[flatthump]

Originally Posted by WolfPlayer

This mystery has been solved. I cannot take credit for this. The idea came from Freon. I fully tested it tonight. As far as I can tell, I am the first person to test this and report on it. So, here goes.

There are a group of undefined tables that appear to be added into the final timing values. It was unknown if these tables functioned in this manner. It was also unknown if these tables had something to do with the knock routines or if they were related to something else.

My car followed these extra timing tables exactly. I could easily log a run and see 1d missing from 5200, 2d missing from 5600, 3d from 6000, and 1d from 6400. At 6800 I always had full theoretical timing back. A picture of these tables is attached below. This is for Calibration ID A2ZJB10J.

The first thing I did was to reconfigure my primary ignition a tad bit in order to have lower timing in the 5200-6400 region. This was important for these tests because if I was able to get the ECU to add full theoretical timing, then I wanted to do this with end result timing values that I knew were safe. I got a good baseline with my current map in terms of logs. I did 3 runs in 2nd gear from 2000-7000rpms. Again, my total timing between 5200-6400 was always missing degrees when compared to the theoretical timing. The degrees it was missing corresponded very well to the first map shown in the attached image.

The next test was to flash a new map setting the knock correction to be disabled above 5000rpms. This was Jeff's suggestion as it would lend evidence as to whether these tables may be related to knock control. I completed the 3 runs in second gear like above. The end result was the same as the first test. Timing was still not equal to the theoretical timing. There goes that theory.

For the last set of logs I flashed a new map with the tables below zeroed out (and my knock correction re-enabled). I completed the 3 runs in second gear again. Woot. Timing now equals the theoretical. Cooooool.

I do not know what these tables are for. However, they definitely remove timing from the theoretical. It doesn't appear to be related to the knock sensor.

t

Good work. Thats kind of strange.