lean idle after longer runs, perfect idle AFR every other time

[ccrp]

this one stumps me a bit, perfect idle AFR when cruising through town at steady 14.6 with quick rebound after every throttle blip.

If I kick it around on the highway for longer time like tuning WOT it goes straight to idle AFR of 14.6 but slowly creeps up to 15.5 and stays there no matter what. Changes to MAF airflow or VE table will not affect it, however runs fine in OLSD and does not do this, does it when CLMAF is enabled. Weird thing is STFT show zero % error as well when going up to 15.5 even if commanded AFR shows 14.63

This is a supercharged cammed/headed 382 LS1 @ 8 psi boost making about 600hp. Runs otherwise superb.


VE table got tuned with wideband AFR % error, same for complete MAF table. Plans for near future are to take MAF out and run 2bar MAP SD but until then want to make it run perfect. 60lbs injectors are dialed in like always so offsets etc. are correct.

tune attached, really open for suggestions here

[oren m]

[quote=ccrp;175980]Weird thing is STFT show zero % error as well when going up to 15.5 even if commanded AFR shows 14.63

Your both STFT and LTFT disabled . Please post log.

[ccrp]

LTFT are left disabled on purpose, but where do you see STFT's disabled?

I only have this log on this laptop, see idle AFR at the beginning of log.

[oren m]

I receive an error when open your log file.

[ccrp]

hmm, don't know oens fine on both laptops

[ccrp]

try again, renamed file

[oren m]

Frame 0 to 120 you are in OL then PCM switch to CL and activate the STFT. Your VE-STFT hyst is seems to be far from normal.

[C5 2000]

Definetely your VE need lots of work and you are runing lean with your injector pulsewidth change these values.

[ccrp]

VE table is near perfect from idle to 105kPa, wide band % error everywhere between -3 to +3% max. However, I do see quite a difference from what wide band reads at part throttle in terms of % error and what narrow bands read.

First I tuned lower portion of VE with the good old STFT CLSD method and car was rather rich and needed a lot of gas and did not feel happy. I then retuned using the wide band % error method in CLSD and values changed a lot. Car now runs perfectly and smooth at every MAP and rpm level up to full boost, just occassionally it creeps up to higher AFR values at idle.

The injector values are correct ones for those 60lbs and I have used them on the last 10 supercharger installs and never had an issue like this. Verified them and these are the recommended values from A&A too.

Could the stock narrow bands have issues causing confusion in the PCM? Like how comes narrow band wants to take out even more fuel if already running lean and STFT show like -10 at AFR 15.5?

Also why do changes in VE table at idle cells not show any effect at all, same for frequencies below 3500 Hz (this is where car idles in this case). A change like 10% up or down should show some response but doesn't.

[oren m]

In the log you posted the NB O2 are not in the log. Bad O2 can cause issues in close loop. How old are your O2?

[EaglePerformance]

Possible exhaust leak? Gets real hot, expands a bit, sucks in some air, cools down, seals up, idk, just thinking!?