Weird RPM drop in 3rd gear WOT pull

[ysu]

Gents, after the beautiful ACE header install I've applied a great stage 2 tune to my car (many thanks to @Wayno!). Running 98RON (aus).
Now, trying to log and see how things go, I've found (and felt) a strange reluctance to go around 3.6k rpm in 3rd gear. Interestingly, I did not feel it in 2nd at all.

I'm wondering if anyone has any ideas what could the cause be? And even better; how to fix it. (although it's unlikely that I'll accelerate hard at this rpm, ever...)
I'm not even sure if this question is in the right section.

Sadly I did not have my oft to log things before the header install, so I've no idea how it was before. And I did not try to do 3rd gear pulls from ~2k rpm either.

Here are the logs, 3rd gear pulls as I said.
The second is a lot more damning, the rpm actually *drops* in the middle.
http://datazap.me/u/ysu/log-14685771...a=0-1-2-6-7-13
http://datazap.me/u/ysu/log-14685771...a=0-1-2-6-7-13

[steve99]

Looks like several parameters not logging correctly like flkc, can you log KC Learned instead will let you see any knock, the ignition advance drops a bit in yhat area. might also be traction controll cutting in if their was a bump in road or dirt\gravel. try with traction controll full off ie 5 secoind button push

[ysu]

Originally Posted by steve99

Looks like several parameters not logging correctly like flkc, can you log KC Learned instead will let you see any knock, the ignition advance drops a bit in yhat area. might also be traction controll cutting in if their was a bump in road or dirt\gravel. try with traction controll full off ie 5 secoind button push

IIRC the 2nd one was TC full off (5s) actually. I may have time to try logging other data, but "sadly" we're leaving for a week of holidays tomorrow morning

[ysu]

Well, thanks again for your help - it must have been the TC interfering indeed. I've made absolutely sure it was off now. And there was no trace of that drop.

Here's a few pulls - altho not full - in 3rd gear (and I've changed the logging as you recommended)

http://datazap.me/u/ysu/log-14686383...0-1-2-6-7-8-13

[ysu]

ah wait, there's something on this one after all:
http://datazap.me/u/ysu/log-14686388...0-1-2-6-7-8-13

And here are some 2nd gear pulls, not sure if these help at all...
http://datazap.me/u/ysu/log-14686388...0-1-2-6-7-8-13

[KoolBRZ]

Have you used Shiv's ECU speed learn technique since you loaded that tune? If not, you need to run it from 2k to 5k rpms slow enough to keep it in closed loop all the way. Do this at least 3 times, then pull some logs. The reason is, the ECU actually uses engine knock to determine the best ignition timing and A/F mixtures. It populates an ignition compensation table with what it "learns". Logs taken before this table is populated will be erratic and unreliable. On the other hand, if you've had that tune in for 2 weeks to a month already, the compensation table is probably already populated.

[steve99]

Originally Posted by ysu

ah wait, there's something on this one after all:
http://datazap.me/u/ysu/log-14686388...0-1-2-6-7-8-13

Ok now you have logged kc learned their appears to be no significant knock in that area where the rpm levels off.


The afr looks ok


The ignition timing or throttle doesn't drop off so it does not appear to be the VSC/Traction cutting in.


I assume you can actually feel this drop off ? as by the look of it it should be quite noticeable.


If you cannot feel it it may be some weird logging anomaly


If you can feel it the only thing left is likely a cam timing issue causing a loss of efficiency in that area.


maybe log intake/exhaust cam timing and see its hitting the desired values

[Kodename47]

I'd personally say something looks amiss with regards to ignition timing.

[ysu]

Thanks, gents, I'll do some more logging when I get back home next week.
@KoolBRZ yes, I've done a few pulls in 2nd beforehand on this tune, but thanks for the heads-up it's good to cover all angles every time.
In fact, I probably haven't done them right (i.e. slow enough), because if I output & look at the "fuel sys status" on the graph linked above, it seems to jump 2-4. If I remember correctly, that means it's learning, still?

[steve99]

Originally Posted by Kodename47

I'd personally say something looks amiss with regards to ignition timing.

Originally Posted by ysu

Thanks, gents, I'll do some more logging when I get back home next week.
@KoolBRZ yes, I've done a few pulls in 2nd beforehand on this tune, but thanks for the heads-up it's good to cover all angles every time.
In fact, I probably haven't done them right (i.e. slow enough), because if I output & look at the "fuel sys status" on the graph linked above, it seems to jump 2-4. If I remember correctly, that means it's learning, still?

As @Kodename47 said might be something weird in tables, maybe give the ignition tables like base B and Knock correction max A tables a look over check the values for typo,s and the axis both x and y for something weird like maybe a decimal point in wrong place ie load goes from 1.0 11.0, 1.2 or something weird, might be worth looking at a few others like maf scaling, maybe fueling etc.

[KoolBRZ]

Originally Posted by ysu

Thanks, gents, I'll do some more logging when I get back home next week.
@KoolBRZ yes, I've done a few pulls in 2nd beforehand on this tune, but thanks for the heads-up it's good to cover all angles every time.
In fact, I probably haven't done them right (i.e. slow enough), because if I output & look at the "fuel sys status" on the graph linked above, it seems to jump 2-4. If I remember correctly, that means it's learning, still?

The 2 - 4 change indicates the transition to Open Loop. It does mean the run wasn't done slow enough, or it could mean something weird is in a table, like @steve99 just mentioned, pushing it into Open Loop before the engine is ready for Open Loop. You want to keep it in Closed Loop for the three Learning runs before you Log.
Perhaps your tune reacts the way it does because of large transitions between the cells of certain tables. They could cause certain peak requested load conditions to spike briefly, resulting in the logs and results you are seeing now.
Try copying over #2 of the three tunes I sent you. It's in American values, so temperatures and pressures will need to be re-scaled and recalculated This is the tune I am running now. It already has fueling compensation for MAP above 0.0 psi added in. The AVCS and PI/DI tables are very smooth. The only things you would need to change, would be higher compensation ranges for your higher boost.