That log is looking good. Don't worry about that negligible .03 higher lambda reading on the wideband. Wideband compared to commanded being so close (while running closed loop!) is confirmation that the wideband is accurate. If there is a tiny exhaust leak the effect will disappear as rpm's increase. Accuracy under load is probably what you're concerned about. There will be no exhaust reversion pulling in air through a pinhole when you're on it.
Fuel trims are nice and tight. It's normal for one bank to be a little off from the other. As LTFT's populate the fuel trims should stabilize.
IAT change of 41F just from moving the MAF/IAT combo? Idk about that. A standalone IAT close to the throttle body will be more accurate.
Last edited by SiriusC1024; 06-02-2023 at 09:35 PM.
i had to move the car out of the way so i flashed the tune from the local tuner....the maf tuning starts tomorrow i hope, but i notice as soon as i start logging the wideband....or one of the wideband pid's disappears.what causes that?
when i took the first log it was around 2pm, the second log just at dark....so the ambient probably dropped 10-15 on it's own.
correction 38* difference....both logs at very end
Last edited by JR92RS; 06-02-2023 at 11:54 PM.
ok i need help, i searched the forums looking for an answer but the only thing i've been able to conclude is that while maf tuning in open loop with stft's and ltft's disabled....what ever the narrowbands are reading doesn't matter.
but what i need to know is:
1.
with a commanded lambda of 1.00
and an actual lambda of 1.035
with an afr error ratio of 2.94
is it normal for b1 narrowband to be close too pegged out rich at 936mv, and b2 narrowband to be pegged lean at 38mv....and just hanging in those regions (no oscillating)????
or do i have some other problem?
2. during maf idle tuning i noticed the 3250hz cell is not populating any data....i'm using special functions to control rpm, so i can hit all of the idle area's of the maf range 1500-4000ish, and i'm using averages of 20 cell hits too populate
no matter how long i stay there nothing is showing up? update: just went back and looked at my own log....and data is there after the fact, but doesn't show while logging....that makes even less sense
including a log please help as i have stopped maf tuning to figure these questions out!!!
crickets all up in the forums again.....is it just me that has no life outside of working on cars?
Last edited by JR92RS; 06-03-2023 at 06:25 PM.
1) That's normal. Narrowbands are extremely sensitive to variation. A slight differential will cause that, and previous logs showed slight bank-bank imbalance. Remember they show rich or lean - Hi or Lo. Not how rich or how lean. That's why you'll see them oscillating during closed loop in a log as the ecu maintains fueling at stoichiometric. Hi-Lo being equal with respect to time averages out to the target value. Higher frequency of oscillation means quicker reaction of the control loop.
nernst o2.png
Don't forget that your wideband has been shown to be about 3% off, so actual lambda matches commanded.
2) No idea about that.
Last edited by SiriusC1024; 06-04-2023 at 02:49 AM.
thanks for the reply it's been awful quiet in here lately....all day yesterday i was chasing my tail. i would get 1500 thru 3000hz dialed in (afr gauge bouncing between 0.995 too 1.005) then when i made adjustments higher than 3000hz it would throw everything off and i would go in the opposite direction than i started from....like i overshot the mark. so i would drop back to step 3 and try to zero in on it again. i did that like 3 times getting to step 7 then falling back to step 3.
this is getting super frustrating
oh and i got this screen grab of the bug im up against when it comes to special functions and resetting stft's....hopefully calling them out brings a faster response with the ticket i submitted
screen grab problem.png
i'm also including everything i'm looking at around step 8 of the maf tuning, can someone please take a look...maybe something will stand out too you and you can give me a direction to go in (some step numbers may not match up disregard)
update: deleted all logs and tunes, starting my current tune back to step 1, and pulled 1% fuel from entire maf table...gonna try and sneak up on it again today. on a side note after a good heat soak on everything under the hood i think the highest iat was around 129* so the relocate of sensor maybe dropped about 13*
is there a good writeup somewhere that explains in detail every step too tuning the maf on a 3rd gen pcm? (which tables if more than one)
Last edited by JR92RS; 06-04-2023 at 08:12 PM.
What exactly are you trying to get working? Should be plotting EQ error vs MAF Hz and then copy-paste>multiply by percent onto tune table.
Here's what we'll do:
- Post the current tune file you're working with.
- Post a log of that file.
- Post a tune file of your changes to the MAF curve based off the log.
- I'll make changes to MAF and we'll compare.
