Cobb BIG SF intake fuel trim issues

[SeeeeeYa]

Originally Posted by anjuna

A MAF cal is a MAF cal is a MAF cal is a MAF cal.

It should be accurate no matter the weather or if the tune is just being flashed.

When I had my 15, my trims were ALWAYS +/- 1-2% even if it had been a month down the road.

In my OPINION, trims are excuses.

Couldn't agree more... but his post included an intake with, shall we say, a less-than-perfect MAF cal.

My quoted response was in reference to that intake and that MAF cal. And in the case of an imperfect MAF cal, and an intake with a reputation of erratic behavior, trims change significantly between just flashed and after a learning period.

The MAF cal won't change with a tune reflash. But he has subsequently explained he's an experienced logger and has accounted for the learning.

Changing tunes to one that doesn't show significant issues will somewhat mitigate the intake issue, which he has done.

I suggest he polish his modifications with a known good intake and continue tuning to potentials, instead of working around problems.

* Registered users of the site do not see these ads.

[SeeeeeYa]

FWIW, the trims on my stock WRX are very close, except during weather swings and depending on constant or stop/go driving. But they are never out of the single digits. Understandable to me.

But once the car has been run a while and all systems and temps have stabilized the LTFT is 0ish and the STFT moves around in the single digits depending on what I'm doing, often close to 0. This stock tune and intake is very good.

[sisomphonE]

I'd love to provide my feedback when driving home tonight. Can you tell me what numbers/values I should be looking at? I have a 16 running their BIG SF 91 OTS map.


Edit: New to turbo cars/Subaru's please don't flame me too hard

[sisomphonE]

Here's a quick log I did after going to the gas station:

http://www.datazap.me/u/sisomphone/l...og=0&data=5-18

[simpleJ]

Originally Posted by sisomphonE

I'd love to provide my feedback when driving home tonight. Can you tell me what numbers/values I should be looking at? I have a 16 running their BIG SF 91 OTS map.


Edit: New to turbo cars/Subaru's please don't flame me too hard

Okay, to do a log you need to basically go pedal flat to the floor from 2/2.5k rpm and hold it to redline or near redline.

You will also need to adjust the parameters you logged so that we can assess the target fueling and boost levels.

I would turn on your comm fuel final monitor and your TD boost error fields in the AP and give us a log at full throttle from 2kish to around redline. 3rd gear or 4th.

Without having this type of log, it's really hard to say

[sisomphonE]

Originally Posted by simpleJ

Okay, to do a log you need to basically go pedal flat to the floor from 2/2.5k rpm and hold it to redline or near redline.

You will also need to adjust the parameters you logged so that we can assess the target fueling and boost levels.

I would turn on your comm fuel final monitor and your TD boost error fields in the AP and give us a log at full throttle from 2kish to around redline. 3rd gear or 4th.

Without having this type of log, it's really hard to say

Got it and thank you. I will upload another log tonight.

[sc00ter]

Originally Posted by sisomphonE

Here's a quick log I did after going to the gas station:

http://www.datazap.me/u/sisomphone/l...og=0&data=5-18

Your log isn't too terrible, but it's not great. You have no feedback knock but you have some fine knock learn once you get above 5000RPM. Coincidentally, your AFR's above 5000RPM remain in the low to mid 11's. Ideally, you want AFR in the mid to low 10's once you get in that higher RPM range to richen things up and make it a bit safer.

For now, I'd recommend not going WOT above 5000RPM until you get a proper tune or get rid of that intake.

[sisomphonE]

New log
http://www.datazap.me/u/sisomphone/l...=3-5-7-8-20-21

[tomferd]

So I'm pretty new at the whole AFR learning etc. I understand you want to keep the AFR rich under load. Can someone explain the closed loop logic or point me somewhere that does? I'd like to learn about the issue.

[Hunter_B]

First off I have to thank you guys for turning me on to this datazap datalog site. Its amazing.

Here is my most recent log after adjusting the intake. I don't think it looks too bad but I am also very very new to this.

http://www.datazap.me/u/hunterb/log-...g=0&data=16-18

[anjuna]

Originally Posted by tomferd

So I'm pretty new at the whole AFR learning etc. I understand you want to keep the AFR rich under load. Can someone explain the closed loop logic or point me somewhere that does? I'd like to learn about the issue.

In a tune, there are tables with X and Y axes.

In one of these many tables on the 2015, there is a target AF/R (air to fuel ratio) that exist for each X and Y value. This table is what the ECU used to "lookup" the amount of fuel to be injected for a given load (X axis) and RPM (Y axis). The values are not arbitrary like some of the Honda stuff, but instead actual ratios, like 14.47:1

Now how does the computer know how much fuel to inject to obtain this ratio? We have a given volume of fuel from the ratio, so we need another unit of volume to determine the air. We get this volume of air via the MAF sensor (mass air flow) which uses a heated coil, air temperature sensor, and/or combination of both that results in a voltage change across the sensor. (More air requires more voltage).

