, boost where applicable, and whether the controller is repeatedly pulling timing in the same cells. Codes are helpful when they exist, but an engine does not need to set a code before the data becomes useful.

Do not use a correction as proof of cause without checking the chain behind it. Premium fuel may hide too much timing. Retarded timing may hide excessive heat. A richer mixture may cool the chamber while leaving poor distribution untouched. A colder plug may survive better while the engine is still running an unreasonable load or temperature. Diagnosis is finished when the mechanical story makes sense, not merely when the sound stops.

The final test is the original condition. Put the engine back under the same load, temperature, rpm, and gear that produced the knock. A fix that works only in the driveway has not passed anything. The cylinder complained under load; the repair has to prove itself under load.

Silence is not the only pass/fail standard when the engine gives you better evidence. On a sensor-equipped engine, repeated knock retard under the same load means the controller may be hiding the sound by pulling timing. On an older engine, plug condition, repeatable fuel or timing sensitivity, and the absence of damage signs after correction may be the best evidence available. The goal is comfortable combustion margin under the original load, not merely a cabin quiet enough that nobody complains.

Guessing may get lucky. Judgment gets there on purpose.

↑ Back to the Top, Before the Cylinder Starts Complaining


Bottom Line

Engine knock is the cylinder telling you the burn has lost control. The sound is only the warning. The important event is inside the chamber, where the normal flame front is supposed to finish the burn before the remaining end-gas gets hot and compressed enough to react on its own.

Detonation happens after the spark starts normal combustion. Pre-ignition starts combustion before the spark because some hot source in the chamber lights the mixture early. They can feed each other, they can damage many of the same parts, and neither one deserves to be dismissed as “just old-car noise.”

Load is where the combination tells the truth. More cylinder filling raises pressure. Low rpm and tall gearing can make each firing event carry more work. Heat reduces the remaining margin. Timing moves the pressure event around the crankshaft. Fuel determines how much autoignition resistance the end-gas has. Chamber shape, quench, plug location, piston shape, deposits, and mixture distribution decide how hard the fuel has to work.

That is why knock cannot be diagnosed from one number. Compression ratio alone is not enough. Total timing alone is not enough. Octane alone is not enough. Coolant temperature alone is not enough. The cylinder experiences the whole combination at once.

The repair should therefore follow the evidence. Reproduce the load, change one variable at a time, verify the timing curve, check heat and mixture distribution, make sure the fuel matches the actual pressure and chamber, and do not confuse a quieter symptom with a solved cause.

If the burn is arguing with the parts, do not keep driving until the parts lose.

Related Pages:

↑ Back to the Top, Before the Cylinder Starts Complaining