Narrowband? Wideband? Exhaust leak? Running Lean.

[JamesTheriot]

2003 C5 6 Speed. Stock car, Cat delete, AEM 30-0300 wideband.
I tuned the MAF and VE tables separately with fuel trims off.

When I turn the FT back on the car pulls ~8-10% fuel out right away. Now since both narrowbands agree I'd want to think my wideband is wrong or I have an exhaust leak after the O2. The only thing making me thing differently is if I command the desired AFR with my VCM scanner and Closed Loop disabled, My wideband almost perfectly follows my commanded AFR. It's as if both my narrowbands are skewed and want to cycle around 16.6:1 AFR.

I had an idea that since the stock stoic is set to 14.63 and modern fuels have more ethanol than in 2003 that changing the stoic to 14.08 would fix the issue, but when I tried it made no difference.

I'll post my stock file, current file, drive log, and idle log of me controlling the commanded AFR.

JAMES C5 V003 MAF+SD FINAL+PE ADJUST.hpt
1G1YY22G235125468 2003 Corvette 6Speed stock file.hpt
weird o2 drive.hpl
idle log, scanner afr control.hpl

[SiriusC1024]

Of course the wideband follows commanded with the fuel trims off. This is because you calibrated the airflow models with the wideband with fuel trims off.

It's common to see the wideband reading lean at low RPM/MAP when CL is enabled, especially at idle. Nothing about the 'weird o2 drive' log looks weird.

You should be tuning in CL, though. Save the wideband for PE.

[JamesTheriot]

Ok. This is my first time that the wideband has read this far off at low rpm so it had me worried. Of course I haven't tuned nearly as many vehicles as alot of you guys. All of that makes sense. I'll do some more adjustments today. Thanks for the clarification.

[Alvin]

These are meant to switch lean at part throttle/idle.. in normal closed loop operation.


This is done because if you go lean.. there is some oxygen left over that can make it the cat and burn off anything left over (combustion inefficiencies). If its burning perfect Stoich there is no oxygen left over to do this.

Therefore.. If the car is in closed loop and your taking a average of how the wideband reads.. it will average a touch lean. This is completely OK.


My opinion on this setup.. Lean on the fuel trims for tuning other than wide open throttle. This advice changes if you have a huge cam or something.. but in this application...trust the fuel trims.

[kingtal0n]

2003 C5 6 Speed. Stock car, Cat delete, AEM 30-0300 wideband.
I tuned the MAF and VE tables separately with fuel trims off.

The wideband signal is extremely noisy. Top noisy signal=30s zoom of your file. Bottom is a typical canbus aem wideband input.



It throws a giant red flag that there is a problem somewhere in the communication of the wideband , how is the wideband connected?

[kingtal0n]

Introducing wideband accuracy. Witness the accuracy of a wideband effortlessly crossing threshold



Narrowband is a digital switch behavior, a light switch from "off" to "on" or on from off. We know for sure threshold is being crossed when the voltage begins to rapidly climb or descend without pause and completely switches from "HIGH" to "LOW" or back.
Notice how closely the wideband 'switches' from one side of the threshold to the other.
This is how logs should look in canbus wideband applications. This is normal wideband accuracy and curve shape behavior.


A wideband is a narrowband. The two will always agree unless there is a problem.
When turning on the closed loop in your ECU From a hand held commander unit, while watching the wideband gauge (you MUST Have a gauge to verify wideband controller output) you will see the wideband start crossing the 14.7:1 gasoline threshold at some rate.
I've seen it so many times I would steak dinner on it.


When they do not agree, like yours,

there is a problem. This is a problem. This is not normal.