This relationship is not linear, but instead polynomial equation. We find this relationship defined on another table in the tune. Volts vs g/s. This is the MAF calibration.

Yay. Basic understanding.

Now, if our MAF cal is wrong, or we have general fueling faults (bad compensations, ethanol content, etc) we will end up needing a value correction to obtain our closed loop target. This is "A/F correction 1" for our FA owners. Now there is some logic used, because it would make sense if we are correcting by -10% everywhere, we should just learn that value. This gets placed into one of the 5 "A/F leaning 1" options (a,b...,e) which correspond to load values with A being low load and E being high load.

A properly set up tune will have a lot of things ironed out like compensations (barometric, intake air temp, coolant temp, gear, etc.). There aren't really injectors for the FA series yet, but if there were, that's a whole new story. The key piece for proper fueling, especially for those that have modified the MAF position (aka intake) is the MAF cal, since this is a pre turbo measurement of the volume of air flowing into the engine.

When you see logs with huge negative or positive learning values, the tune is rubbish or simply a pump and dump. You'll see a lot of "e-tunes" produced by high volume pump and dumps claiming your tune needs time to "learn" it's boost and fueling. This is because they are sending you a basemap of one to few runs they may have made in their shop on a dyno. There is no tuning, or else you'd have a proper tune that runs proper instead of a junk tune that requires learning.

To make matters worse, all this learning for fueling only applies in closed loop. You're stuck with whatever open loop fueling (aka, no correction - strictly MAF cal driven) for the tune.

/rant.

However, Cobb OTS maps are designed this way because their intent is different. They are designed to run rich (safer) and learn their to being proper. Much more of a one size fits all, and I don't blame them since the maps are free and designed to be safe.

Now all this about the MAF cal on their new intake is worrisome. Could it simply be the OTS map that needs tweaking? Probably. Could the intake have issues? Less likely. All I know is I would not trust the provided MAF cal given the logs I've seen.

[sisomphonE]

Originally Posted by anjuna

In a tune, there are tables with X and Y axes.

In one of these many tables on the 2015, there is a target AF/R (air to fuel ratio) that exist for each X and Y value. This table is what the ECU used to "lookup" the amount of fuel to be injected for a given load (X axis) and RPM (Y axis). The values are not arbitrary like some of the Honda stuff, but instead actual ratios, like 14.47:1

Now how does the computer know how much fuel to inject to obtain this ratio? We have a given volume of fuel from the ratio, so we need another unit of volume to determine the air. We get this volume of air via the MAF sensor (mass air flow) which uses a heated coil, air temperature sensor, and/or combination of both that results in a voltage change across the sensor. (More air requires more voltage).

This relationship is not linear, but instead polynomial equation. We find this relationship defined on another table in the tune. Volts vs g/s. This is the MAF calibration.

Yay. Basic understanding.

Now, if our MAF cal is wrong, or we have general fueling faults (bad compensations, ethanol content, etc) we will end up needing a value correction to obtain our closed loop target. This is "A/F correction 1" for our FA owners. Now there is some logic used, because it would make sense if we are correcting by -10% everywhere, we should just learn that value. This gets placed into one of the 5 "A/F leaning 1" options (a,b...,e) which correspond to load values with A being low load and E being high load.

A properly set up tune will have a lot of things ironed out like compensations (barometric, intake air temp, coolant temp, gear, etc.). There aren't really injectors for the FA series yet, but if there were, that's a whole new story. The key piece for proper fueling, especially for those that have modified the MAF position (aka intake) is the MAF cal, since this is a pre turbo measurement of the volume of air flowing into the engine.

When you see logs with huge negative or positive learning values, the tune is rubbish or simply a pump and dump. You'll see a lot of "e-tunes" produced by high volume pump and dumps claiming your tune needs time to "learn" it's boost and fueling. This is because they are sending you a basemap of one to few runs they may have made in their shop on a dyno. There is no tuning, or else you'd have a proper tune that runs proper instead of a junk tune that requires learning.

To make matters worse, all this learning for fueling only applies in closed loop. You're stuck with whatever open loop fueling (aka, no correction - strictly MAF cal driven) for the tune.

/rant.

However, Cobb OTS maps are designed this way because their intent is different. They are designed to run rich (safer) and learn their to being proper. Much more of a one size fits all, and I don't blame them since the maps are free and designed to be safe.

Now all this about the MAF cal on their new intake is worrisome. Could it simply be the OTS map that needs tweaking? Probably. Could the intake have issues? Less likely. All I know is I would not trust the provided MAF cal given the logs I've seen.

