Erratic results from AFR Error

[Hondaeater]

bk2 life, did you confirm the MAF failed and that the AEM WB gauge/ controller is the same value as what is showing up in the channel list? The gauge tab in scanner are showing what's in channels but what is the AEM gauge/controller in car showing vs Scanner? You're using the MPVI interface analog input so it's probably the same.
You might want to ADD injector pulse width when you tune WOT in PE. You need at least 1 bank to show the injector duty cycle%.
I'm a little confused when you said you enabled the LTFT and OL? Either set the CL enable table ECT to stock temp for CL, or raise to 280* for OL.
Cheers,
Hondaeater

[bk2life]

bk2 life, did you confirm the MAF failed

yessir, it was failed

AEM WB gauge/ controller is the same value as what is showing up in the channel list?

gauges shows what the scanner shows, verified.

You might want to ADD injector pulse width when you tune WOT in PE. You need at least 1 bank to show the injector duty cycle%.

Ive always had inj data being logged. At 6800rpms im around 55%

My main issue I want to figure out, is, these two pictures.
This is a datalog from a long drive this past weekend. I installed new spark plugs, old were fine, and new plug wires. The misfire I developed was two plug wires hitting the headers.

I moved the cell hits up to 60 on EQ Error and STFT+LTFT
Why does my wideband error show me so lean, and the STFT+LTFT show me either rich or slightly lean?
My WB is a AEM- X series I purchased from HPTUNERs. Now I have people telling me these WBs aren?t any good. Why would hptuner sell them if they were no good?
Lastly, how do I verify if my WB is working correclty? I cant go further in tuning if I cant trust my tool.
If I needed to buy a $2500 WB then i would have, but, that?s besides the point now. How can I verify it works?
Looking at nothing else on the scan log, how do I get what is being recorded to be accurate?
Should I not use the WB for cruise non PE tuning? And only plug in the WB for WOT tuning? But if I don?t trust the WB, I cant do WOT due to not knowing if I?m really lean or rich.

Thanks for any info on this.

[gtstorey]

Compare what the EQ error and the STFT+LTFT at any particular instance and see if that tells you anything. I would graph EQ Error and make ST+LT as a graph together.

[bk2life]

Compare what the EQ error and the STFT+LTFT at any particular instance and see if that tells you anything.

if im understanding what you said, both the pictures are the exact frame, just clicked on the different tab. Meaning, "at any particular instance and see if that tells you anything" it tells em they arent reporting the same thing, not even close. Why not?

I would graph EQ Error and make ST+LT as a graph together.

What do you mean by this? they both are graphed and screen printed above?

[gtstorey]

Unless I’m missing something, try he charts are an average over time (or max/min) which could be the problem. The graphs are a moment in time or individual values. If you compare the individual values it can help tell you why they are so much different. Is your math correct, or is there something throwing the average of one way out and needs filtering.

[Hondaeater]

BK you keep trying to mix apples and oranges. The STFT + LTFT is logging your closed loop computer corrections. Your WB it's very sensitive and picking up the level of oxygen after that correction. If the computer has to correct 10% rich by taking fuel away then you're wb going to see a rise in the amount of oxygen left over. The point is either log open loop and look at your wide band and YOU make the correction or log closed loop and look at your fuel trims and adjust the table based on the ecm correction but don't mix and match or compare or worse make changes based on two different fuel status.
As far as injector duty cycle, that picture that you posted 3 days ago is all I looked at and in your chart vs time there's nothing displayed there under injector duty cycle in your top group 1.
If your gauge is the same as what's showing in your scanner it's working just fine and you can trust it for OPEN LOOP EQ error correction.
There's a green button in the scanner for special functions you open that go to fuel tab you can turn off closed loop that way you don't have to change it in the tune file. If you're driving along and you're closed loop fuel trims show at -10% error just keep driving steady turn off closed loop you should see 10% error between the commanded EQ and your WB. Since you're using lambda that would be 1.0 equivalency andt 0.90 WB because you are actually richer than the commanded. That's the only way you can compare the WB to your closed loop fuel trim.
If you're in open loop are you going to look at your fuel trims to make the table correction? Then don't look at the WB in closed loop to make corrections.