Last edited by SiriusC1024; 06-06-2023 at 06:58 AM.
i was having zero luck going the eq error ratio vs. hertzs method to tune maf, and found a ltft+stft vs. mass frequency method i'm trying out now while dialing in the sd side. and when you look at the tune file you will notice some injector data missing....im doing a process of elimination thing trying to figure out what was interfering while i was trying too tune maf only, but currently i'm attempting to fine tune the original sd tune enough i can re-enable ltft and forget it.
Last edited by JR92RS; 06-06-2023 at 07:12 AM.
Zero luck? I was getting valid numbers. When it's close to 5% error try doing mulitply percent half.
ltft+stft works fine. Not going to get PE that way, though.
Last edited by JR92RS; 06-06-2023 at 07:19 AM.
I'll make it work if you want. SD or SD/MAF. How does the short pulse adder portion of injector data just disappear? Come on man tag me in.
I will do it for free. We'll start with SD. Get a log with the channels like you've been doing.
Since you'll be using SD/MAF we have to revert to the old tune OS with the new tune config applied. Can't use HPT SD Enhanced OS. I've made the changes already. Whichever you pick, for the first time, you must write entire. Tunes will be fine with a write config after that. Write entire is to change OS, and it only needs done once.
You may do both if you want to. I'll get back to it when I get time.
Last edited by SiriusC1024; 06-06-2023 at 08:25 AM.
Sorry. At first I had MAF and SD. I changed it to start with SD so we don't complicate things too much.
i flashed the tune in it's entirety (thank you for not making trans changes at the same time, makes me nervous changing too much at once)
i logged 3 drives today and applied 2 tune updates too the car (primary ve table) using the stft+ltft - map - rpm chart (fuel trims sd)
i will post all 3 tunes and the corresponding logs, i would like to know if i'm doing it right, or if i'm overshooting.
question on that process:
is the goal to paste special multiply by half as many times as it takes for each individual cell to be as close to (0.00) in the follow-up log as possible,
or is the goal to have a correct reading on the afr gauge....or are both true and the same time provided afr gauge is reading correctly?
notes:
under this tune for some reason i'm getting popping from exhaust on engine/trans deceleration (i'm using a lower gear to hit different boxes in the table, like 2nd @55 or 1st @20-25 then just come completely off gas pedal) it sounds like the car has glass packs, but it doesn't.....i can't imagine that much backfiring can be good for it, and what causes that?
the 4th tune file has not been applied yet, i figured that one would be good to compare too #3 tune file to see the changes i made.....just learned how to use the copy with axis function today.....wish i'd have known that a week ago
Last edited by JR92RS; 06-06-2023 at 07:57 PM.
We got similar results on VE adjustments.
Started from tune/log 3. Since VE is so close, I pasted-multiply percent-half LTFT then the same for WB EQ Error (30 count to remove transients) from 80kPa and up except for 400rpm. On the 80kPa row there's some blending of LTFT and EQ Error, so pay attention that something isn't being double-multiplied. From there I manually smoothed along rpm splines from each adjusted area outward.
Things changed:
- VE
- Fuel tabs
- Spark tabs
- WOT Trans shift points 5500>6150rpm
Next log of this go for a 30 min drive or so. Fill in as many fuel trim cells as possible. Lock it in 1st or 2nd if you have to. Slow throttle inputs. 0-floored pedal should take a 1-one thousand 2-one thousand count. That's the rate you should be moving your foot. Keep an eye on the wideband. If you're not seeing 0.9 lambda or less at WOT let off. This time only do logs for this tune, please. Post the tune file with the logs. Even though I have it I want to make sure we're on the same page.
Popping in exhaust is due to DFCO being disabled.
Last edited by SiriusC1024; 06-07-2023 at 07:18 AM.
i was not making any changes with wb eq error, just ltft+stft, what do you mean about the 30 count to remove transients?
and what do you want me to watch for....too make sure it isn't double multiplied?
you changed the trans to shift at 6150 or 5500?
i don't want to rev out the trans to where it could be damaged
here you got me confused....you said 0-floored pedal.....are you sayin do no w/o throttle pulls, or you do want a w/o throttle pull for 2 seconds provided she shows on rich side....is this to be a gentle cruising log or more aggressive driving one?
oh and the most important....she is showing like 20% lean on startup until it goes into closed loop, is there something in the tune that can richen up startup before damage occurs????
Last edited by JR92RS; 06-07-2023 at 07:50 AM.