OK ... here we go ... get comfortable ... I got some stuff to say about this!
Believe you me, I've paid my dues with the things you now speak of. I also learned a thing or two but, as usual, I had to learn them the hard way, lol. Maybe, just maybe, I can save you some grief and frustration.
Here is my take on this topic and a little info on what I've been through looking for Closed Loop (cl) tuning nirvana, lol.
Dialing in the injectors using the slopes, bp, min pw, offsets, etc is nothing new.
I remember when the whole deal came about years ago from a guy named Walt on the Yahoo Tuning site. Everybody was all excited, including me, so we all loaded his Excell program and drove who knows how many miles around in cl as you most likely have done yourself.
I got the K's closer and closer but not as close as I felt I could.
Then it was discovered that there was a flaw in the program so shortly there after, everybody started to do it all over again using the Walt 2 method, lol.
I watched everybody else to see what would happen (might of been burned once but it wasn't gonna happen again) and sure enough ...... peeps seemed to be getting better results with the new and improved method.
I loaded up Walt 2 and had at it for the second go around and got the K's closer than the first time.
Now fast forward to a guy named Clint Garrity and his EEC Analyzer program.
It does the same thing but has a slick, modern look to it and is way more user friendly but after it is all said and done ........ I got no better results from his program than the one before.
btw ...I really do like EA and tell peeps to use it cause it will make their self tuning life more easy and all.
Now ... Here I am with a new ProM 80 maf, throttle body, and this new fangled EA program getting ready for my third go around at dialing in my maf transfer curve and injectors.
I quickly got the cl part of my maf curve good to go as the curve I built from ProM's sheet was almost dead on except for the idle points. I was a bit lean at idle but wanted to see if the low slope tweecs would take care of that.
I started the EA assisted tune with the following baseline stuff or values
j4j1 cal file
low slope@36
high slppe@30
oem t4m0 bp value
oem t4m0 min pw value
offsets@standard values for 30's
It did not take too long to get my K's to a range of 6% with, IIRC, 4% above the 1.00 line and 2% below. The first time I tried this whole process, I bet I spent 4 to 5 times the amount of effort & time for a 6% variance.
I got the K's centered with 3 up and 3 down pretty quickly but there was this nasty lean spike at idle or around the 1.5 to 2.0 ms area.
I followed EA's recommendations of slope values and got to 2 up and 2 down for a 4% range pretty quickly but that lean spike at idle still remained. With the Walt 2 method, I could get no closer than 4% K's and it took me so much longer than with EA.
I decided to see how close I could get to perfect 1.00 K's with EA since I had gotten this far with not too much difficulty on my part at this time.
At this point EA was telling me to make changes smaller and smaller the closer I got to the coveted prize of 1.00 K's. I mean we're talkin like I got down to 25% of 1%, lol. IIRC, the smallest recommended value EA spit out at me was something like 10% of 1%.
I got to 1.5 above and below and then, things just went to pot but I did not have an understanding of why. All I knew was the K's just went hay wire.
After looking at all my notes, double/triple checking all my pcm file values, download/upload Tweecer interface, yada, yada, yada, I decided my cal file might be corrupted.
Since I'm a GT boy, I had to get one of the Cobra Tweecer Site boys to send me a fresh bin download from their pcm and load all my values and, well, you know the drill, lol.
This time I got the K's to 1 above and 1 below all up and down the 1.00 line. I had been working on killing the lean idle spike for just a very short time and ......... dadgummit if it didn't happen again. The K's just went wild.
The adaptive was just going crazy but I had no explanation for it!
I asked the Tweecer Site guys if they had any ideas but none of them they gave seemed to make me wanna try any of them.
Here is what happened the 1st time the K's went crazy ...... It got dark!
Here is what happened the 2nd time the K's went crazy ...... It rained!
Here is the reason the K's went crazy both of those times ...... The battery volts dropped!
You See ...... Its all about the Function called ...... Injector Offset vs Battery Voltage
I did all of my data logging test drives in the summer time mornings and afternoons. This was with the radio, & ac working.
The two times the K's went crazy I had the lights or wind shield wipers working which pulled the battery volts (bv) down and the Offsets (O's) caused the K's to go crazy.
Now Then ...... If you have stuck with me so far ...... this post is already way too long for all the nitty gritty details so I'm gonna try to sum up things to the point that you won't make the same mistakes I did.
Briefly ...... the O's + or - fuel to try and keep your inj pw's stable if the bv moves one way or the other.
Can you not see the following reasoning?
You are only gonna get so close to 1.00 K's, like I did, and then if things change, like the bv's ....... the K's are gonna not be as close to the 1.00 line as they once were cause you tuned under different driving conditions.
Now at this point ...... We gotta go down a different path for a bit but we will return to this stuff and I hope you will see how it all comes together.
A bit of info about BV's.
I wanted to know what was happening with all this bv stuff cause ...... after all ........
How are you gonna improve on something if ............................
you don't know how the thing works in the first place, lol?
I saw a range of bv from a low of 8.xx to a high of 15.xx.
Before the next bit of bv info, remember my Stang is not a dd so that could make some difference to those of you who dd your Stangs.
It would take me 30 min give or take a few until my bv would stabilize. You could watch the bv be like 15.xx and over time it would settle to an average for what was applicable to your driving conditions. My Stang might sit for several days before I had time to do more testing and the battery needed that long to charge. For you dd Stang guys, your bv's would stabilize some what quicker.
Anyway ................
In order to gain the above knowledge, the research did take some time and btw .........
for those who are wondering about the value of the Tweecer's data logging feature .......
I would hope you can see how it came into play and the way it helped me gather the data to arrive at these findings in just this one instance.
Data Logging Rules!
Now that I had some data on what was REALLY going on, I was able to get back to hosing around with my tune and played with the O values to get the K's to to the point of not going crazy if the bv went + or -.
Summing up here .................. bet you thought this was gonna go on and on.
I ain't ever gonna type out all this junk again, lol.
You and only You can decide what is acceptable for the range of drift the adaptive strategy is gonna make under all different kinds of driving conditions.
I've seen some of the guys say 10% and I've seen some peeps get so frazzled with this subject ................. they admitted to just giving up.
My final reasoning was to tune close (2% variance) for my average driving conditions and get the extreme conditions as close as possible with the O's.
I almost forgot about the lean idle spike!
The O2's being mounted in the collectors of my LT's were ....... over a long period of time, causing the adaptive strategy to ever so slightly drift.
To fix the prob ...... I just took the easy way out and went to a method of ................
Tweecing things for an Open Loop idle which gave me total control over, well, stuff.
I'm tired of typing Guys and if you wanna know more or try it .............
The details are on my site.
I'm Done Now!
Grady