snflupigus

First question...

I have been running around tuning AE in open loop VE learn from wideband for a while now. Today I switched both idle and cruise back to closed loop and drove around. The car ran fine and I datalogged about 20 minutes. Then I VE learned on playback and it is trying to take out 3-4% fuel in a bunch of cruise locatations above 1k rpm but add in 5-7% on decel and idle locations below 1k rpm.?

My car idles perfectly now and I have no intentions of adding fuel back to those VE tables for idle. As it is now It spits out black spatters on the concrete when cold idling. I think from all the carbon still burning off in the cat which I think needs replaced.

Suggestions?

2nd question...

which is going to be more accurate and run better? Running constant open loop wideband learn or closed loop blm's incremental learn? Some of my blms went to 148. This is after it had everything damn near pegged with the wideband learn.??? Leave the wideband AE tables tune and forget the blms?

RBob

iTrader: (1)

The real reason for closed loop fueling is the catalytic converter. It requires an oscillating AFR to properly operate. Which is obviously an emissions item.

Running open loop does make a difference in driveability. The vehicle runs better, as is the transfer from closed loop to PE/WOT mode. There isn't a sudden step.

I'd say try it both ways for a while and switch back & forth. Then post what you think of each. Quite a while back I had folks telling me that open loop wouldn't work. But I tried it anyway and found that it ran rather well. And the vehicle not having cats from the factory didn't require closed loop.

RBob.

snflupigus

back.. from the... dead...

New intake w.out egr. Has been running for 2 years (but only about 3k miles) in open loop (using the wideband 100%). I've never been able to get back to the smooth idle I had before. It will idle fine for a few minutes then start to hicup and dip, recover and repeat. Sometimes it dies. It doesnt matter if its in park or drive - same thing... and lately it dies a lot if i go from park to drive without tiping the throttle a bit before shifting which I dont think my transmission likes...

Ive tried to add fuel at low rpms, move the target idle for IAC in both drive and park. Even when they are the same i get a huge dip in rpm from park/neutral to D. I've tried tweeking very slightly the timing at idle too all while emulating and not emulating loading to the emulator to run. TunerproRT crashes everytime I alt-tab from the dashboard to TP. annoying and is the only reason why I havent paid for it.

On decel enlean it will surge a tiny bit sometimes and in cruise control it has a VERY slight rpm surge that I lived with. Other than that it doesnt hesitate when I stomp on it.

who has ideas of where to start.

graebz28

iTrader: (5)

Are you using the option to use a constant SA at idle? I found it good to keep most VE and SA cells the car idles at very close if not exactly the same. I also backed off the amount of timing pulled when trying to correct idle and tuned a lot using a vacuum gauge. With my idle SA now at 28, I actually got the map down to 40 (T-56 car).

snflupigus

sa and map fluctuate slightly - i think it was a fuel issue. Since i turned off the wideband idle and let the computer adjust the old fashioned way it seems to be idling more smoothly with less hicups. It just took a while of driving stopping and going I think for it to put some fuel back in. I think my fuel table had too little fuel and so it kept having to compensate for it.

so... heres my graph from emissions a few minutes ago. looks like its accel that needs fixed. but all the gasses spiked? A bit confused I think i'll attack the CO first

Dominic Sorresso

Snuf,

I fought something
like that for over a year on my old 84 Xfire.
The motor would sit and idle fine for a few minutes then
and then begin to oscillate. Sometimes it would then shutoff.
It would start again and it appeared to cruise ok. I later discovered I had a
head gasket failing, having used the FelPro 1010 when I installed my
TFS23 heads.
Also, I did have a similar issue with VE Learn putting fuel back in the
Decel Region. It just kept adding and adding. I later found out I had a vacuum leak at the muffler and decel was drawing fresh air back through the exhaust
and causing the O2 sensor to report lean. Once I fixed that leak, the VE Learn began removing fuel in those same areas.