i recently just switched back to closed loop because open loop idle was driving me nuts in a daily driver. anyways, everything's functioning right, except my afr is always about 14.3.. not around 14.68 in closed loop. i thought adjusting the switching points would be my answer to getting the narrow band sensors to be more around stoich, but no dice. raising or lowering switch points seems to have no effect. what gives? what can i do to solve this? its all the time like 2-3 % rich...

 

What is your commanding afr at idle? There is a setting in the tune that set the max afr lean limit for idle. Perhaps you adjusted it to 14.3ish.

 

uhh, i thought in closed loop the computer automatically commands the afr?? but i can stil see it on my scanner.. 14.68 is what its commanding in closed loop once warmed up

 

It is entirely possible that your wideband and/or NBO2'a are not accurate. A few percent isn't that much error for a wideband. What my NBO2's think is stoich is not what my wideband thinks is stoich, but it's within 1-2% though.

What wideband and NBO2's are you using? Mileage on them?

What switchpoints did you try?

 

using afx wideband, mileage on it is just about 10k maybe... stock 02's... prolly 90k on them. wideband is spot on, just recalibrated it. my afr's never see stoich no matter what i set the switchpoints to. it's always 2-4% off. i tried stock 450mv, 500mv, 550mv, 400mv, and 350mv.. doesn't seem to do much to be honest

 

Yea changing the switch points doesn't do ****. The 14.3 vs 14.68 is just an error between the two different sensors your using. It is pretty normal for the wideband and the narrowbands to not agree with each other. Nothing to worry about. Just make sure you use your narrow bands/STFT to tune idle and part throttle cruising and use your wideband for WOT regions.

 

My point was that all you're doing is comparing the the wideband to the narrowband O2's, neither of which may be indicating exactly stoich. I'd trust the NBO2's to find stoich over the wideband since that is their only purpose in life. However will 90K on them I'd think about replacing them.

A guy in efilive forum did a test between different NBO2 brands and found that they all had different opinions of what stoich was as indicated on his wideband.

You might also consider that your fuel may have a small percentage of ethanol in it, which will lower the stoich AFR of the fuel. I adjust the stoich AFR in my tune to account for the amount of ethanol in the fuel, which is usually around 8% here

For reference, when mine is running in semi-open loop (using STFT's only) the wideband averages out to 1.012 lambda.

 

Ferocity pretty much nailed it. The narrowbands are much better at their job of finding stoich than the wideband is. So if the narrowbands are switching properly (look at the voltage swings) then your wideband is off.

Also the fuel content will not affect what the wideband says. If e85 is burning at stoich the wideband will show whatever it thinks stoich is.... ACTUAL stoich for pure gasoline thats 14.68, gas with e10 is around 14.42, e85 is like 9.5, but the wideband will show (if its calibrated to gasoline) 14.68 for all of them. Widebands use lambda and convert to afr later. Lambda =1 at stoich by definition.

 

As an additional point, I would encourage you to start thinking in lambda. Lambda=1.0 is stoich for ALL fuels and mixtures of fuels, whereas AFR=14.68 is stoich only for gasoline. When you start getting into ethanol blended fuels, E85, methanol injection, etc, it becomes much more difficult to think in terms of AFR. Tuning with lambda is much simpler.

Unfortunately the AFX can only display AFR, but you can have it show up as lambda in your logging software. A wideband is an lambda sensor, but it is converted to AFR for you. The AEM wideband is nice because it can display lambda.

As per the AFX manual, "The AFX’s display is designed to show AFR values based on a gasoline scale with 14.57:1 AFR as the stoichiometric ratio."

So the AFX assumes stoich is 14.57, not 14.68 like you mentioned. So that should cut down on your error slightly.

 

if you are using HP Tuners check out this write up https://www.performancetrucks.net/fo...lambda-524099/