Questions about spark timing

[pauly24]

On the E38s, from what I understand the ECU will use any value inbetween the high and low map.
It constantly increases the spark timing and when knock has occurred it pulls some out. Only to keep re-iterating the process.
These values becomes the learned spark values.

So if that is all true, then does that mean you dont have to ever touch the spark maps because they are self adjusting?

The only thing I could see that would need adjusting is the sensitivity from when to pull knock and/or how much.

Sorry for any ignorance if any, I've just purchased EFIlive but it hasn't arrived yet so trying to wrap my heads around it all still.

[joecar]

Hi pauly24, welcome to the forum...

When the ECM detects knock, it does two things:
- it immediately determines how much knock retard is required and applies it,
- it moves the octane scaler away from the HO and toward the LO table;
(two distinct concepts: octane scaler and knock retard)

The octane scaler is viewable in the pid E38.OCTSCL_DMA;
The scaler is used to compute a weighted average between the HO and LO tables.
E38.OCTSCL_DMA = 0% indicates full HO weight;
E38.OCTSCL_DMA = 100% indicates full LO weight;
(this is in the opposite direction compared to the LS1 PCM).

If knock persists, the scaler keeps moving toward the LO table;
If knock stops, the scaler moves back toward the HO table;

The E38 has an extensive set of tables/parameters to control:
- how fast the octane scaler moves toward either of the LO and HO tables,
- how fast knock retard is applied and decayed.

You shouldn't have to adjust the HO and LO tables... It's not a good idea to adjust the HO/LO tables without a dyno...

A dyno allows finding timing that produces the best peak torque that avoids detonation.

[pauly24]

Ok so back to my original question (kind of)

Does all that mean, when playing with these cars there is very little gain in playing with the spark, because technically it should always be at the optimum setting.

Why would it matter what the HO table is set at? whether it was set at its normal value of say 80, or an extremely high number say 50000, it wouldn't make a difference because it will know when to stop advancing the spark.

Obviously you wouldn't set the numbers like that but i just use those numbers to help prove my point.

What, if any advantages can you get from playing with the spark maps then if spark is always optimized?

EDIT: I think i answered my own question, the only advantage is if you know you are using 98 octane fuel, you can tune the HO map to be at MBT using a dyne (thats if MBT is before detonation, which I believe it is now with most modern engines)

But anything apart from 98 fuel will knock and will pull timing out so the HO values are irrelevant.

[joecar]

By advancing timing there are gains in torque, but at the risk of knocking... some people do it and get away with no knock, others find lots of knock.

The octane scaler and knock retard mechanisms only kick in after knock has been detected;

Each time knock occurs, damage occurs... the amount of damage is related to both the intensity and duration of knock... if it lightly knocks for a moment, then only a tiny amount of damage (microscopic); if it heavily knocks for a period of time, then large damage occurs (piston flakes visible on spark plug ceramic, melted/holed pistons, blown head gaskets).

If you set the HO table high, say 40 degrees everywhere, then the engine will constantly go thru this:
- knock detected,
- PCM applies knock retard and moves octane scaler toward LO,
- KR is ramped out, OS is moved back toward HO,
- knock detected, repeat.
i.e. engine is continually driven to knock.

This is bad for various reasons including damage... but also, when knock is induced, to stop the knocking, timing has to be retarded to some value below the point where knock would have not occurred in the first place... this means that some amount of torque is given up when KR is applied, this is not acceptable if trying to win a drag race.

The stock timing tables are not necessarily optimized from the OEM, altho they are usually good... sometimes the OEM adjusts timing for model price range of the vehicle (the cheaper model makes less power, slightly retarding timing is one way to achieve this).

When you add/modify parts, you change the engine's burn rate/ability... the only safe way to modify timing is using a dyno and to carefully monitor for knock.

In some cases detonation may occur at or before MBT.

Knock still occurs with 98RON (93(R+M)/2)... it takes race gas (>100(R+M)/2) to prevent knock.

The other aspect of this is that over-advancing timing makes cylinder pressure peak at the wrong crank angle, either impeding crank rotation, or not taking advantage of crank-rod angle (i.e. leverage).

[pauly24]

Yes thank you for the reply.

