One of my frustrations with DIY tuning has been that my BLM's are reading that the car is running lean so the computer dumps more and more fuel until the car barely idles and smokes from being too rich. Can this be caused by a leak at the head-to-header seal sucking in air? Is there a way to check where the leak is?

 

Look for vacuum leaks. One way to tell if you have a vacuum leak is to look at what your IAC reads at idle.

 

Yes, the O2 sensor is very sensitive to any air getting into the exhaust. If you partially block the exhaust outlet that will increase the pressure in the exhaust. It may be enough to help in finding any leaks.

RBob.

 

(sorry), on a side note, what would the iac read at idle if there was a vacuum leak?

 

It would read zero or close to it to compensate for the leak. I think at idle you should be at around 30-40 counts? (guessing here, I don't remember the actual spec)

 

What you said about IAC & vac leak.

For TBI, a low IAC count of 5 - 10 is best.

For port, an IAC count of 25 - 35 is good.

These are both with no additional loads: in P/N, A/C off, backlight heater off, etc.

RBob.

 

Well I finally did some datalogging from a cold start and just let the car idle and what I noticed is the integrator value wandering up and down. As the O2 cross counts went up to 255 the INT value would also go up to about 160 and the car would go very rich. Then when the cross counts reset at zero, the INT would start to decrease to about 80 and the idle would smooth out. Then the cross counts reset to zero again and the INT starts to climb back up. It would never even out. I could also see the injector PW vary from about 2.5mS to 5mS. What am I missing here? I'm not really sure what to do with this now.

Items of note: In my tune I have the "Idle O2 Rich/Lean Threshold Fast" set to 500, I have a heated O2 sensor, my MAT sensor is just hanging next to the distributor because the LT1 intake doesn't have a spot for it so the MAT gets pretty high but the INT wandering happens long before that. IAC counts at idle are 45-55, depending on the INT at the moment.

Here's a graph of what it's doing and I've uploaded the .adl and .xls file of the datalog, and also the bin file, if anyone cares to have a look. Thanks for any help!
http://www.fabiens.org/alex/V41024a.BIN
http://www.fabiens.org/alex/0918a.xls
http://www.fabiens.org/alex/091810a.adl

 

what code are you running? what ads file should I use?

 

730 ads, $8D, SAUJP xdf.

 

Ok, will take a look tonight........

 

looks wierd, I would start with raised the timing (24-26) and locked it at idle too see how it responds to that.

 

It looks like once warm my total spark advance is already around 26.

RBob mentioned in another post I found that the issue may be with the proportional gains. I need to research that further I think. I just seems like the ECM doesn't respond to the INT or BLM quickly enough to adjust fuel. It lets it fatten all the way up and then goes the other way and leans out.

 

Dont have TP in front of me but you could try to slow down the blm some.

I didnt look in your bin so much but I dont think the ecm is to slow for your setup....wierd as it switches from rich/lean as soon as the CC starts over try go back to a "virgin*" bin?

I would try to lock the timing sp you know it wont affect it, its a simple change........