Loading-up at idle? WB02 getting richer on extended hot idle

[Mr. P.]

Has anyone run into this before, and what can I look at to diagnose the issue?

When I start the vehicle, it idles fine and WB02 readings are within 1% of commanded EQR - but as the vehicle continues to idle, the WB02 reading gets richer, and richer, and richer like it is loading-up with fuel.

Does this mean my injector small pulse adjust table is wrong? The alternator voltage never changed the whole time, so I don't think the Injector Voltage Correction table is involved. Maybe injector timing, or timing trim, or ???

Maybe I need to do a RAFIG? I would not think so, because the long-block is bone stock LQ9.

TIA for any ideas/feedback!

Mr. P.

ps - log file showing the problem: Garage Hot Idle.efi

[joecar]

Are you running a COS in OLSD...? Then maybe look at table A0014.

[mr.prick]

^+1^
CL or OL?

NBO2's show rich but Lambda/AFR is lean.

[joecar]

Oh, I meant with MAF failed...

[Mr. P.]

Yup gents I am running COS-3 in CLSD. My O2's are getting really old, I think the heaters in them have gone bye-bye and they're not switching normally until the vehicle has been on the road a few minutes and the exhaust gets them up to operating temp.

Mr. Prick you are right, I spoke backwards describing the problem, the WBO2 is leaning out, not loading up.

Attached is a copy of my {A0014}

Mr. P.

[mr.prick]

WBO2 and NBO2 contradict each other so choose whichever one you feel is correct.

The problem w/CL is you need NBO2 to be accurate.

[Mr. P.]

You think that's screwey - look at my LTRIMS, I'm not so sure the PCM is running in closed-loop - the O2's are switching, but the LTRIMs never report anything but zero. The only DTS code I have is P0102 (MAF is not present), there are no other codes present. It takes about 5-minutes of driving before the NB O2's start switching (the WB02 works fine), my commanded EQR is 1.00, and the truck is getting the most awesomest MPG ever (18.8 @ 70-MPH yesterday, originally it was 14) so I was assuming that the PCM was fueling in CL. But when I look at my LTRIMs they are ZERO, and my VE tuning certainly isn't that perfect!

One thought, I did flash the truck immediately before making this logging run - do you need to drive for a while before the PCM starts calculating LTRIMs?

* 4th of July_0003.tun
* Apt to Chevron.efi

Mr. P.

[WeathermanShawn]

You have LTrims B3801 set to Disable.

[SSpdDmon]

To answer the underlying question, AFR will deviate from commanded for a couple reasons as temps change. I think the two biggest reasons are:

1) Actuals vs. Reported: Sensors will suffer from heatsoak and other deviations as a vehicle sits over longer periods of time. Because the density of oxygen molecules in a given volume of air changes with temperature, fueling requirements will vary as well. If the PCM is told that a given amount of air is entering the cylinders - but in actuality there's less or more going in - you are going to be rich or lean (as measured by the O2 sensors).

2) PCM modeling: I think the engineers did a killer job trying to model the physical nature of our world. At the end of the day though, it's still a model that will work better in some conditions and worse in others. As you deviate from the normal operating temps (specifically IAT), I'd argue that the model isn't as strong/accurate. However, it's still close enough to allow for the parts of the PCM that account for varialbe change to manage those inaccuracies (fuel trims, idle trims, etc.).

This is my hypothesis based purely on my observations...

[Mr. P.]

You have LTrims B3801 set to Disable.

DOH! Rookie move on my part! And I have the STFT disabled too, no wonder I don't have a good warm idle lol.

I'm going to keep them disabled as I continue VE tuning.

To answer the underlying question, AFR will deviate from commanded for a couple reasons as temps change. I think the two biggest reasons are:

1) Actuals vs. Reported: Sensors will suffer from heatsoak and other deviations as a vehicle sits over longer periods of time. Because the density of oxygen molecules in a given volume of air changes with temperature, fueling requirements will vary as well. If the PCM is told that a given amount of air is entering the cylinders - but in actuality there's less or more going in - you are going to be rich or lean (as measured by the O2 sensors).

2) PCM modeling: I think the engineers did a killer job trying to model the physical nature of our world. At the end of the day though, it's still a model that will work better in some conditions and worse in others. As you deviate from the normal operating temps (specifically IAT), I'd argue that the model isn't as strong/accurate. However, it's still close enough to allow for the parts of the PCM that account for varialbe change to manage those inaccuracies (fuel trims, idle trims, etc.).

This is my hypothesis based purely on my observations...

Point #1 - I disagree strongly; I cannot see NB02's response changing that much due to heat soak; now the NB's cooling off is another issue entirely - but from my current knowledge/understanding, once a NB is hot it works fine, more heat & heat soak is good for them until they are so hot as to be permanently damaged, and that's beyond 'heat soak' IMO.

Point #2 - Agree completely; but don't we have IAT-based fueling modifier tables to account for this? Isn't this what the A0014 & B4901 are supposed to be for??

Mr. P.