wideband vs narrow band discrepancy

[pauly24]

Finally installed my wideband today.
Have it all working beautifully.

Logging the car, stock tune, just at idle in CL mode.
Commanded air/fuel ratio is 14.68 (of course)
But my wideband shows a pretty constant 14.88

Does this mean either my wideband is out, or my narrow band is out?

Will this affect my tuning? should I correct the figure firstly? (it will throw out the BEN factor wont it)

[mr.prick]

Where in the exhaust is it installed?
Do you have cats?
What WBO2 are you using?

Are you converting Lambda/EQ into your stoich AFR or are you using the default PIDs for AFR?
Most WBO2 give you the voltage to AFR conversion for 14.7:1 but some are 14.64: or 14.57:1
It may just a discrepancy between the EFILive PID expression and what your stoich AFR needs.

[pauly24]

The car is a stock VE SS. The cats are pretty much right after manifold (there is a narrow band in there) but I didn't place it here as it was extremely close and I thought it might get too hot.
So I placed it straight after the cat, just before the second narrow band (these car have 4 O2s sensors)

I am logging EXT.WO2AFR1 (I have no idea how it works, if it reads lambda and converts to AFR?)

I have a techedge 3A2 connected through serial to my flashscan V2.

Its a bosch LSU sensors, cant remember if its the LSU 4.0 or LSU 4.2, its the most common one though.

[joecar]

EXT.WO2ADR1 assumes stoich AFR is 14.7.

What mr.prick is saying is that there are various assumptions on the value of stoich AFR, and the differences between those assumptions introduces an error.

Log GM.EQIVRATIO and EXT.WO2EQR1 and compare those (just enable and log all the EXT.WO2xxxx pids, they don't cost any pid channel count).

If those are not identical then your tables need correcting.

[ScarabEpic22]

Im a believer in having the WB pre-cat, I think you wont get proper AFRs if the WB is post-cat under normal cruise situations. I could be completely wrong, but my LC1 reads almost dead on with the NB on that side and bounces between 14.6-14.8 AFR (expected).

[mr.prick]

The manufacturer defines the value of stoich for serial WBO2.
If it is programmable then you can match (closely) EQ1.

Where are the serial PIDs expressions? (http://forum.efilive.com/showthread.php?11596-Where-are-the-serial-PIDs-expressions&highlight=serial+expression)

[pauly24]

Thanks joecar/me.prick, Yes I understand now.

Lambda is the value that comes directly from the controller, so if I'm reading AFR, as you mentioned there is some assumption going on somewhere.

Once I log the lambda values from the controller and the commanded value from the ecu, I will see exactly what is going on, if these values are correct then my problem is the definition of AFR somewhere from when the computer converts lambda to AFR.

If these values are incorrect, I will have to reprogram my controller with new lambda/Voltage values to shift it over slightly.

Am I on the right track now?

Also is there any difference if I log EQ or Lambda? (im guessing no as one is just the inverse of the other so nothing can go funny in the calcs)

[joecar]

Sanity check your wideband: in CL or SOL it should show Lambda cycling tightly (very small excursions) around 1.00.

Then, after the wideband passes the sanity check, if wideband/actual/measured EQR does not match commanded EQR, then your calibration has to be corrected.

Either of WO2EQR and WO2LAM can be used.

[swingtan]

Try this.....

Log the NB voltage for the PRE and POST cat NB O2's. (GM.HO2S11 & GM.HO2S12) and watch how they react. This should indicate to you that for accurate AFR readings, you need the WB placed BEFORE the catalytic converter. The reason is that the nature of the CAT is to use residual oxygen in the exhaust to "crack" the exhaust molecules into less harmful gases.

Simon.

[pauly24]

Ok, my W02LAM1 stays pretty constant at 1.012 in CL mode (commanded lambda is 1)

This is probably because of the placement after the cat.

So what is the best way to correct the table, should I just shift the points .012 values across the voltage?