[bk2life]

--Your WB it's very sensitive and picking up the level of oxygen after that correction.

--don't mix and match or compare or worse make changes based on two different fuel status.


--If you're driving along and you're closed loop fuel trims show at -10% error just keep driving steady turn off closed loop you should see 10% error between the commanded EQ and your WB. Since you're using lambda that would be 1.0 equivalency andt 0.90 WB because you are actually richer than the commanded. That's the only way you can compare the WB to your closed loop fuel trim..

Thank you again for this info. I like to learn and understand each aspect before implementing it. I ask many questions in order to fully understand what happens when changing something. I never thought of the WB showing different to what NB is showing doing to corrections being made. That gives me a whole new outlook on the subject. Next outing I will try turning off CL to see the commanded to WB error. I?ll report back then.

[Alvin]

Honestly, and i'm surprised that no one has said this already.

That cam is on edge/likely not going to be OK with the stock narrowbands/fuel trims. They will consistently fight you to be too lean in spots, too rich in spots. They are just not happy with a certain amount of overlap.

I would tune this with the fuel trims off and just with the wideband and see how it does for you.

[kingtal0n]

Honestly, and i'm surprised that no one has said this already.

That cam is on edge/likely not going to be OK with the stock narrowbands/fuel trims. They will consistently fight you to be too lean in spots, too rich in spots. They are just not happy with a certain amount of overlap.

I would tune this with the fuel trims off and just with the wideband and see how it does for you.

Yeah, somebody should have said that already

IMO turn off the narrowbands, use a wideband to populate the a/f datalog

https://forum.hptuners.com/showthrea...l=1#post721215

It is the proper way to tune a performance engine.

Even if they work, the 14.7:1 from narrowbands is good for.... emissions
Ideal-Air-Fuel-Ratio.jpg


We are trying to pull some people out of the dark ages. Old dogs, new tricks. It isn't 1985 anymore.

[bk2life]

Honestly, and.

It is the proper way to tune a performance engine.

very true to both of your statements. I'm having an issue with the WB not reading correctly down low.
I'mg going to try moving it closer to the head collector, but, currenlty it is right after my Y pipe.

I have commaned 12.5 (8.5L) for PE and it logs/shows .84-.86 so its doing its job there. At cruise, 14.68 (1L) comanded it shows/logs 1.0 to 1.12-im sure this fluctuation is due to NBs doing their job, and the WB isnt right next to the collector to get super accurate readings.

Remember, this tune i am working on is 3 years and roughly 13k miles old. Since i have changed the injectors i wanted to datalog it and see where everything stands. This is when i noticed it was super rich, and needed work to the VE/PE and MAF.
The guy who did the 'street tune' had my vehicle for i think 6 hours. There is no way he could have done an accurate RAF/MAF/VE/PE in that time.
Now i know he pulled a NB and installed his WB there. I think i'll try that, and see what the scan log looks like in OLSD.

I do appreciate and thank you all for giving me comments, suggestions, it really is appreciated, and I've almost got my heaed fully wrapped around it all now.

[kingtal0n]

I don't recommend Lambda tuning


there is a huge difference between 0.8 and 0.76 lambda under some conditions.

https://www.theturboforums.com/threa...2#post-2056605

Make sure your exhaust is leak free. Make sure the wideband does not overheat. It should be around 2 feet from a turbine and perhaps 2.5 to 3 feet away from the head if non turbo. They can work for a while up closer, but once they get really hot say over 1300*F they start to read rich

[Hondaeater]

bk2life, "I have commaned 12.5 (8.5L) for PE and it logs/shows .84-.86 so its doing its job there. At cruise, 14.68 (1L) comanded it shows/logs 1.0 to 1.12-im sure this fluctuation is due to NBs doing their job, and the WB isnt right next to the collector to get super accurate readings."
Yes looking at the WB bounce is due to the nature of NB in CL. You can run OL in MAF only or VE only OR in blended mode when you are finished calibrating those tables.
You might want to run stoich of 14.12 which is 10% Ethanol blend and most common in US gas stations. You might be lucky enough to get pure gas at the pump but it's more common to have E10.
Give yourself 15 minutes to watch this video, maybe it'll shed some light on your tuning and the desired afr/lamba Target.
https://youtu.be/JzbLrn-2jyw

