Question about knock, predetonation, octane (recent Time to Kill Weekly)


This week’s Time to Kill email had a link to this article:

It states:
"…the heat of the engine and the compression of the fuel and air can cause “pre-ignition,” the ignition of the mixture without a spark at all. It causes knocking that not only is unpleasant to listen to, it’s inefficient and can damage the engine."

I have wondered this for a long time and I hope someone can answer this. So here is my question: Why is it that changing engine timing can reduce the incident of knock? Engine timing affects neither 1) nor 2) above. Seems to me predetonation will occur regardless of timing as it precedes the spark.


I’ve always had the same question.


I don’t know for a fact, but my thought is this: too-advanced timing results in knocking when the spark occurs while the piston’s on the way under conditions where the combustion front propagates before the piston head back down. The pressure increase tries to push down the rising piston. Delaying the spark (retarding the timing) results in the cylinder experiencing combustion with the pressure increase working to push the piston down.


It isn’t the preignition that is detrimental to engine components but the resulting detonation. Detonation is where unburned mixture explodes rather than combusting in a controlled flame front. The optimum situation is to have the spark ignite the fuel mixture and have the flame front progress to the last of the mixture as designed. This provides an even pressure build on the pistion and other components. The spark plug is usually positioned to allow the shortest distance to the completion of combustion. If the mixture spark-detonates, the flame front progresses as designed; the unburned mixture is compressed by the movement of the piston and the heated expansion of the combusting mixture; a residual unburned mixture explodes rather than burns; and causes a sharp impact to the combustion chamber, piston, and valves. This transfers excess heat to the components and can overstress the components. Mechancial signs of sever sustained detonation are broken spark plug ceramics; burnt holes in pistons; and valve burning.

The problem with preignition is that there is more time for the compression/mixture heating to affect the unburnt mixture. Also the position of the ignition point may extend the flame front travel distance giving more time for detonation to occur.

If you notice the method the ECM uses to reduce detonation is to retard the spark timing even to the point where ignition occurs after top dead center. That reduces the efficiency of the combustion cycle; adds more heat to the combustion chamber and components; and sends hotter exhaust by the exhaust valves and out the tail pipe. Another strategy the ECM can use is to richen the misture to add fuel cooling to the combustion process. The standard procedure for a “Go Around” with a single engine Cessna is to remove carb heat; go full rich on mixture; full throttle; and pitch for best rate of climb.

In the days when pistons got carbon build up, a piece of glowing carbon would become the igintion point for the preignition. Again, going to a rich mixture was enough to cool the mixture so that detonation was avoided. It might be enough to quench the glowing carbon spot.


I didn’t read the link but you have to consider the explosion that takes place occurs in so many milliseconds. (It varies…)
The explosion must take place before the piston reaches the top of the compression stroke.
Ideally, the peak of the explosion should occur right when the piston reaches top center and on the initial fractions of a degree on the downstroke.

If the peak occurs before the piston reaches the top you’re going to have pinging.
If the peak occurs a fair amount of time after the piston starts down on the power stroke then you’re going to have reduced power.


Thank you Researcher.

To summarize, your first and second paragraphs are all about why detonation is bad, and what constitutes optimal burn.

But the the key is in your third paragraph where you say that the engine is run such that it stays cooler by “sends hotter exhaust by the exhaust valves and out the tail pipe”. So it affects #1 (heat) of the combustion chamber.

I am also familiar with the richen mixture approach as some cars smell like unburned gas when they accellerate.


@ok4450: I understand WHY predetonation is bad. But that is not the gist of my question. The question (which I think Researcher has answered) is HOW the ECU is able to reduce pinging by controlling timing. In my concept of the sequence predetonation occurs BEFORE the spark.

So first occurs predetonation and THEN spark from the ECU. If that is true, why would delaying the spark help? I think I have an answer now.


When the timing is advanced too much, the spark and resulting flame front travelling across the piston are causing an increase in pressure while the area above the piston is still getting smaller (i.e. the piston is travelling upward). This causes the pressure to increase more than designed. Higher pressure means higher temperatures, and those higher temps cause the unburned gas still at the far ends of the piston to spontaneously ignite.

Now when the timing more retarded, much of the burning across the top of the piston occurs after the piston reaches the top (TDC). There is less opportunity for the pressure to rise so much (and hence temperature) because the flamefront is growing as the area above the piston is getting larger and larger. (from the piston travelling downward).


My '80 Bronco did not have the benefit of computer timing control.
I let the cackling / pinging ( just under acceleration, especially long upgrades ) go on too long and one night near the end of the two hour drive home from Albuquerque . . kablooey !
Coughing , sputtering, loud knocking, smoke from under the hood and tail pipe, . . pull off to the side, family in the truck . . now what ?
Waited a bit and tried to restart . . not a chance . . ‘‘blew’’ the engine as they say . .I thought.
. . . . walked two miles to the nearest phone and called a friend for a ride and had it towed to my shop the next day.
Joe customer would have just put in another engine but was too curious and told my mechanic I’d pay him to tear it down to learn just how bad it was.

    • so he took it apart to discover . . . a burned hole clean through the top of just one piston.
      one 10 dollar used piston and a few gaskets ( 351M cast iron engine ) and down the road we went !