So much to learn! Did you get a chance to analyze my latest log? I had a friend from a shop take a quick look and they said my log was okay.

[SeeeeeYa]

Originally Posted by anjuna

In a tune, there are tables with X and Y axes.

In one of these many tables on the 2015, there is a target AF/R (air to fuel ratio) that exist for each X and Y value. This table is what the ECU used to "lookup" the amount of fuel to be injected for a given load (X axis) and RPM (Y axis). The values are not arbitrary like some of the Honda stuff, but instead actual ratios, like 14.47:1

Now how does the computer know how much fuel to inject to obtain this ratio? We have a given volume of fuel from the ratio, so we need another unit of volume to determine the air. We get this volume of air via the MAF sensor (mass air flow) which uses a heated coil, air temperature sensor, and/or combination of both that results in a voltage change across the sensor. (More air requires more voltage).

This relationship is not linear, but instead polynomial equation. We find this relationship defined on another table in the tune. Volts vs g/s. This is the MAF calibration.

Yay. Basic understanding.

Now, if our MAF cal is wrong, or we have general fueling faults (bad compensations, ethanol content, etc) we will end up needing a value correction to obtain our closed loop target. This is "A/F correction 1" for our FA owners. Now there is some logic used, because it would make sense if we are correcting by -10% everywhere, we should just learn that value. This gets placed into one of the 5 "A/F leaning 1" options (a,b...,e) which correspond to load values with A being low load and E being high load.

A properly set up tune will have a lot of things ironed out like compensations (barometric, intake air temp, coolant temp, gear, etc.). There aren't really injectors for the FA series yet, but if there were, that's a whole new story. The key piece for proper fueling, especially for those that have modified the MAF position (aka intake) is the MAF cal, since this is a pre turbo measurement of the volume of air flowing into the engine.

When you see logs with huge negative or positive learning values, the tune is rubbish or simply a pump and dump. You'll see a lot of "e-tunes" produced by high volume pump and dumps claiming your tune needs time to "learn" it's boost and fueling. This is because they are sending you a basemap of one to few runs they may have made in their shop on a dyno. There is no tuning, or else you'd have a proper tune that runs proper instead of a junk tune that requires learning.

To make matters worse, all this learning for fueling only applies in closed loop. You're stuck with whatever open loop fueling (aka, no correction - strictly MAF cal driven) for the tune.

/rant.

However, Cobb OTS maps are designed this way because their intent is different. They are designed to run rich (safer) and learn their to being proper. Much more of a one size fits all, and I don't blame them since the maps are free and designed to be safe.

Now all this about the MAF cal on their new intake is worrisome. Could it simply be the OTS map that needs tweaking? Probably. Could the intake have issues? Less likely. All I know is I would not trust the provided MAF cal given the logs I've seen.

Listen up peeps.

As a self-tuner I have never seen a MAF calibration that was without flaw from Cobb... not that it didn't "work," per se, but the OTS tune is only a starting point, something that isn't terrible but certainly not a perfect set of calibrations to take to the bank... especially when pushing the envelope beyond the limited performance of the tune as provided. Again, it is a starting point. Knowing what I know I'd NEVER push my car on an OTS map.

For those who then want more via "pro tuning," those tuners simply do not have the time nor inclination to perfect the base tune MAF cal, and most often rely on it for their tunes. A true pro tune where the tuner ensures that your MAF cal is optimized for their results will be rare. Push the envelope too much without refining ALL the critical elements of a tune... like the MAF cal...and toss your dice.

Unlike earlier platforms, where there were many who were self-tuners, it appears the FA20DIT is woefully bereft of those people... partly because the software is not as developed and partly because the ownership crowd is apparently all too willing to hand the responsibility over to others.

Anjuna is an anomaly here and now, one who has taken the reins of his own horse. Listen. Learn.

[anjuna]

Originally Posted by SeeeeeYa

Anjuna is an anomaly here and now, one who has taken the reins of his own horse. Listen. Learn.

Too bad I sold my 15 and don't really do this professionally... I'd kill to be in a shop working as a calibrator. I doubt it pays more than where I'm at now, though the oil business is tanking and I'm one of the last few holding on.

[sc00ter]

Originally Posted by anjuna

Too bad I sold my 15 and don't really do this professionally... I'd kill to be in a shop working as a calibrator. I doubt it pays more than where I'm at now, though the oil business is tanking and I'm one of the last few holding on.

What did you replace your '15 with?

[anjuna]

Originally Posted by sc00ter

What did you replace your '15 with?

Full JDM STi swap 02 sedan purpose built.

