NefMoto

I've been trying to figure out the source of this very random issue for a while now, and I finally have logs of it. 

My engine runs very well, except for a few random occasions where everything is way out of wack and the car idles for shit.  This has happened like 5-10 times in the last 6 months and this car is a daily driver. If I clear codes with VCDS, all returns to normal immediately.

When this issue is present, engine runs very lean, o2 sensor voltage is very low, and wideband reports lean as well.  Attached, is the "bad" log, and then the "good" log was taken right after I cleared codes with VCDS.  It's like there's a big vacuum leak that I can fix with VCDS.

Fuel trims are just fine, before and after in this case, but I have seen it affect trims if it's in this state for too long. I am stumped as to what would do this.  Hopefully someone has an idea.

80+ views, 1 download of my logs.  Does this post need more pictures!?!   :P

My observations from the logs:

In both cases, injector duty cycle and rpm are the same.   How can this result in such a lean (and different) mixture, if in fact there is not a large vacuum leak?  Perhaps evap valve is opening, and clearing codes in VCDS closes the valve?  I have never had codes that seem to indicate a evap leak.

You can see in the "bad" log that o2 sensors (wideband and narrowband) still respond when I go into open loop (full load) fueling.

clogged injectors, failing fuel pump, clogged fuel filter?

Thanks nyet, but why would clearing codes in VCDS fix the issue immediately?  Again rpm/injection time is the same in both cases.

Fuel pump is original, filter is new(ish), injectors are new. I don't believe any of those are the issue here.

Here's a couple screen shots of both of these plotted.

Notice how the injection time and rpm are the same, although MAF is a tad higher in this state.  I might attribute the raise in maf signal to the engine running rough.

In the good log you can see the o2 signal oscillating, in the bad it's flat-lined at idle.

I'm just not sure how to interpret this.  The engine runs terribly when this happens, as you can see from the logs.  Is there anything else that I should try to log when this is happening?

It is as if the ECU doesn't think the O2 sensors are ready, and it is just running open loop.

Perhaps some warm up or o2 heating state that is being set when it shouldn't?

Might look through the FR to see if there is something blocking DAC readings from getting to usvk/ushk

Yes from long ago.. but I Just wanted to say from 200-500who maestro I never got 02's working consistently. I swapped 02's obviosly. Wires. everything. They would just be 0 stft&lt Wish I had something useful to add like a solutioon, but all I can say is you weren't going crazy.