A co worker brought me his 2007 Outlook XE 3.6L to repair for him after a transmission shop returned it to him not running. They replaced the torque converter and somehow in the process the engine "lost" compression. Im not here to discuss what they did or the circumstances, this is only relevant as the car was brought to me on a tow truck and it was still half way torn apart. After a leak down test found a majority of the valves were leaking.
I have since rebuilt the engine: honed the cylinders, new rings and bearings, sent the heads out for a valve job and surfacing. Installed a new timing chain kit, new plugs. ect ect.
The engine is back in the car and runs. Albeit with an issue. At idle it runs good for about 10 seconds and then starts stumbling with a flashing check engine light (p0300 random cyl miss fire) for about another 10 seconds, then back to running smooth. It continually does this loop unless i hold the throttle open, then it will run good. It seems to me the ecu is trying to make an adjustment but not getting the desired results.
Now for some data (only a cheapo scan tool not a $5k tool)
Claculated load at idle is ~42%
Short term fuel trim Bank 1 -33% (max range of taking fuel away)
short term fuel trim Bank 2 +33% (max range of adding fuel)
MAP 17inHg (indicating good vacuum, so no leaks)
Why are the fuel trims polar opposites bank to bank? When the car runs smooth and isnt miss firing, both values go to 0% indicating the ecu is in a base map where its not trying to adjust anything and it runs just fine. I have checked many times for vacuum leaks, replaced all the plastic hoses and hosed everything down with brake clean to find any vacuum leaks. Nothing. And with the MAP value in spec i believe there are no vacuum leaks. I swapped injectors from bank 1 to bank 2 and visa versa to see if fuel trims followed, they did not. Checked compression, 170psi (pretty good considering its not even broken in yet). Only code i get is p0300 which comes and goes. The exhaust is not restricted. Temp sensor shows correct data. Also note worthy how high the calculated load is at idle. on a quick drive under light acceleration and cruising its above 75%. On the same test drive i watched o2 sensor voltages and they were switching as they should. Bank 1 short trim came down to a reasonable number but bank 2 stuck at 33%, even though the sensor voltage was switching the correct rich/lean/rich switching.
Also note the engine does run just fine...when the ecu allows it. It starts right up, idles smoothly then runs rough. runs smoothly if play with the throttle.
I am at a loss, which is why i am here.