Intermittent Erratic AFR's

[trapstar31]

Hey, this is an issue that cropped up a week or two ago. My AFR readings become very erratic and seem to stay that way until the car is parked for a while (few hours?). I've tried upgrading my firmware, and re-downgrading but it doesn't seem to have helped. I have an Innovate wideband that's only a few months old, hooked up to the EMU analog input. I tried recalibrating it after this issue cropped up but it didn't help.

These logs are from the same night, good readings on the way downtown, then parked for 15-20 minutes, then driven home with the erratic AFR readings. I'm at a loss for where to begin troubleshooting this.
Since my one log was over 2MB, I uploaded them here:
erratic readings: http://s000.tinyupload.com/index.php?fi ... 8661498914
good readings: http://s000.tinyupload.com/index.php?fi ... 4817183067

[PSI-Motor]

Innovate MTX-L ?

[trapstar31]

Innovate MTX-L ?

Yep.

[PSI-Motor]

Seen a lot of them fail on the analog output.
Check if the voltage signal out of the guage is stable.

[trapstar31]

Yeah it basically mirrors the lambda reading computed by the EMU. I've noticed the gauge itself showing these erratic readings as well so I guess it's possible the gauge/sensor itself is failing, which is really frustrating but at least it isn't something worse. Really not impressed with this gauge or the AEM one it replaced, both POS's.

Should have just bought the EMU wideband addon after my AEM gauge died instead of the stupid Innovate one

[nickbmw]

Ι suspect you had the same problem that i had. If the iat sensor is heatsoaking and shows very high numbers when you drive in traffic, or park it for 10-15 minutes and drive it again after, the afr then becomes very lean and unstable, until the iat shows lower temps. My Honda S2000 shows up to 70 degrees (yes, Celsius) with the stock sensor if you let it idle for a while, and the afr is going very lean. I installed a new iat sensor at a different position and it solved my problem as the temps now are closer to actual and it doesn't heatsoak. If you change the sensor and position, you will need to redo the ve tables.

[Jadzwin]

For heat soaking you should use air charge temp. table .

During the idle the temperature of the air depends on the air temperature and the engine temperature (the air is heated by intake manifold, valves and so on.)

Air charge temperature table is created for proper configuration of this phenomenon.

[nickbmw]

For heat soaking you should use air charge temp. table .

During the idle the temperature of the air depends on the air temperature and the engine temperature (the air is heated by intake manifold, valves and so on.)

Air charge temperature table is created for proper configuration of this phenomenon.

I would agree if the chargetemp was a blend between the ambient temp and the intake manifold temp. But it is not, and using the ect as a reference point and chosing a percentage of "gravity" between ect and iat, the higher the percentage towards the ect, the more you choose to just ignore the iat sensor. Once the engine is warmed up, the ect is pretty stable. Using the charge temp table doesn't solve the problem, it just makes it less noticeable. If you offer the change to blend between ambient or airbox temp and iat sensor, it will be much better. Chargetemp via ect is just a bandaid.

[trapstar31]

Nah mine is erratic even at IATs of 20 degrees C. The only time it pretty consistently isn't is under boost. I'm starting to think that it might be the MTX-L's heater circuit to blame