Idle fuel smell with trims +/- 4%

[JimMueller]

I helped a friend tune his '98 car yesterday. He has heads, a Vinci 060 cam, Pacesetter headers and no cats. He doesn't have a portable wideband, so I disabled the MAF, DFCO, LTFT, LTIT and reset the fuel trims. Then used STFT's to tune his VE.

After three sessions the LTFT's were +/-3% and the STFT's were +/- 1% at idle, yet the smell and fumes from his exhaust was obnoxious. I'm not sure if it's completely related to not having cats, but what else can I check/log to see if it can be improved upon?

[Bill@HPTuners]

Pacesetter headers and no cats.

You answered your own question, with no cats its going to smell like that.

[svede1212]

that's just the smell of old school muscle car

[jimmyblue]

It doesn't have to be eye burning stank, thing is trims work to
an average and don't care if the mixture spends half its time
pegged fat and half pegged lean, or boogeying tight to center.

Also trimming to a narcoleptic O2 sensor's output can give you
nice -looking- LFTFs but a whack actual mixture.

Observe the real time O2 sensor waveform and whether it's
"tight" or "sloppy" (bottoms & sits for a second, then rails high
and sits, ...).

[JimMueller]

It was sloppy at idle. It'd hang out between 200-300mv for maybe 10-15 seconds, sometimes spike to 700mv, then drop back. It definitely wasn't a pretty sine wave. He bought the car used and as far as he is aware they are original O2's. But he's had to spend some unexpected funds on the car and I didn't want him to go down the path of buying Bosch 13111's until we covered all the bases. Are they still the O2 'to have'?

It'll be a couple of weeks before we can get together to look at it again.

[jimmyblue]

I have the NGK 'vette rears and still got a pair of switching codes.
But they do switch, still looking kind of soggy though.

One thing I've been meaning to do, is play with the transport
delay and proportional fuel scaling factors to see if I can get
it to swing tighter.

Saw something on LS1Tech referring to a Honda 'board, where
people were taking those anti-fouling adaptors and pulling the
O2 sensor -out- from the pipe. Not sure why except for the
idea that this might cut the conductive thermal losses, back
to the header wall and maybe allow the heater to keep the
sensor hotter. Seems kooky but the rice-boys seem to think
it's helping them (but then, they probably have Turbonators
too).

Anyway I think from the sound of it sleepy O2s are the problem
and the question is, can it be fixed for just brainpower (tune)
or cheap (wrap, adaptors).

I haven't seen any good guidance on how to play with the
transport delay but I'm thinking that just scaling it by the
ratio of primary tube length to manifold O2 distance-to-
head, would be a start. Not too clear on which way to go
with the proportional gain either but cut-and-try, I guess.
Maybe make up a rack-'o-tunes with 0.25, 0.5, 2 and 4X
just for that and see if anything affects switching waveform
quality observably. Hopefully someone who's done this with
success will chime in and point the way.

[Billiumss]

I haven't seen any good guidance on how to play with the
transport delay but I'm thinking that just scaling it by the
ratio of primary tube length to manifold O2 distance-to-
head, would be a start. Not too clear on which way to go
with the proportional gain either but cut-and-try, I guess.
Maybe make up a rack-'o-tunes with 0.25, 0.5, 2 and 4X
just for that and see if anything affects switching waveform
quality observably. Hopefully someone who's done this with
success will chime in and point the way.

I asked about this almost a year ago:

http://www.hptuners.com/forum/showthread.php?t=8765

[JimMueller]

Did you come up with a method to properly alter the integrator table?

[JimMueller]

How about finding out which airflow mode were in when the symptoms are occurring and then lowering the O2 switchpoints?