View Full Version : Rich Decel leads to lean bucking?


GuitsBoy
09-05-2006, 01:48 PM
Since I can remember, ive been tuning my VE table off of the averaged histogram. My deceleration seems to be very rich, and then when it slows down and starts to level out or when you slowly accel, it seem to lean out and buck like crazy. Im sure that my VE table is off quite a bit because the rich decel is is screwing up my histograms... Any Ideas? Im lowering and smoothing out my offset in the 13.0v - 15.0v rows in the injector offset table. Maybe transient fueling is the culprit here?

SSpdDmon
09-05-2006, 02:02 PM
Did you try tuning it with the NBO2 sensors since it's decel? What I do with mine is log a good amount of light throttle and decel. Then, filter out FTC's greater than 19 to make sure DFCO isn't interfering and also filter out STFT data points where -0.1%<STFT<0.1%. This will give you a better average by elminating all of the zeros.

Also with that cam, timing should be close to 44* for cells .16 grams/cylinder and under for 1000~2000rpm. At least, that's where I'd start with timing.

GuitsBoy
09-05-2006, 02:25 PM
Did you try tuning it with the NBO2 sensors since it's decel? What I do with mine is log a good amount of light throttle and decel. Then, filter out FTC's greater than 19 to make sure DFCO isn't interfering and also filter out STFT data points where -0.1%<STFT<0.1%. This will give you a better average by elminating all of the zeros.

Also with that cam, timing should be close to 44* for cells .16 grams/cylinder and under for 1000~2000rpm. At least, that's where I'd start with timing.

Sadly, NBO2's are useless since im in openloop-MAF mode. I do however filter out all AFRs below 10 and above 17 in order to weed out obvious bad data. The H/O timing from 1000 - 2000 rpm, from .08 - .16 g/cyl is as follows:

1000 1200 1400 1600 1800 2000
.08 43.5 44.6 45.9 46.9 47.4 47.7
.12 39.2 40.5 42.2 43.7 44.6 45.1
.16 35.7 37.1 38.7 40.2 41.3 42.2


The base tables look like:

800 1200 1600 2000
.08 27.0 40.5 43.7 45.1
.12 27.0 40.5 43.7 45.1
.16 27.0 37.1 40.2 42.2

SSpdDmon
09-05-2006, 02:38 PM
From what I understand, as it leans out changes in timing become ever more obvious. When you look at a log, does it show timing jumping around during the surging? I'm still learning the whole M6 side of things since you guys use the spark smoothing.

On an A4 when the TC is locked, what helps is to command one number for decel in that rpm range down to maybe 1200. That way as you change cells, the timing doesn't change and road surface inconsistencies don't trigger surging as easily. So, I'd try 43* from 1200~2000rpm for all .16grams/cylinder and under. Do the same in the base tables and see if that helps. The other thing that I've seen is to trick PE into coming on. If you set the PE Enable table to 0% for 1200rpms and under, it'll richen idle up for you.

Getting the car to drive smooth with a bigger cam like that is rather challenging I've found. But, don't get discouraged... :)

GuitsBoy
09-05-2006, 02:47 PM
From what I understand, as it leans out changes in timing become ever more obvious. When you look at a log, does it show timing jumping around during the surging? I'm still learning the whole M6 side of things since you guys use the spark smoothing.

On an A4 when the TC is locked, what helps is to command one number for decel in that rpm range down to maybe 1200. That way as you change cells, the timing doesn't change and road surface inconsistencies don't trigger surging as easily. So, I'd try 43* from 1200~2000rpm for all .16grams/cylinder and under. Do the same in the base tables and see if that helps. The other thing that I've seen is to trick PE into coming on. If you set the PE Enable table to 0% for 1200rpms and under, it'll richen idle up for you.

Getting the car to drive smooth with a bigger cam like that is rather challenging I've found. But, don't get discouraged... :)

I dont see much jumping around at all. Ill try a solid block of 43* up there. As for PE, Im running open loop, so i dont need PE to command a richer AFR. I can command it simply with the VE table without trims screwing with it. The problem is that I cant rely on my histogram logs to tell me which way to skew VE or by how much...

...And dont worry, im not getting discouraged. Ive been in the game for some time now, not my first time out. im just chasing down perfection. :)

GuitsBoy
09-05-2006, 06:56 PM
Im gonna try cutting the stock injector offset table in half to see if that aleviated the rich decel problem and lets me bump up the VE accordingly....

98Z28CobraKiller
12-02-2006, 03:46 PM
I dont see much jumping around at all. Ill try a solid block of 43* up there. As for PE, Im running open loop, so i dont need PE to command a richer AFR. I can command it simply with the VE table without trims screwing with it. The problem is that I cant rely on my histogram logs to tell me which way to skew VE or by how much...

...And dont worry, im not getting discouraged. Ive been in the game for some time now, not my first time out. im just chasing down perfection. :)

Have you achieved perfection yet? I threw in the towel and am waiting for some 42's right now.

GuitsBoy
12-02-2006, 05:46 PM
Have you achieved perfection yet? I threw in the towel and am waiting for some 42's right now.
sadly no perfection - not yet. I have gotten a little further playing with the transient fuel > minimum fuel setting until decel no longer runs rich or lean. Helps dialing in AFR easier and more accurate.