The problem could be position, temperature, water contamination, age, lead, exhaust leaking, etc... It is very easy to have an unknown problem with a sensor, luckily they are cheap and generally easy to access

And that is not the only problem

The top pink problem shows an extremely noisy signal in the log


I attached a proper signal below in the light green vomit color. It is tempting to think that the noisy signal is part of the inaccuracy problem, but it may be separate problems.
Noisy signals are often caused by nearby EMI from some other wires, there are some AEM noise issues out there after all
https://rennlist.com/forums/944-turb...ut-signal.html
https://www.miataturbo.net/diy-turbo...isy-ts-100700/

things to try to solve quickly
re-wire it
replace it and compare
revisit the configuration
replace sensor
relocate sensor, swap sensors
consider the ground and wiring of the ECU or logging device



Et al;
The correct way to diagnose this issue is by looking at the wideband gauge.
When you don't realize how important the gauge is. The wideband gauge is integral to a wideband microcontroller, which is designed to work with both gauge and sensor perfectly already before we come along to ruin it. There is a certain way the gauge behaves when tuning engines over time I got used to it and I can tell immediately by the behavior of a signal or gauge whether the wideband is communicating properly with the sensor and output behavior is traditional.
https://www.theturboforums.com/threa...2#post-2056605


Experienced tuning experts never use a log to tune a vehicle, unless the log can directly be confirmed to match the 'gauge', which is not just a physical gauge but really a representation of microcontroller behavior and accuracy.
Different controllers have different code to take the wideband sensor behaviors and translate into gauge behavior, for example Innovative LC wideband I can't even use it. I will not tolerate the behavior of that wideband.


Analog Offset ----
If the wideband gauge is reading correctly then you know the microcontroller and wideband are good and the problem is downstream, something in the communication from the Wideband microcontroller to the Logging device, and there are many types of logging devices, for example a logging device might be a serial 9-pin DB on an 0-5v ADC which is what I used in 1998 to record oxygen sensor voltages to a Windows 95 laptop. The analog signal between microcontrollers will always have some offset innacuracy which needs to be accounted for by comparing the wideband gauge with the logged values. You must be aware if using EGR or A/C Input to the ECU that the scanner will never log the correct air fuel ratio, ever, there is always some offset + or - and you must mentally keep track and adjust your analog offsets for every run/pass/condition/day to account for this offset.
https://forum.hptuners.com/showthrea...oltages-How-to

In most vehicles, as RPM/LOAD changes so does the analog offset, therefore a different offset should be used for different rpm to maintain gauge accuracy. Notice we always reference the gauge is matched to the log before trying to tune with any log.
-------

//Canbus
In the canbus application the advantage is no more analog offset, in theory. Although there is still some change for a noisy analogy signal going IN to the controller from the wideband sensor (which could be the problem before signal processing even takes place its already noised up). This is why I never mentioned until now anything about analog offset- the wideband in this thread is a canbus application based on how it is configured in the HPtuners scanner which does not seem to require any end user maths to make it work, thus no accuracy corrections are to be made after reviewing the gauge.


The one in this thread, is not delivering the behavior I am used to. There is a clear problem when trying to look at the data.


This is completely un-tuneable. If I Saw this I would not try to tune the vehicle. You cannot tell any actual air fuel ratio because the value is impossibly oscillating. There is too big difference between those numbers, is it 13.5 or 12.3? Nobody can tell.

I analyzed some of the noise at various rpm and the frequency is the same around 4.8Hz so I do not believe it is anything to do with the coils or injectors wiring at least, or the firing order of injectors, or a misfire or anything like that. It seems like something in the hardware or wiring of the controller and communication between ends (sensor -> wideband controllers -> logging device) For example the wideband controller could have a bad capacitor or maybe a voltage or conversion math in the controller settings is incorrect for the communication in the logger and it is counting part of the packet sending as the air fuel ratio because of misplaced or skewed bits, canbus works in bits which often occur in brackets of 8, 9 or 16 but they can also be weirdly bracketed 12 or 17 or whatever and it depends what the code is expecting to receive and maybe an extra bit is causing a rate term to have high magnitude causing rapid switching in the logger.

In any case this is a kind of general tutorial on the kinds of wideband issues and different kinds of widebands and how to think and go about tuning a vehicle with some kind of logger- always verify the accuracy of what you log and make sure the data signal is clean, it should not be jumping up and down like the pink lines we see here.

