RomRaider

Hey All,

Recently I was looking through the forums to see if anyone had managed to crack the code for the O2 sensor gain / PID logic. Seems many people have had issues with this, especially after exhaust work where the O2 sensor had been moved much closer or further from the exhaust ports.

Seems none of the other threads ended up revealing anything except for the fact that the AF Learning Limit vs ECT tables might be clipping the I term of the CL routine, but I think its definitely something worth looking into considering how many people have run into this issue.

I started playing around with my car (EZ30 06 Liberty, D0XJ001M) to see if I could turn up anything. I recently had a stainless Y pipe installed to match my headers and during that process the front O2 sensors (Sensor #1 and #2) were moved further away from the Exhaust ports vs stock, causing more extreme AF oscillations especially at idle. My engine was now prone to stalling in some cases due to the correction going extreme in the lean direction after the exhaust work. Hence I suspected that O2 gain was to blame here.

After a bit of work, I was able to locate a table in my ROM @ c5dfc, directly above the OL AF Target Tables (See Screenshot 1 - My table has slightly tweaked values in this SS vs stock). Its a 3D table scaled with RPM and raw Mass Airflow, and based on its values, I suspected it might be a transport delay table used to tell the O2 routine how extreme the delay in exhaust gas reaching the sensor will be (Perhaps the units are revolutions or ms). As a test I set all its values to 1.0 and this caused the O2 oscillation to become so bad that at idle the car wouldn't run anymore once the correction got into a tank slapper. IMO that's suspicion confirmed, and after increasing the values at low airmass I was able to stabilize my Idle AF, and my car no longer tends to stall. I'd be interested to see if this table can be located in any other roms.

I also think I may have found the CL o2 gain tables directly below the CL AF Target Tables (Screenshot 2), which in my case are also 3D and scaled with MAP and RPM. I recently lost my good disassembly due to a drive failure, but I suspect that the tables are selected between based on AVLS State.

Happy to give anyone any info that might help shed some light. I'll be testing the suspected gain tables too soon and I'll report back!