[jsf55]

bk2life,
https://youtu.be/JzbLrn-2jyw

Well worth a watch, thank you

[smokeshow]

A wideband is not a magic bullet. Hand-picking wideband data to calibrate fueling with the ECM in open loop doesn't mean a narrowband sensor is obsolete or inaccurate. It means YOU have become the closed loop part of the equation simply because you know more about the hardware than the ECM does, specifically when and what to ignore in the data. The old school algos in the P59 are not designed to filter out misleading data during scavenging. So it becomes your job to filter it out when the camshaft reaches the 230+ range for duration.

"Narrowband old and bad because cam puts fresh air in exhaust" is a non-sequitur. Where you can use the onboard fuel trims to correct the VE and/or MAF, you should. Just use your judgment.

[Hondaeater]

Could use some help here. P59, DBW, mild cam 231/242 (advertised 280/292) 3600 stall, truck intake with 50# 12613412 injectors. Back story for those whom haven?t seen my previous post. This truck is running on a tune that was done 3 years ago and roughly 15k miles. Now that I have more time to get back into tuning, I started datalogging the vehicle and noticed it is very rich, and timing looks like California after an earthquake. I have done datalogging with and without a wideband, using STFTs and without. Currently I have the MAF failed, disabled LTFT, disabled DFCO, disabled COT, set EQ ratio to 1.0, copied hihg octane to low octane, and disabled PE.
I have an apprx 10 mile ?course? that enters a highway, can get up to 75 mph, then exits to regular surface streets back to start.
When data logging, I am showing lean. So I added 10% to the VE table. I have run this route 4 times today. Engine is above 170 degrees, I start datalogging once I start driving, and the wideband is fully heated up as to not have any odd variables. Once I complete the ?course? I take the data copy, and past special, multiply by half %.
For example, if you look at the axi/row at 55kpa/2400 rpms there is virtually no change on 4 of the charts. And on the 5th I cant explain that one.
I have watched many videos, and have done research through the forums, and either I am making a blatant mistake, I?m positive I?m doing this correct, but why am I not getting the numbers to change, or the inconsistencies?
My previous post has a copy of my tune, and I was told its too rich, injector data was wrong, and a few other things. Those suggestions and comments were implemented, and I really could use an extra set of eyes as I have to be missing something.
Sorry for the long post, I wanted to get as much information as to not having people guessing at things that have been done.
Thank you in advance..

Originally it was running RICH. You discussed the 10 mile course you use, When data logging now it is LEAN?
You posted a log, a tune, and 4 or 5 pictures of it lean. You have disabled PE as mentioned above, and you went WOT? on the log "drive from hobby lobby"?
You hit 5200rpm at stoich AFR and WOT. Repeatedly went WOT to above 4500rpm at 14.68 AFR, Not good.

One thing you need to turn off is STFT in OL. ECM is still making adjustments in OL, you need that to be isolated using EQ ratio error only in your graph, Commanded EQ vs WB data. This is probably why you were getting odd results in OL.

PE ... Set the enable kPa to something like 80kPa so it will enter PE above the delay 2400rpm and WOT.
May I recommend tuning the idle and part throttle first so you limit engine damage before winding it up. Get the fuel sorted out below 4000rpm then progress to higher RPM. I looked at the details section under the channels list and it does not show the MAF failed, but as you mentioned it does when scanning Diagnostics and info>> Scan for DTCs right? i know it is set up to fail, but sometimes it still does not. You should have a CEL on too yes? The DTC section in tune is still checked.
Here is a channel list organized for you, but you need to add the WB, and I left a blank spot next to EQ and AFR Commanded.

What is your Fuel pressure?? and are you running a vacuum referenced FPR?

Cheers,
Hondaeater.

[kingtal0n]

Hand-picking wideband data to calibrate fueling with the ECM in open loop doesn't mean a narrowband sensor is obsolete

Actually that is exactly what it means. No stand-alone computers which use widebands will also use narrowbands because- its obsolete!

It means YOU have become the closed loop part of the equation simply because you know more about the hardware than the ECM does, specifically when and what to ignore in the data.