[joecar]

Place the WB upstream of the cat.

[pauly24]

is there another work around?
I just paid someone yesterday to install the O2 for me.

Ill be installing a new exhaust in a few weeks and I will have to re do all the system then.
So of course then ill place it in the right spot, but is there a way I can offset the value for now?

It seems i can only calibrate the analog output of my controller not my serial, is there someway i can offset the values in flashscan?

Can I create a calc.pid that takes the lambda value and subtracts 0.012, and then I can use this value for my BEN corrections?

[pauly24]

Ok I have created a calc pid which offsets the value of my wideband due to the error in reading it has by being positioned after the cat.

Does anything think that having this offset will be dangerous when it comes to tuning?

Im not sure if I should subtract 0.012 from the 1.012 lambda reading, or if I should subtract 1.2% of the lambda reading from itself?

Obviously if I went the first way, when I commanded an AFR of 13.23 my lambda reading will be 0.9, but if i subtract .012 my new lambda reading will be .888 giving me an AFR of 13.0536 but if I subtract 1.2% from 0.9 then my new reading will be 0.8892 which is AFR of 13.07.

Hmm I guess the difference isn't all that much anyways.

[WeathermanShawn]

Ok I have created a calc pid which offsets the value of my wideband due to the error in reading it has by being positioned after the cat.

Does anything think that having this offset will be dangerous when it comes to tuning?

Im not sure if I should subtract 0.012 from the 1.012 lambda reading, or if I should subtract 1.2% of the lambda reading from itself?

Obviously if I went the first way, when I commanded an AFR of 13.23 my lambda reading will be 0.9, but if i subtract .012 my new lambda reading will be .888 giving me an AFR of 13.0536 but if I subtract 1.2% from 0.9 then my new reading will be 0.8892 which is AFR of 13.07.

Hmm I guess the difference isn't all that much anyways.

I am not sure I would trust the accuracy of a post-cat location for WOT. Your Calc.Pid may address your offset at idle, but I highly doubt it is linear as fueling & exhaust temperatures change.

Anybody remember if post-cat reads leaner or richer? Seems like it reads leaner, but the exact relationship might take some more calculation..:confused:

[swingtan]

Pauly, the reading post CAT will not be a constant offset. The CAT works as a "chemical reaction" to "crack" the gases in the exhaust. As you may know, chemical reactions change rate depending on the heat involved. The heat will be changing in the CAT, so the amount and rate of reaction will also change. If you replace the upstream NB O2 with your WB, the ECM should default into OL mode as the NB will not be switching. When doing the VVE, you want to disable the NB O2's so it's no big deal.

Simon

[pauly24]

Thanks guys.

[joecar]

Which WB do you have...?

Some WB's can also do NB emulation at the same time as doing WB function (e.g. LC-1 and others).

[mr.prick]

Ok I have created a calc pid which offsets the value of my wideband due to the error in reading it has by being positioned after the cat.

Does anything think that having this offset will be dangerous when it comes to tuning?

Im not sure if I should subtract 0.012 from the 1.012 lambda reading, or if I should subtract 1.2% of the lambda reading from itself?

Obviously if I went the first way, when I commanded an AFR of 13.23 my lambda reading will be 0.9, but if i subtract .012 my new lambda reading will be .888 giving me an AFR of 13.0536 but if I subtract 1.2% from 0.9 then my new reading will be 0.8892 which is AFR of 13.07.

Hmm I guess the difference isn't all that much anyways.

Don't subtract, multiply actual Lambda by your stoich AFR or divide your stoich AFR by actual EQ.

[pauly24]

Yeah I've got a techedge 3A2, this does emulate narrow band, but i didnt want to cut into the harness, maybe i can make up a plug from an old oxy sensor.

The Holden VE O2 sensors have 4 wires, can someone explain whats what? I thought they would only have 2? Just the simulated signal from the techedge is just a single voltage so which of the 4 wires would I have to connect to?