[SiriusC1024]

Don't let this clown try to bamboozle you with all this talk about signals. He's had 3 months to come up with arguments supporting his fantasies after I dumbassed him, but he still didn't bother to educate himself about CANBUS.

There are no "extra bits" flying around majikally through the ether. You don't need to know anything technical about CANBUS to see the problem with this charade posted above. CANBUS controls airbags. You think engineers would use it if it wasn't fault tolerant?

No they wouldn't. They also wouldn't use CANBUS on the industrial equipment I work on as part of my profession to control things that could kill people if they went out of control.

Now, if you go look up the IEEE literature on the subject you'll see what I mean.

Tune in CL.

[edcmat-l1]

Capt Klingon is a joke.

All you need to know about this guy is right here in his own horrific toon.


uploadmetryboilingtime9-22-23.hpt


Just wow!.jpg

[SiriusC1024]

https://forum.hptuners.com/showthread.php?108310-Timing-Tuning-off-EGT

[SiriusC1024]

Capt Klingon is a joke.

All you need to know about this guy is right here in his own horrific toon.


uploadmetryboilingtime9-22-23.hpt


Just wow!.jpg

Terrible. The fuel injector data gets me, too.

Look at the date on that tune file. Instead of doing it the right way, he's been fucking with his tune for at least 5 years because of the goofy tuning methods that he swears by. Mind you, he does so on a daily basis with some garbage potentiometer on a dial in his dash that's wired to the IAT. His ego can't handle being incorrect.

https://en.wikipedia.org/wiki/Time_Cube

[kingtal0n]

Notice They don't address the issue they just complain like women. Smear campaign and ignore the thread topic.


There is nothing in their posts to address the problem issue at hand, they ignore the issue and try to sling mud together as a married male couple.



user posts an awful impossible to read wideband log and they think its 'normal'.




He said its normal lol. Lets see what a real tuner like Ghuggins says about the wideband signal in the log to see if he thinks it is 'normal'.



They are empowered only by hate speach, there is no moderation so they can off-topic bashing threads all day and then point to it like it means something

No guys all you've really done is out yourselves as a gay couple with a agenda and the users are suffering due to their ignorance transgression of proper engine tuning

[bk2life]

No guys all you've really done is out yourselves as a gay couple

ive spoken with you before.
I listened to what you said about tuning.
I have tried your ways--leaner is meaner.
Cleaner plugs, all that stuff.

I take what everyone says and put into a basket and use it all here and there when the situation calls for it.

But I too have noticed the things you say dont make alot of sense. Now, maybe in F1 or some type of crazy racing where it's high RPM WOT track usage I bet your methods work for that type of stuff.
But, for all the rest of us, CL, NB tuning is the way to go. WB for PE tuning, everything else NB.

Your paragraphs of ongoing relentlessness gets old. State your facts, and move on.
Pages and pages of words doesn't look good to anyone.

This is all i will say on this topic.

[SiriusC1024]

user posts an awful impossible to read wideband log and they think its 'normal'.

Nope. The wideband is wrong, and you're wrong about why it's wrong.

Notice They don't address the issue they just complain like women. Smear campaign and ignore the thread topic.

No guys all you've really done is out yourselves as a gay couple with a agenda and the users are suffering due to their ignorance transgression of proper engine tuning

Sounding more and more schizophrenic every day there, kt. Just like the Timecube guy.

"A large amount of self-invented jargon is used throughout, often never defined. In one paragraph, he claimed that his own wisdom "so antiquates known knowledge" that a psychiatrist examining his behavior diagnosed him with schizophrenia.^[6] Adi Robertson of The Verge commented that Ray's theory of time is "an incredibly confusing one peppered with racism and homophobia"
https://en.wikipedia.org/wiki/Time_Cube

Now, bake our cake.

[edcmat-l1]

No guys all you've really done is out yourselves as a gay couple with a agenda and the users are suffering due to their ignorance transgression of proper engine tuning

Your own personal example of "proper engine tuning" huh? LOL

You're a joke buddy.

uploadmetryboilingtime9-22-23.hpt


Just wow!.jpg