That is just called tuning. If you know how to tune an engine, you will understand what data to ignore and when, while tuning, so you don't make a bad tune based on bad data you should have ignored.

The old school algos in the P59 are not designed to filter out misleading data during scavenging.

In other words, they become obsolete when engines are modified past a certain point which will require actual thought and tuning efforts on behalf of the owner- such as using a wideband and ignoring bad data. Tuning!

So it becomes your job to filter it out when the camshaft reaches the 230+ range for duration.

"Narrowband old and bad because cam puts fresh air in exhaust" is a non-sequitur. Where you can use the onboard fuel trims to correct the VE and/or MAF, you should. Just use your judgment.

I don't mean to provoke, Just observation. I feel that you do not understand how to actually tune an engine or what the word tuning truly implies. It means, 'get the engine running properly regardless of what the ECU thinks is happening'.

I do believe you understand how the OEM ECU works and basics about engines to some extent- but you seem unable to disparage between which features are necessary for tuning and which are superfluous or extravagant add-ons used to patch up holes in the OEM algorithms like bandaids on a hull full of holes. The ECU doesn't need to know what you are doing or what engine is in the vehicle to be tuned. It doesn't even need to know the actual pressure, timing, torque, any of that, if you actually know how to tune an engine and how computers work.


After your claims about the GM patent for TPS based enrichment during cranking turned out to be false I realized you lack the ability to actually read the patents and comprehend the recursive function removing fueling during cranking to infinity, which helped me understand your personality, most importantly that you will pretend to have knowledge that you do not really have while using seemingly intelligent talk-down holier than thou type style while hoping nobody else can decipher them which is depending on the stupidity and laziness of those around you unable to see through the charade. Your post I just pulled apart is a good example of so many words saying nothing other than 'tune the engine properly by ignoring bad data' as if people need to hear that in order to get better at tuning- you've added nothing useful and helped nobody. You've been doing it so long successfully that when I finally started pulling apart the threads you lost your mind and created that wonderful sig, which by the way is like provoking a dangerous animal. I am not inside a cage as you will continue to experience painfully and regret but once and that is continuously.


The correct way to tune an engine is to select the appropriate air fuel ratio for any given situation. This is achieved in modern times used a closed loop wideband which GM ecu does not have. But that does not mean or imply that using the onboard fuel trims are the correct way to tune the engine. Lack of a fork does not make a spoon the proper choice of tool to use in stead. That is a major hurdle in your thinking that every feature of the ECU must be used just because its there. Again, the ECU doesn't need to know anything about what is happening, the engine just needs to be run properly, thats it.

[smokeshow]

I'm not reading that lol.

[bk2life]

What is your Fuel pressure.

showing 62 psi on gauge from key on to 6800 rpms

are you running a vacuum referenced FPR?

corvette style fuel filter/regulator on a ?truck? style intake manifold. I would like to add a sensor to accurately be able to ?see? fuel pressure every time the scanner is hooked up, and not have to watch my analog gauge sitting on my wiper cowl through my windshield.

Today I set my vehicle up as OLSD and went for 3 drives. I was able to pull and or add fuel via the VE table and actually see it make a difference on the scanner. (see picture) -I think my vehicle actually runs a bit smoother like this as well. I might continue to be MAFless/no O2s and keep it that way. Especially since I will be adding a turbo in the very near future.
One thing I did noticed is a have a ?bucking? I can feel at cruise speed. 45 mph steady throttle. Not ?cruise control? but it is something to do with the follower I believe, as I can feel it down until about 10 mph then my idle drops to where its supposed to be.
I also think a portion of my issue is this isn?t a stockish combo. Meaning, it is very hard for me to hit the 400-1200 cells at any KPA due to the loose convertor I have. Even normal acceleration on a city street I go from idle to roughly 2500 rpms when the stop light turns green. I command my convertor clutch to lock up at 40 mph to limit the continued slip.
I?ll continue to log my trips, and will go on a much longer drive to get a good data to finish dialing in my under 4k RPMs VE table.

fuels.jpg

[Hondaeater]

