[shanec]

In trying to debug my infamous longtube header issue with the P0133 code (front passenger slow o2 response), I recorded this among other things:


This is a datalog beginning with cranking the engine and then driving a couple minutes.





Anybody understand o2 voltages? What's it mean when one of the O2's voltage data doesn't swing as low as the other? Notice on both banks, the peaks are about the same. But on bank 1 (driver or passenger?), the troughs (bottom side peaks) don't go as low as bank 2. What's that mean?

[kraigster]

Allright here's the deal,Bank 1 is on the passenger side, it is not switching as much as bank 2. What I would look at next is the fuel trims and see if bank 1 is adding fuel (more than bank 1) if they are about the same, then I would switch the 2 front sensors and see if the problem moves to bank 2. You can also ohm out the 2 sensor heaters and compare them, they are the 2 white wires of the O2 sensors. Good Luck !

[shanec]

Here's my fuel trim data. This was taken from the same run as the data above. These numbers are an average taken after the closed loop operation started (I logged a flag which indicated when that started).

Short term fuel trim Bank 1: 0.99332375
Short term fuel trimBank 2: 0.997336951

Long term fuel trim Bank1: 0.955347113
Long term fuel trim Bank2: 0.992085435


Not exactly the same. Close enough?

[kraigster]

It looks close enough. I am used to seeing Percentage values like + or - 1-5% not .995 or .999. I have to assume this is like a Lamda value where 1 is perfect stoichiometric. I would get under there with an ohm meter and ohm out the heater wires of the o2 sensors first and if they are within 1-2 ohms of each other then try swapping the right with the left and see if your problem moves.

[qwik48r]

close enough.. but very rich.

remember narrowbands are only really accurate right around stoich.. they really can only tell you if your leaner or richer than 14.7.

but your afr is around 11.8 going but these..

datalog you dynamic airflow, stft (disable ltft's if you can, as well as PE) and maf reference in Hz.. look up you maf table and see if the dynamic g/cyl is corrosponding this actual mass air flow Hz. scale it back where it needs the fuel out.

then, scale back (or up) my % multiplying stft's to primary VE...

set your PE to 1.117 to achieve a commanded afr of 12.5.. (easy, 14.7/1.17 = 12.5)

should pick up quite a few ponies... watch for KR.

[UrbanRedneck]

What would he learn from ohming the heaters? He has a response code, not a heater circuit code, and most of this data is from long after the 02s are warmed up.

For the response thing I think taht other guy told you to record the delay or something, I can't remember. But as far as voltage and switching, they look fine. They will never be the same left to right, I'd say they are close enough. But the slow response is the time from which the PCM switches to rich and the time that the 02 sees rich, and vice versa. All we see is what the 02 sees, not what the PCM wants to see.

[shanec]

There's delay-related data I have logged, but I'm not sure what it means. Here are their names:

ECU1:TOTAL_TRANSPORT_DELAY
ECU1:TRANSPORT_DELAY_FROM_TABLE_BANK
ECU1:LEARNED_TRANSPORT_DELAY_DIFFERENCE

Intuitively, you'd think the "table" number is what the tuner programmed in, the "total" number is what the PCM sees the delay actually is, and the "learned" is some adjustment the PCM has figured out it needs to make.

In the data, I'm seeing "total" is ALWAYS greater than "table". Not only that, but they track each other perfectly with an offset of .6 or .7 between them. So "total" is always some fixed number greater than "table". And..."learned" is always equal to this offset of .6, ,7, or whatever it is for that particular run.

So the data is empirically showing me that "learned" = "total" minus "table" and that this result is some constant number like .6. None of that jives with what you'd think these parms might mean.

Hopefully I can get SCT to tell me what these mean.