Auto High Octane Map

[Gelf VXR]

I'm thinking about an Auto High Octane Spark Map using a spread sheet.

Enter your Hi Octane Map calibration
Enter your KR map from the scan tool
Round recorded knock values upto the next integer (2.1 = 3)
Add 3 degrees to the rounded up value
Cells where no knock was recorded, change to minus 1
Smooth the map
Round the values up and down to the nearest integer
Subtract from Hi Octane calibration (plus's subtract, minus's add)
Paste back to tuning calibration

Cells with knock are reduced by knock value plus 3 degrees
Cells with no knock are advanced by 1 degree.

This gradually advances timing a little at a time until knock is detected, which is then retarded by the knock amount minus a further 3 degrees.

This should only be applied to logged cells and smoothed to their adjacent cells. I don't know how to do this in excel? Ignore empty cells (but not 0.00 cells entered from the log) Anyone??

Spread sheet attached, you may recognise it, its a modified interpolation sheet for smoothing BEN's from this forum

Thoughts????

[macca_779]

Nice idea mate. I do something similar but in a simpler manner and without the adding of timing your doing. I simply apply kr logged values to the high octane table twice (effectively doubling the value) I then select the cells either side of the modified cell and smooth it to a level I like. The reason for applying the logged map twice is that with smoothing the retarded value it will rise back up. If you only apply the logged value once in most cases when you smooth you return the cell back to pretty much what it was in the first place. Everyone has their own way of doing things.. Thats mine and I've found it works very well. Plus I hate non smooth timing tables as a smooth table transfers into a smooth power delivery. Nice

[Wasted Income]

Great ideas, guys!

[Gelf VXR]

I've played with this idea for a bit now, I think I have a working model.

I have removed the previous rounding functions.



I have entered my last 6 logged GM.KR from AutoVE in the spread sheet (both cell count and knock values) into Auto Hi Oct Spark Table -data Tab

Entered the number of logs entered (for the average calculation)

Averaged both cell counts and knock values.

Smoothed the cell count values so that surrounding cells have values greater than zero.

Converted all the zero values for knock to minus 1

Look up smoothed cell count values table, if cell count = zero then zero, if cell count > zero then cell value of value of knock/advance table

Smoothed these values

In the Auto Hi Octane Spark Map Tab, Entered my current Hi Octane Map calibration, subracted the smoothed knock/advance table to give new Hi Octane Spark Table to paste back into tune calibration.

Retard/Advance modifying Table

http://i5.photobucket.com/albums/y164/andysa1966/RetardAdvanceTable.jpg


Of course the logged knock data needs to be validated before use.

Before

http://i5.photobucket.com/albums/y164/andysa1966/SparkBefore.jpg

After

http://i5.photobucket.com/albums/y164/andysa1966/SparkAfter.jpg

[ChipsByAl]

Thanks for the update.
Al

[Gelf VXR]

Another similar method is to compare the actual spark advance to the hi octane table, the results show where the PCM's adaptive spark table is at, lower in knock areas and higher than the hi oct map in others. If the PCM is advancing above the Hi octane table, can we assume the hi octane map can be increased in this cel?? We let the PCM learn the optimun spark from the adaptive spark function. We can copy this table back to the Hi octane Table. (No spreadsheet)

The table below shows the difference between filtered SAE.SPARKADV (changing > 0.5 degrees per 100ms) and B5101, green cells higher and red lower than B5101. Empty cells were un logged.

http://i5.photobucket.com/albums/y164/andysa1966/SAESPARKADV-B5101.jpg

Before I do this, I am looking at my knock in detail, I have around 1.3% of my log with knock from GM.KR, of which around 45% accounts for less than cell count of 10 which I am ignoring for now. I'm going to try illiminating knock by other means before retarding the timing

http://i5.photobucket.com/albums/y164/andysa1966/Comparecellcounttoknock.jpg


Other considerations are my BEN's are not close enough to zero yet, and PE was not enabled in a lot of the high map cells, shown below

http://i5.photobucket.com/albums/y164/andysa1966/PEnotenabledinhighMAP2.jpg


I'm just putting my thoughts on the table, comments are appreciated?

[N0DIH]

Do you actually see better throttle response or power with pushing to the highest timing? With an efficient head like the Gen II/III/IV is peak power STILL being seen at highest tolerated timing?

I am running E85, so having 100+ octane in my LT1 I can see in the mid to high 30's for WOT timing, but it doesn't FEEL faster there (need some dyno time or some Elapsed Times to verify). So I tend to be only around a couple degrees more than high octane (LT1 Corvette base map) timing for these engines, but there is a difference, I have FE heads, the Vette has AL and slightly more compression....

Good work on the SS, I still like it!

[Gelf VXR]

I found one table {B5112} that can add or reduce spark, only affects cells over 4000 rpm and is multipled by {B5113} MAP table also. So doesnt really count for the spark added to the Hi octane table shown above.