Looks like it is responding predictably to changes. Did you disable the STFT in OL? Once the BASE VE table is correct, you may find CL drives smooth as OL. If you remove the MAF sensor you will need to add the IAT sensor.
ICT Billet has LS3 card style MAF Mass Air Flow Sensor Weld Bung (for LS Custom Intake Tube)
$21.99
IN STOCK
SKU 551545

You can tune the MAF in OL too, for practice if nothing else. Once you get VE DONE 100% satisfied (+/- 1 - 3%) , tune MAF, (highly recommended), now re-load OL VE/SD tune. Create a VE/SD log, look at your Dynamic airflow value in the channel list. As long as your MAF is failed in a DTC scan, normally this would have a 0 in the mass air flow g/s, (this was discussed earlier and your ecm may still populate an airflow value even when failed, I do the plug & unplug procedure to force the failed mode ) anyway, now playback log, pause to compare the Dynamic airflow and use the MAF Frequency Channel to look up what airmass is in the MAF table of your tune file. The Dynamic and MAF air mass should be near perfect match.
You would want to compare during a steady airflow NOT a bunch of TPS moving around.

Full time OL... If temperature doesn't change much where you are, then sure go OL all the time, but if the temps change much you either have to be constantly re-calibrating the fuel tables for environmental changes OR go CL and let the ECM do it's job as GM intended. If this was a track car where you're going to make fueling changes all day as track conditions change, then sure go full time OL.

If you plan on driving it DAILY, I would have to carry a laptop everywhere. There will be no ECM corrections. Where I live it can be 75* in the late morning and drop > 30* by late afternoon, snow by evening/morning. No way I would go full time OL here, too volatile. I could go to work and on the way home need a completely different tune to drive it safely !
As far as the bucking.... in an automatic misfires can create drivability issues. You mentioned, you set the TCC to lock at 40mph. That might be for economy but watch your cylinder airmass. If it has a drastic change when this bucking is happening, you may not want to lock it that early or keep it in a lower gear like 3rd and locked at 40 but 4th and TCC lockup may be too much load for this cam. Use Scanner special functions green button when you feel the bucking, command 3rd and see if it goes away, unlock the torque converter and evaluate. Command an absolute advance and lower it to see if that helps the bucking.
You can do a lot with that scanner.
Apart from that, reference your cylinder air mass and RPM during this bucking event.
Your timing is less than ideal. Lower by 5* increments at your trouble spot/ bucking until it takes that bronco. Throttle follower is probably not the issue, that air adder is for idle, but I have seen some poor calibration of the decay and delay values cause issues too but are usually broader in scope not just that speed / rpm or kPa. The cracker air flow on the other hand uses speed as an axis and can give that cruise control, self driving feeling.
Too much or too little has poor drivability.
Try using the channels I posted and monitor the follower and cracker air channels when bucking.
Can you send me your AC user math for the WB you are using.
Stay tuned,
Hondaeater.

[bk2life]

Looks like it is responding predictably to changes. Did you disable the STFT in OL? ..

yes?I have nothing showing on the scan for the lt/stft?s My o2 sensors are powered on and showing voltage changes. Without physically unplugging them, is there a way to make sure they are disable and not causing any fueling issues. Or will they still scan but not affect the fueling unless unplugged?

If you plan on driving it DAILY?No way I would go full time OL here..

im in az, right now its 75 ish, but soon itll be 110 daily. Ugh. I do drive this vehicle as much as possible, I?ll keep the maf/and go back into closed loop once im done playing with it.

As far as the bucking.... ..

Zero has changed in 15,000 miles other than me now playhing with AF ratios etc. It?s never bucked before, only now/currently. So, im not worried about being in 3rd or not locked up, because it has never been an issue, ever.

Your timing is less than ideal?.. cracker air flow on the other hand uses speed as an axis and can give that cruise control, self driving feeling. ..

why do you say this about the timing? I only ask, because the timing and fuel tables are what I am learning with, and had no way to check the tune I had paid for 3 years ago. Now that I am able to, I was able to see the fueling was off across the board and once I get that dialed in, ill adjust timing. I have very little retard, maybe a degree or 2 when I either pull to much fuel or add too much, but mostly zero retard across the scale.

Can you send me your AC user math for the WB you are using..

I don?t know what AC means? Does the picture have what youre asking?