What causes ignition retardation?

[onlinebiker]

My 99 SC2 keeps retarding the ignition while driving at steady speed -50 to 55 mph - while on level ground.


It will go from 42° btdc down to 10 ° btdc (or less) - resulting in a loss of felt power....

I replaced the knock sensor, and throttle body and sensors with no change.

[BV22]

The only thing I can think of based on my years of reading this site is that there have been cases where the clunk of a failing serpentine belt tensioner is read as a knock by the knock sensor which retards the timing. How old is the tensioner on the offending car?

[onlinebiker]

Unknown..... Probably original with about 130k miles on it....

[BV22]

There are many folks here more clever than I am, and hopefully they will chime in with other ideas. Given the size of your fleet (if your signature list is of currently owned cars), swapping out the belt tensioner should be rather easy and would either fix the problem or eliminate the tensioner as the source of your issue.

[billr]

How are you reading that retard, with live-data? Doesn't the live-data identify if it is "knock retard" or "knock detected"? Also watch MAP, see if that if flaky. spark advance is primarily controlled by MAP.

[onlinebiker]

My INNOVA doesn't say why it is retarding spark.

I cleaned the MAP sensor passage and replaced it with a good one.

No change

[fdryer]

If I'm not mistaken, ignition timing for spark starts as advanced (up to 45 degrees BTDC) and retarded from various factors. Service manuals never describe degrees of timing for any specific rpm or load as this varies. What is known is any retard of timing is either momentary or long lasting with programming always attempting to advance timing to maximum. Since engine loads, xmission gearing, temperature, humidity, and other factors are used, timing will always vary. Did you consider a fuel issue, possibly lower fuel pressure? Are spark plugs fouled, worn? As covered in spark timing descriptions in aa1car.com, a faulty egr valve not allowing exhaust gases into the intake manifold to mix with the fresh air/fuel mixture to lower combustion temperatures may cause the pcm to retard timing without any error codes for egr valve faults.

What's normal fuel mileage and has it decreased significantly since noticing spark retard on your scantool and seat of the pants feel (butt dyno)? One way to tell if spark timing and loss of power is genuine is by fuel mileage.

If the belt tensioner is weak, the drive belt may flutter at idle and easily depressed (engine off) with fingers. Normal belt tension, if measured with a spring gauge, would be around 45-55 pounds. Drive belts should be banjo tight. When the tensioner weakens, it may hit its own end stop with a noise most here describe as engine knock. Confusing actual engine knock, pre detonation in cylinders, can be confusing with many swearing engine knock is real when its more likely the belt tensioner weakened over the years. Some try suggestions of high octane fuel but never report their findings - genuine engine knock from pre detonation of fuel will go away when premium fuel is used. Some with readers may use another suggestion of tapping the engine block at idle and monitor spark timing; metal tapping (small hammer, wrench, etc) against the engine block will be picked up by the knock sensor with timing retarded and seen on readers capable of live displays. Some readers need screen refreshing in order to see changes in values.

One good source of info would be aa1car.com. Spark knock is described in detail; https://www.aa1car.com/library/spark_knock.htm

[onlinebiker]

Belts are tight....

Will swap out the EGR and see what happens.....


I'm beginning to think I may need to get a goat....


A sacrifice may be in order....

[alordofchaos]

I think people have mentioned a loose heat shield as rattling and being picked up as knock but it was never tested nor proven.

Carbon build up causing knock?

Engine temps normal during cruise (actual temp per OBD, not instrument panel)?

I'm assuming correct plugs and gas not an issue.

IAT and ECTS temp readings are close when cold?

[onlinebiker]

Changed the EGR - No change.

I did note that at certain speeds - in excess of 57 mph - the timing stayed up near 43° or more btdc.....

Also - during heavy acceleration the timing drops to 15-20° btdc... That seems wrong - under heavy acceleration I would expect full or near full advance..

[billr]

Nope. Under heavy load (accel, high MAP) the spark advance will be reduced.

Rather than a goat, how about a scanner that yields better live-data?

You might try disconnecting the knock sensor, see what effects that causes.

Some vehicles retard the spark a lot when "over-run" is detected, based on the MAP, VSS, rpm, and TPS. You may be looking at a problem with the VSS or TPS. Or still the MAP sensor circuit. Cleaning/checking/changing the sensor does not ensure the entire circuit is OK. The live-data does that...

[onlinebiker]

Originally Posted by billr

Nope. Under heavy load (accel, high MAP) the spark advance will be reduced.

Rather than a goat, how about a scanner that yields better live-data?

You might try disconnecting the knock sensor, see what effects that causes.

Some vehicles retard the spark a lot when "over-run" is detected, based on the MAP, VSS, rpm, and TPS. You may be looking at a problem with the VSS or TPS. Or still the MAP sensor circuit. Cleaning/checking/changing the sensor does not ensure the entire circuit is OK. The live-data does that...

Good idea. I will try running with knock sensor disconnected. If that makes no difference it should eliminate any sonic problem - such as a bad tensioner or idler (or alternator, water pump or p/s pump) from tricking the knock sensor into action....

[RobertGary1]

Do you have any codes? I know some of you live in states where you just drive around with the MIL light on all the time.

-Robert

[onlinebiker]

That's part of it that's strange. No engine codes...

[billr]

Not so strange. There would probably be no (knock sensor) code as long as some signal is reported to the PCM. False triggering, like from a banging idler, would be a "normal" knock. When you disconnect the sensor, that should set a code.

[98_Gold_Uranus]

Originally Posted by alordofchaos

Carbon build up causing knock?

This is the first thing I thought of. Carbon build up on top of the pistons can increase the compression enough to cause detonation and pinging. Often times the knock sensor can pick it up and pull timing.

Try putting some octane booster in the tank or just fill up with a higher grade of fuel next time you fill up and see what happens.

Use to have an old Ford Ranger, the ones with the dual spark plug per cylinder 4 banger, that after about 200k miles I had to start putting premium in it or it pinged horribly... After I started putting 93 octane in it it never did it again.

[OldNuc]

A failed(disconnected) Knock Sensor Forces the timing into the COLD engine base timing table where it will stay.

[onlinebiker]

Will fill with 92 octane and see what happens.

If it improves - I'll have to try the old steam cleaning trick.

[OldNuc]

Be sure you do not have restricted or badly leaking exhaust.

[onlinebiker]

I put in 8 gallons of 93 octane.

I did not get any retarding at 52-55 - but did get a bit of stumbling at lower speeds while accelerating.

I need to put some miles on it and see what happens.