Knock advance exceeding max advance table

[engineermike]

Anybody ever run across the knock advance not stopping at the limits set forth in the max knock advance table? I even zerod out the entire knock advance max table and flashed-entire but still logging over 3 deg of knock advance.

[engineermike]

IMG_2938.jpg

[RobCat030]

Are you sure it's advance? On my car negative numbers are advance and positive are retard.

[engineermike]

I am 100% positive it’s advancing. I know there is confusion around which pid shows negative or positive as advance, but you can tell this is advance by the even steps it takes as it ramps up, plus there are times where the knock correction tag goes negative. Even if the sign is backwards, I’m still getting + and - even with 0 for knock advance max.

[veeefour]

Probably borderline adder like lambda modifier.

[howeverthe5oh]

I am 100% positive it’s advancing. I know there is confusion around which pid shows negative or positive as advance, but you can tell this is advance by the even steps it takes as it ramps up, plus there are times where the knock correction tag goes negative. Even if the sign is backwards, I’m still getting + and - even with 0 for knock advance max.

You're looking at knock retard right?

[howeverthe5oh]

I am 100% positive it’s advancing. I know there is confusion around which pid shows negative or positive as advance, but you can tell this is advance by the even steps it takes as it ramps up, plus there are times where the knock correction tag goes negative. Even if the sign is backwards, I’m still getting + and - even with 0 for knock advance max.

or final spark advance?

[engineermike]

I’m talking about knock advance (ecm 44684), the amount of timing added if no knock is detected when spark source is “borderline”.

[engineermike]

Sooo...the Ford timing algorithm, during warm steady-state and not during an SAI event, summed up:

MBT - First looked up in the MBT tables as a function of measured Mapped Point, load, and speed. Then, corrected using the lambda modifier table.

Borderline - First looked up in the Borderline tables as a function of measured Mapped Point, load, and speed. Then, corrected using learned octane, lambda, MCT, and ECT.

Cylinder Pressure - First looked up in the cylinder pressure table as a function of load and speed, then corrected using a lambda modifier.

Pre-ignition - First looked up in the pre-ignition table as a function of load and speed, then corrected using a lambda modifier.

Then, a low-select is performed when the above 4 sources are compared.

If knock is sensed, timing is retarded regardless of spark source.

If the source is Borderline and no knock is sensed, it advances timing up to the max knock advance lookup table. If at any time, the total advance exceeds any of the other 3 sources above, then it stops advancing (though the spark source doesn't reveal which of the other 3 stops the advancing).

If at any time the above yields a number lower than min final spark, then min final spark is used.

If at any time the above yields a number higher than the max final spark, then max final spark is used.


The issue I'm having is that it appears to be ignoring the max advance lookup table.

[GHuggins]

Not sure if this is your problem or not, but all of my gt500 files have 8 rows for the knock advance limit table - is it possible that your just not seeing all of the rows? Also these tables are all 0 from factory and they still add timing, so is it possible it's just getting the timing added in from another source? I know it's typically the borderline knock main timing tables.

[Ruiner46]

I ran across a similar problem a few years ago, but at the time I chalked it up to something being handled differently in the Whipple tune. I made a thread about it here: https://forum.hptuners.com/showthrea...etard-limiting