Mysterious AFR issues at cruise.

Supraman

New Member
Mar 30, 2005
64
0
0
Sunnyvale
While my car has had plenty of strange problems, this is the most mysterious one I've run into.

First, my car is running MAFT-Pro Speed Density. It has been running well with the setup for perhaps 1,000 miles, including a 500 mile road trip that was nearly all cruise and did not show this problem. I noticed the problem spontaneously while driving a short time ago.

While cruising, especially at engine speeds around or above 2,000RPM, the AFR can show one of three behaviors. If I am very light on the throttle, such as 4th gear at 2,000RPM barely maintaining or even losing speed on level road, the car will run close to stoich (14.8-14.2). If I maintain speed or push enough for a very very tiny acceleration, my AFR instantly hits the 12s. Any reasonable acceleration (we're still talking very slow, less than 25% throttle) will immediately yield AFRs around 10.0. I do not notice any missing, which would produce falsely rich readings. Idle is usually close to stoich, but after driving around for 10 minutes or so, tends to move to the 12s and stay there. All of the readings I mention are with a warm engine.

Adjusting fuel pressure or MAFT-Pro settings has not helped. I initially suspected the oxygen sensor, even though no codes were evident, so I wired in a simulated signal from my wideband. The problem remained unchanged. I next suspected the TPS, but it passes the TSRM diagnostic, assuming I read the diagrams correctly and E2 is the bottom pin when looking at the unit from the front of the car.

And finally, the car's performance at WOT is simply golden. Great AFRs, blistering performance. No complaints whatsoever. That implies to me that the ignition system is strong, increasing mysteriousness :p

Any ideas would be wonderful. At this point, I'm quite frustrated, despite my extreme love for this machine. Thanks all!
 

Supraman

New Member
Mar 30, 2005
64
0
0
Sunnyvale
cjsupra90 said:
What wideband are you using and where is the wideband sensor located?

I have an LM-1 with the Bosch 4.2 sensor. It's mounted at the bottom of the downpipe, at the bend. It has been working great since I installed it, shortly after getting the car.

Mods are as below, but with the E-Manage torn out and replaced by MAFT-Pro speed-density (haven't updated the image yet).

supraspecifications.gif
 

Supraman

New Member
Mar 30, 2005
64
0
0
Sunnyvale
I took it for another drive, and it continues to behave strangely, as though in open loop. Indeed, I have observed that the Vf as measured by the MAFT-Pro is constant at 0V.

The two reasons I know for this are a) Open-loop operation, especially due to O2 failure, and b) Very rich condition, full fuel subtraction. I'm wondering why, if A is the case, I have no codes--and moreover, using a simulated signal from the LM-1 didn't work (assuming my wiring was good, which I believe to be the case). And, I don't understand why B would be the case, especially because the car sometimes floats into lean ranges such as 16.x for a while after returning to idle, or during some accelerations.

Any ideas?
 

Supraman

New Member
Mar 30, 2005
64
0
0
Sunnyvale
drjonez said:
PT AFR tracking.
PT tracking is not currently enabled, because I don't have the wideband connected to the MAFT-Pro yet (found the analog out cable just before this problem occurred, and figured tracking is not important until the car runs well without it).

Perhaps I could run the ECU in open-loop and have the MAFT-Pro do PT control; has anyone done this?
 

Supraman

New Member
Mar 30, 2005
64
0
0
Sunnyvale
drjonez said:
might want to double check that AFR tracking is indeed off....

Last night I wired up the LM-1 to the MAFT-Pro, and set all of the tracking features to be triggered by TPS greater than 4.9. My max WOT TPS voltage is 3.33, so that should disable it. I don't think they were 'on' before because the configuration was set to no wideband, so that should disable the features by default.

I checked the narrowband O2 signal wire, and it has good continuity from the sensor to the ECU plug. Is it possible that anything is wrong with the heater circuit, or would that throw a code for sure? Any other ideas, or part-swap-test recommendations?
 
Last edited:

Supraman

New Member
Mar 30, 2005
64
0
0
Sunnyvale
Well, the car is back to health, after I reconnected the stock narrowband signal wire (having cut it earlier to see if a simulated signal from my wideband would fix the problem). Continuity to the ECU is good; I did not check it before. It did seem like an O2 problem all the while, though it is disappointing that the ECU did not throw a code.