That’s a good story @“ken green” … It must be true what they say about those older Ford with the Detroit Iron, they are so easy to fix, they can be fixed with chewing gum … lol …

About the OP’s question, there’s two different problems that can happen. The first is if the spark fires too soon on the upstroke. That causes a big force pushing down on a piston still going up. It’s the immoveable object vs the unstoppable force, never a good thing, and will often cause a noise often referred to as a ping, but when it happens big-time, like if the timing gets way too far advanced, it can sound more like shaking a tin can of nails, really loud clattering sound. The first problem it is still the spark that ignites the mixture. The second problem that can happen is the mixture ignites before any spark has fired. If a hot spot forms in the cylinder for example, that can ignite the mixture prematurely, before any spark even fires. Timing too advanced if it goes on long enough can cause the second problem too, so they are sort of related. But if the only problem is the first one, the spark is occurring too soon, then retarding the timing so the spark occurs later, nearer to the top of the compression stroke will eliminate the pinging sounds.

On newer cars the experiment isn’t easy to do, but on older cars where you could experiment with the timing, this is all easy to verify. On my truck, it has a vacuum advance. I can hook up a vacuum pump to that and if I pump it up to 15 or so I’ll hear the pinging & clattering sound clearly.


There is detonation and preignition. There’s no such thing as “predetonation”. You can have preignition without detonation, you can have detonation without preigition, and you can also have both together.
Detonation is the near instantaneous burn of the fuel air mixture with a resulting shock wave that makes the sound you hear. The desired burn of the fuel air mixture is called deflagration, which is a rapid but smooth burn of the fuel air mixture.
It doesn’t have to be either a deflagration or a detonation, it usually starts off as a deflagration but transitions to detonation as the remaining gasses get compressed by the heat of the combustion.
When you think of detonation, think of dynamite. We want dynamite to detonate so the resulting shock wave destroys stuff.
On the other hand, we want gunpowder to deflagrate, so it will push the bullet down the bore without destroying the gun.


Back in the “GOOD OLE DAYS” the factory spec for timing an engine had a foot note that read, "these are approximate settings, engine design, altitude, temperature, fuel and engine condition will all influence timing. The determining factor limiting advance will still be the ‘knock point’ of the engine.


That was always the last step of the tune-up: points, plugs, condenser, cap, rotor, dwell, timing (stock + 3 or so). If it knocked in top gear up a hill, set it back some, if not, bump it up some.


The most spark advance without detonation is not necessarily the optimum ignition timing. Power and efficiency will drop with an over advanced spark even if no detonation occurs. In fact, detonation can even go away if you give the engine a ridiculous amount of spark advance. What happens is the spark ignites the mixture when the piston has only compressed it around 6 to 1 and the burn is finished before the piston compresses it enough to get into the detonation region. This makes for an engine that kicks back hard while cranking and overheats badly while running, but doesn’t detonate.


Since we’re talking about modern day engines and not the realm of the points and carburetor days, remember now we not only have ignition timing that changes but valve timing as well.


It wouldn’t be smart to let a car leave the shop with a fresh tune up that hits back on the starter @B.L.E. It seemed that first using a timing light and then a vacuum gauge to get the maximum vacuum then watch the timing marks and drop back 2 to 3*. And after a good road test checking for detonation, repeatedly cranking the hot engine to test for kick back. An old well worn engine could sometimes run well with the spark so advanced it could break starters and flywheels.

But the quote re timing is word for word from a 50 year old Chilton manual which states that all the information is from the manufacturers.


Adjusting the ignition timing to maximize vacuum would be a great way to get the optimum ignition timing for an idling engine, and would minimize idle fuel consumption.
To get the optimum full throttle spark advance you would have to adjust it while under load at full throttle and hunt for maximum power, which may or may not happen just short of the advance that produces detonation.
I suppose with a Model T Ford that used a manual spark advance, one could open it up on the highway and then use the spark advance lever to find the highest top speed.
Run premium gas in this car with its around 4 to 1 compression ratio and you likely won’t get detonation at any degree of spark advance, however, there will be a spark advance that produces the maximum amount of power.

The main reason that glow plug engines used for model airplanes were so sensitive to fuel air mixtures was that leaning or richening the mixture actually changed the ignition advance. Richening the mixture retarded the timing and leaning it advanced the timing. The main reason these engines use a methanol/nitromethane mixture for fuel is because this fuel will produce a lot of power even when the fuel air mixture is stupid-rich, not because glow plugs won’t ignite gasoline. That way, the richness of the fuel mixture can be used to control the ignition timing.
I learned this from a Cox .09 cubic inch engine I ran in one of my small model airplanes. I would run one to three head gaskets on it to change compression ratios. With the lower compression, the engine needed to be leaner to run well. With higher compression, it needed to be richer. I usually ran two or three head gaskets when using high nitro fuel, 30% or so and used a single gasket when running 10 to 15% nitro.


Preignition and detonation are two different conditions…In modern engine designs with todays electronic controls neither condition happens very often…BLE gave a good explanation.


When working on pre digital engines with no concern for techinical terms adjusting the ignition timing was done to either set it at the factory mark to play it safe or adjust it for the best performance relative to all the significant inputs. The single most important input was to avoid knocking/pinging/spark knock, etc. I found that using a vacuum gauge was much quicker and more accurate than using a tachometer. A good road test and hot restart was always done and the proof was always in repeat customers and referrels.