But there is some conflicting information here, you say

"If you set the HO table high, say 40 degrees everywhere, then the engine will constantly go thru this:
- knock detected,
- PCM applies knock retard and moves octane scaler toward LO,
- KR is ramped out, OS is moved back toward HO,
- knock detected, repeat.
i.e. engine is continually driven to knock."

So this means the only time that you wont experience knock is when your running 98 ron fuel.
If you run 95 or 92 ron fuel your engine is continually driven to knock then from factory. So it can't be bad, the system in place to pull timing must be robust enough to not cause damage then.

[WeathermanShawn]

So this means the only time that you wont experience knock is when your running 98 ron fuel.
If you run 95 or 92 ron fuel your engine is continually driven to knock then from factory. So it can't be bad, the system in place to pull timing must be robust enough to not cause damage then.

I don't think that is necessarily what Joecar's conclusions are. Knock is bad. So is too little timing. It needs to be perfect. Then you will have ample torque and a long engine life.

For example, if I ran a stock tune from a 1998 Camaro LS1, my engine would develop KR (too high of spark at WOT). A 2002 Spark Table is pretty conservative. One might lose HP/TQ with just 19 degrees of spark advance at WOT.

Only a dyno can really measure safely that sweet spot. Here is a good link to a discussion of spark timing vs WOT fuel. It may nor may not answer some of your questions. But, I think it is a decent discussion..http:///www.innovatemotorsports.com/resources/myths.php

Hope that helps..

[swingtan]

Another point to remember is that the Octane Scaler is applied across the entire spark map. It is not like a 3rd spark map, just a single "correction factor between the high and low maps. Here's an example....

Let's say you did set the high table to a flat 40' in every cell, then to allow for knock, you set the low table to maybe 5'. You now have a difference of 35' between the high and low tables, plenty of room for knock correction. At light cruise 40' is probably close to an ideal amount of advance, giving efficient combustion at light loads. But at any sort of load, it'll start knocking like a jam tin full of ball bearings. So the ECM will start pulling timing to stop the knock. At low RPM and high load, the engine might only be able take 12', so the ECM will want to pull out 28' of timing. Because the Octane Scaler is a single correction figure, that -28' figure is applied across the entire Hi Spark Table, so now at light cruise, you will only get 12' of advance, way below the optimal figure. Even worse is that the Octane Scaler can only start returning timing to the Hi table when the engine is under load, the very cells that are way over the levels they should be. So it will never bring spark back up and you will be stuck on the low table all of the time.

The idea of a learned Knock correction system providing the "optimal" timing is simply wrong. What it does do, is allow a "generic spark timing map" to be fairly good under most conditions. If the map was to be optimised for a particular motor, then the knock system should never become active and in this case it allows an amount of protection for the use of lower octane fuels.

Simon.

[WeathermanShawn]

Good answer! That explains it very well..

[joecar]

Yes thank you for the reply.

But there is some conflicting information here, you say

"If you set the HO table high, say 40 degrees everywhere, then the engine will constantly go thru this:
- knock detected,
- PCM applies knock retard and moves octane scaler toward LO,
- KR is ramped out, OS is moved back toward HO,
- knock detected, repeat.
i.e. engine is continually driven to knock."

So this means the only time that you wont experience knock is when your running 98 ron fuel.
If you run 95 or 92 ron fuel your engine is continually driven to knock then from factory. So it can't be bad, the system in place to pull timing must be robust enough to not cause damage then.

You will still experience knock with 98RON... you need race gas (110-120) to not see any knock.

Knock is bad... the PCM pulls timing to protect the engine (good), but the engine knocking in the first place has caused incremental damage (bad).

Every time knock occurs (which is an explosion instead of a burn), some amount of damage occurs to the combustion chamber... normal combustion burn produces cylinder pressures of 600-900 psi, while knock produces pressures of 2000 psi and higher which even for a fraction of a second is bad.

If the engine is continually driving to knock, then there is a problem that has to be addressed (bad fuel, AFR too lean, overheating, timing too advanced, high too compression, carbon deposits, etc...).

[pauly24]

Thought I'd ask here instead of starting a new thread.

When the MAF is failed through the tune does it revert to the lower octane table? If so how can this be fixed as I want to have both tables active obviously for the safety net.