I found a similar issue when I first started tuning my car. My solution was to modify the low octane maps so there wasn't a huge difference between the high and low octane maps. After doing this, the problem was gone. I was seeing up to 3º pulled at times though ...

I used a modified stock high octane map #1 for all 3 low octane maps. This added some advance across all of the power band but was still less advanced than the high octane map. I pulled a few degrees in some areas that were a little too close to the high octane map for safety.

 

Are your high octane ign map #2 and #3 the same as your #1 you posted?

EDIT: Nevermind, went and re-read your first post.

 

Here it is:

TimingAdv is MUT06
TimingAdv04 is MUT04
TimingAdv05 is MUT05
TimingAdv33 is MUT33

Conversion in EvoScan is x-20 for all.

 

The updated maps used in the previous log are the following:

High octane (#1, #2, #3 are the same):




Low octane (#1, #2, #3 are the same):




Note: the low octane map is no longer like stock, is little advanced to try as TouringBubble said.
It is very conservative because on street I'm not 100% sure to find a Shell pump everywere...

 

yeah so MUT33 is your final timing from the map lookups which looks correct 3* at 290/3400.

I will have to look through the code to find out what the reason for the drop is...

 

What elevation did this occur at, or was there a recent drastic elevation change? If so then perhaps the temp/baro tables do something else besides load.

I have seen this many times too, just not as severe, could be the ECU's X-Tau adjustment.


What is the reason in the attached .csv for the unusual values in the TimingAdv05 log at 23,08, 23,25, and 23,34 sec?:

-19
235
-19
235
233
-17

Chow

 

I would start by resetting the ECU to see if it forgets about why it insists on pulling timing there. I think it's holding on to a previous knock event in that section...just a hunch.

 

You could always add a splash of race fuel to see if its phantom knock or real knock.

 

the thing is.. he's getting timing pulled and no knock count are being reported. If he adds race fuel and the timing is restored that would be interesting

 

I would think that MUT05 should be logged in EvoScan as "x", not "x-20"...

Altitude is below 100 feet (300 mt) here. Logs are not hill, but flat street.

 

Mmm, I did never resetted the ECU. Is there a fuse (and where)? Does not reflashing reset the ECU?

 

Disconnect the negative battery terminal for a few minutes. That will reset your ECU.

 

I typically just disconnect the battery for awhile, I'm not sure how long it needs to be done to clear I typically go do something else and come back after 15 minutes or so.

I don't know if reflashing the ECU would be the same thing.

 

i brought this issue up before because I was puzzled at the same thing when it happend to me. I think sometimes the cars ECU gets knock but dataloggers dont logg it. MRfred says this is impossible but I think NOT.
The ECU is designed to not be compromised to read accurate data because of a user logging data. As of right now the ECU is designed to not memmory dump but instead pass data through and re-write over old ones.

But as another member have said, maybe its temperature or coolant related. What is the temperatues during this run? Also, I dont think changes in the MIVEC would have anything to do with the ECU pullling timing, why would it do that?

 

Would it be a smart try to log RPM, knocksum and timing only (faster) ?
Or I could try to increase baudrate, I remember a thread for MUT@62500 baud...

I will try to disconnect the battery to reset the ECU.