Killed my motor trying to figure this out

[ayousef]

Seriously! Now that I grabbed your attention, this is on a 2010 Silverado with a forged LS7 427 transplant and a TVS2300 and literally modified every single thing. I was having a transmission problem, and was diagnosing why the truck felt different after just a 1 3/4" with 2" merge collector to a 1 7/8" to 3" merge collector swap. Figured out I was hitting a higher MAF range at WOT, a range that I had not previously touched and thus was untuned for my combo.

Broke a rocker arms, valve spring, bent a pushrod. Replaced the entire valvetrain after a downtime of three weeks and im back to tuning and this is the problem I face:

Im using an NGK AFX wideband, and quite recent beta version of the software. Wideband is placed after the x-pipe because id rather have an "averaged" reading from both cylinder banks, taking into consideration that readings are slightly richer at the collector, like 0.2:1 richer than placing it where I have, so as long as I account for this (which I am) im good.

I disable closed loop operation and long term fuel trims and zero out the trims, I dont disable DFCO because I have deceleration data filtered out to begin with by means of a filter that reads about 2% higher TPS than idle. I cruise around and collect as many data points as possible, the motor tends to get into boost really quick skipping over alot of cells so I drive in a way that allows me to collect as many data points as possible in all cells. Two to three WOT pulls and I work the MAF curve by multiplying HALF the AFR error into a new MAF curve. Few repititions this way and I get a MAF error that is from 0-2% off which is acceptable by my standards.

Now the problem im facing is the following:

- even with the average AFR error as close as possible to zero, there is still a discripancy between commanded AFR and actual AFR.

- AFR varies between a second gear pull and a third gear pull, getting richer the higher the gear

- AFR would jump from 10.9 at night to 11.8 in the morning

- at a given MAF frequency say 5000HZ. actual afr would be lean at cruise and pig rich while coasting down after a wot pull, both while commanded afr is stoich at 14.68

- Truck does not get into DFCO, and when I forced it into getting into DFCO it acted very funny, so Im under the impression that you do not want an automatic transmission car to get into DFCO unless you have the ability to maintain a locked torque converter at coastdown.

1) Im trying to find out WHY am I having those problems? Are there any fueling enrichment tables that HPTuners is not displaying? something like an IAT enrichment table, or ECT enrichment, gear enrichment etc...?

2) why do we filter out deceleration data when tuning using a wideband in open loop? do we filter out deceleration data when tuning idle/part throttle using the narrowbands?

3) Does this sound like a bad maf placement by any chance? Im attaching a picture of how the MAF sensor looks like on the truck.

Im about to smash the laptop trying to figure this out.

*included image of the MAF location, latest hptuners file and log, look at frame 28185 for the post WOT richness, and a config*

[carlrx7]

for starters, id swap the maf and the 6" pipe.

[ayousef]

for starters, id swap the maf and the 6" pipe.

its a 3.x inch pipe, infact there are many who use a 4" pipe and maf with no issues? I think the MAF location is bad too, too close to the airfilter and right at a bend, doesnt seem its getting a good reading since this is a custom cold air intake built by SLP

Anyone else wanna take a shot at this?

[smokeshow]

Should clean that engine bay! I know you're a fan of using histogram filters, but since that function logic implements immediately, it does not take into consideration any physical delay. With the vast majority of people having to tune on the street and not the dyno, we're unable to mimic steady state and we get throttle transition 'noise' in the feedback from the O2 sensors. Filters are very useful, but they can't fix everything So you need to try to mimic steady state. That throttle transition at 28182...no good for getting clean, steady sensor feedback. If you can't hold a specific cell on a loaded dyno when calibrating for fueling, you need to (at the very least) roll into the throttle and roll back out of it. Realistically, the slower the better. Obviously there is a speed limit on the roads so this can be difficult...but just so long as you're aware.

While that is an issue, I don't think it is whats causing the post-WOT richness. It looks to be a symptom of large injectors riding the minpw value and not being able to inject a small enough amount of fuel in the area of <30kpa manifold pressure. You might fix this by lowering your base fuel pressure and lengthening the required minpw and giving the injector more precision with the injection event.

[ayousef]

Should clean that engine bay! I know you're a fan of using histogram filters, but since that function logic implements immediately, it does not take into consideration any physical delay. With the vast majority of people having to tune on the street and not the dyno, we're unable to mimic steady state and we get throttle transition 'noise' in the feedback from the O2 sensors. Filters are very useful, but they can't fix everything So you need to try to mimic steady state. That throttle transition at 28182...no good for getting clean, steady sensor feedback. If you can't hold a specific cell on a loaded dyno when calibrating for fueling, you need to (at the very least) roll into the throttle and roll back out of it. Realistically, the slower the better. Obviously there is a speed limit on the roads so this can be difficult...but just so long as you're aware.

While that is an issue, I don't think it is whats causing the post-WOT richness. It looks to be a symptom of large injectors riding the minpw value and not being able to inject a small enough amount of fuel in the area of <30kpa manifold pressure. You might fix this by lowering your base fuel pressure and lengthening the required minpw and giving the injector more precision with the injection event.

Thanks alot for the reply Jake!

I was trying to get into the throttle as slowly as possible, but something about my blower doesnt like to be at partial boost, its either out of boost or in full boost, I guess my blower bypass valve is acting funny but I try my best since it tends to jump over and skip MAF cells. My engine bay definitely needs some cleaning but if you live here youd get fedup cause two days later this is how it will look like lol.

A friend of mine uses a vaccum pump with some hose and feeds it through the exhaust until it reaches as close as possible to the exhaust ports on the head, and the vacuum pump pulls the pulses real quick to the wideband sensor, but he tunes bikes and says there is no other way to tune 9000rpms+ accurately.

Im glad you thought of the min injector pulse width being an issue cause that was on my mind tonight while I tried to sleep, it does make sense, lots of vaccum, high fuel pressure and a big injector dont help! With that said the injector data I have for the ID850 should be spot on and its not really that big of an injector by today's standards. a Vaccum/boost referenced regulator would come a long way here!

thanks

[SultanHassanMasTuning]

^yousif best of luck with the tune.

since you have a FRP 1:1 wouldnt your injector flow should not be as linear as it is.

in the middle east its simply dusty, most of the cars we work on are very dusty even brand new GTR engine bays.

[MikeOD]

its a 3.x inch pipe, infact there are many who use a 4" pipe and maf with no issues? I think the MAF location is bad too, too close to the airfilter and right at a bend, doesnt seem its getting a good reading since this is a custom cold air intake built by SLP

Anyone else wanna take a shot at this?

He meant the order, and I think he also meant it's a 6" long pipe...ie put the pipe after the air filter, and then put the MAF between the pipe and the coupler to the throttle body. I don't know if that will make a major difference, but it will probably help smooth the airflow across the MAF.

[ayousef]

^yousif best of luck with the tune.

since you have a FRP 1:1 wouldnt your injector flow should not be as linear as it is.

in the middle east its simply dusty, most of the cars we work on are very dusty even brand new GTR engine bays.

Unfortunately I do not have a 1:1 FPR, this is a 2010 Silverado truck that has a fuel pressure control module that has a mind of its own when it comes to fuel pressure!

[darkhorizon]

even with the average AFR error as close as possible to zero, there is still a discripancy between commanded AFR and actual AFR.

I dont see how this is logical... the AFR error IS the difference between commanded and actual.... its a calculated data point from those 2 numbers.