I'm of the opinion that most stuff made after the huge Internet boom is pretty quality. Ej207 v7 certainly doesn't disappoint. I'm still stage one and playing with the tune is quite satisfying. Much faster than the 15 on 93. Just now playing around with e85.

Have some goodies coming this next week to help open up the exhaust.

[simpleJ]

Updating this:

A member here kindly scaled my intake for me. However, with the recall logic I have essentially lost that map. So I will be scaling this myself this go around. The new map/reinstall of the intake yields a much better AFR under boost- but it's still about .5 AFR high at points in the band - so it's time to scale it.

I don't have the time to get protuned right now so I'll be correcting this in the mean time and I am hoping to learn a bit about tuning along the way. My tuner reccomended using the AF learn and Correction variables to dial it in to safe levels quickly with the least amount of effort. As such I will be using the YIKES spread sheet to clean up the MAF calibration. I will post my findings here along the way.

[barbo2]

After reading all this I dont know if i should install my Cobb intake thats been sitting in the box for the past 2 weeks :/

[anjuna]

Originally Posted by simpleJ

Updating this:

A member here kindly scaled my intake for me. However, with the recall logic I have essentially lost that map. So I will be scaling this myself this go around. The new map/reinstall of the intake yields a much better AFR under boost- but it's still about .5 AFR high at points in the band - so it's time to scale it.

I don't have the time to get protuned right now so I'll be correcting this in the mean time and I am hoping to learn a bit about tuning along the way. My tuner reccomended using the AF learn and Correction variables to dial it in to safe levels quickly with the least amount of effort. As such I will be using the YIKES spread sheet to clean up the MAF calibration. I will post my findings here along the way.

Let me know when you get close
Then you go open loop
Then you make your "cal"
Run through linear regression
BAM

[simpleJ]

Originally Posted by barbo2

After reading all this I dont know if i should install my Cobb intake thats been sitting in the box for the past 2 weeks :/

You don't know until you try I suppose.

My car runs a little lean on the stock intake as well.


I've got 0-2volts dialed in pretty well, -2.34,0 for a spread until I get above that threshold. Just keep refining it. Take a log on the way back from work, make a map, flash it before I leave in the am, rinse repeat

[MyCupRunithOver]

I ended up having the same issues after about a week with the Cobb Big SF 93 OTS map. Sent a log in. They said that it was normal and that I probably had bad fuel. I did fill up right before the issues. Who knows.

I flashed to the 91 map and it's much better at the moment. I really think there might be issues with the 93 map. Hopefully, with they will see these posts and redo that map. Their OTS solutions are probably the only reason anyone buys this intake.

[simpleJ]

I did the same thing, it reduced the knock but didn't fix the AFR.

Keep it in open loop and you'll be okay.

Is it knocking at all? Have any logs you can post here?

I talked with my tuner/service/builder about it: he basically said that the FA can take a mid 11 AFR but more is pushing it-but if you get bad gas you'll be pushing your luck. (Him not me)

[shadrag]

This thread got GOOD! I've been waiting for some more consistent warmer weather to re-tackle my MAF cal.

[barbo2]

Finally installed my Cobb intake this weekend along with the Cobb nonresinated downpipe. Running Cobbs V301 stage 2+ OTS map and currently have no issues. Haven't put my foot into yet, but have gotten zero knock and DAM is at 1.00. Air fuel rations look good, just runs a bit rich but thats to be expect on an OTS map.

Being new to this, what MAF Cor volt readings should I be looking for?

[anjuna]

Originally Posted by barbo2

Finally installed my Cobb intake this weekend along with the Cobb nonresinated downpipe. Running Cobbs V301 stage 2+ OTS map and currently have no issues. Haven't put my foot into yet, but have gotten zero knock and DAM is at 1.00. Air fuel rations look good, just runs a bit rich but thats to be expect on an OTS map.

Being new to this, what MAF Cor volt readings should I be looking for?

AFR will always look good out of boost. It's in closed loop, so it's hitting the target.

You won't see MAF V correction - that's not a thing.

You'll want to monitor your A/F Correction and Learning related parameters. These will show how much the ECU is changing the amount of fuel injected from what the tune is telling it, to what the actual amount of fuel is required.

The tune does it's calculations based off the MAF. The MAF says "hey, there's this much air coming in" and the ECU says "okay, let me look up how much fuel I should inject. Here we go!" and then the oxygen sensor comes back and says "hey bro, you're like, 9% lean" so the ECU says "well then I should add 9% fuel" and that's where your A/F correction comes from.

A/F learning is like a trust issue between the MAF and the ECU. It's getting tired of hearing that it's wrong from the oxygen sensor, so it learns to just ALWAYS add X percent (which is an average value over time). So then, it's just taking the MAF value, adding or subtracting the learned value, and then making closed loop corrections after that from the oxygen sensor.