05 GT way too RICH

[dwmccauley01]

Current mods: Catless uppipe, somewhat functional CAI.

 

Hello, I tried my best to do as much research as possible before presenting you all with my issue. I read through Bad Noodle's Subie Newbie guide and I have think I have the basic idea of it all. However, it seems I have gotten myself into problems beyond my knowledge.

 

So to begin, I needed a new air filter. However, with my discount a Spectre CAI was about the same as a new paper filter. I was aware the car would need a tune to compensate for the differing airflow so I did research on MAF scaling for OL and CL. During my research, I came to John M's setup and MAF scaling which I though would be a good base to start from:

 

https://legacygt.com/forums/showthread.php/70mm-intake-12-pics-123246.html

 

For daily driving the AF corrections from learning view looked some what alright, but not the best. The 0-5.59g was around -9% and the rest were all -4-6% . The main concern, however, was the -4.00 timing pull around 5k rpm at 1.6+ g of load. Also, from logs high load caused the AFR to be pegged at 11.02

 

After rescaling the first time, the AF ranges looked alot better except for the 0-5.59 area which was around -6%. The knock correction in high rpms and load was reduced to -2.80 which still isnt in the safe zone.

 

The second rescale yeilded a very acceptable AF learning table which has remained between +/- 1% for the most part. Along with the rescale, I figured I would do a few simple things while I was reflashing the car.

 

The changes include:

CL/OL loop delay to 0

 

Disabling AF correction in OL

 

Timing Compensation Zeroing

 

The Tua tables were going to be fixed because I thought it may help with the richness, but I couldnt find them for my ecu defenition.

 

I also set the MAF scaling to 470 because the max of the MAF scaling is at 440.

 

I pulled about 10% timing alound 2.5 g and 4k-5.5k on my base timing table to try and help with the knock.

 

I still am not sure how to add photos on the post itself so I will attach a photo of the scaling and learning view.

 

After the changes were made I went out for a test drive and all seemed alright except for the random ocassional knock which I assumed is normal bc I wasnt getting much feedback knock correction. However, all went downhill which I attempted a 3rd gear pull. The log will be attached. Im not sure why the throttle was stopped at 86% but I was flooring it. And of course the log stopped before i reached 6000. I noted that during the run I was getting plenty of knock at a very rich AFR, however, the ecu wasnt doing much to fix it with no feedback correction of AFR adjustment. After the throttle was let off though the IAM plunged to 0.5 which was then I knew I had a real problem. Also I noted that I had similar, but less intense knock in that area priop to the changes in tune and intake.

 

I assumed the problem was most likely my OL scaling as I do not have a wideband yet so I havent gotten to the tables yet. Another though was that I was still getting boost error from a leaky uppipe but the boost seems pretty consistent with my tables.

 

This is where I am stuck: running too rich means I dont have enough air coming into the engine however, the mainfold pressure seems to be inline with my tables so shouldnt there be enough air entering the engine? or maybe the CL scaling is so off that the engine thinks there is an abundance of air and decides to pump far too much fuel. Maybe the Intake temperature is too high and needs to be scaled also.

 

Also, this may just be my misunderstanding of knock correction, but shouldnt the ECU have pulled timing and learned the correction with the excessive knock? During the pull the ECU didnt even try to correct anything untill after when it dropped the IAM to 0.5. On top of that my fine learning tables show no learned correction in the load range either.

 

I am just kind of confused on the logic and process the ECU is going through if it went straight to pulling the IAM before correcting or learning any knock.

 

Any input or suggestions for correcting the AFR and knock would be appreciated and I apologize in advance for my subie newbiness.

 

Also excuse my copper wires on the intake, its temporary for now to try and keep it from destroying itself.

 

I put all the files on a google drive folder, but I can post stuff on here if that is prefered.

Edited October 3, 2019 by dwmccauley01

[dwmccauley01]

FILES

[dwmccauley01]

Anyone have any ideas on what is going on? Also the files above is the link to the folder

 

Sent from my VS501 using Tapatalk

[moral hazard]

I would like to offer another guide that may help you, it was called the HolyCrapItsFast_TuningGuide.pdf

 

http://www.netzonexpress.com/Files/HolyCrapItsFast_TuningGuide.pdf

Edited October 4, 2019 by moral hazard

[dwmccauley01]

I would like to offer another guide that may help you, it was called the HolyCrapItsFast_TuningGuide.pdf

 

http://www.netzonexpress.com/Files/HolyCrapItsFast_TuningGuide.pdf

 

I appreciate it, I like how that guide uses excel alot more as a tool to see the math and process more clearly.

 

I think Im stuck for now untill I can get a wideband installed. Right now its hard to tell how accurate the readings are for open loop scaling.

 

Ive been also looking into the Tau tables but I cant find the definition for A2WC510C so if anyone has that it would be appreciated.

 

any other tables I should be looking into altering?

[moral hazard]

 

Ive been also looking into the Tau tables but I cant find the definition for A2WC510C so if anyone has that it would be appreciated.

 

Sorry could you explain where you see the Tau tables? I'm not sure that I've heard of them.

 

Also I know you have the 5eat, but consider test flashing A2WC522N.

 

I have flashed a manual rom successfully on my 2010 and others on the romraider forum have done so for other cars/years also.

 

That rom would have better support and likely more tables.

 

Example :

http://www.romraider.com/forum/viewtopic.php?f=32&t=7892

https://github.com/LegacyNsfw/EcuHacks

Edited October 6, 2019 by moral hazard

[dwmccauley01]

Sorry could you explain where you see the Tau tables? I'm not sure that I've heard of them.

 

 

 

Also I know you have the 5eat, but consider test flashing A2WC522N.

 

 

 

I have flashed a manual rom successfully on my 2010 and others on the romraider forum have done so for other cars/years also.

 

 

 

That rom would have better support and likely more tables.

http://legacygt.com/forums/showpost.php?p=5516111&postcount=10

 

There's a section for them in ecuflash under the transient fueling but all I get is the current map description is invalid or incomplete.

 

Sent from my VS501 using Tapatalk

[dwmccauley01]

I just realized I dont know how to read the open loop table. In higher loads and rpm, the stock open loop calls for under 11 so I really have no idea how rich I'm running with the stock limit of 11.02. I can guess until I get that wideband but I think that has to be the next step. If anyone has one laying around let me know

 

Sent from my VS501 using Tapatalk

[dwmccauley01]

From what I read, some tuners just strap the wideband to the end of the exhaust to dial in the other sensors. In that case, would I be bad off if I just replaced the rear O2 sensor with a wideband even though I still have the cat in the downpipe?

 

 

Sent from my VS501 using Tapatalk

[dwmccauley01]

Fortunatley, a few hours after I posted last a friend of mine basically gave me an AEM AFR gauge with the Bosch LSU4.9 wideband .

 

After additional research, I decided if I do it, I'm going to do it right. I'm going to look into adding a sensor bung. However, its recommended to be 12-18 inches from the turbo so I may have to remove one of the cats.

 

 

 

Sent from my VS501 using Tapatalk

Edited October 9, 2019 by dwmccauley01

[dwmccauley01]

Sorry could you explain where you see the Tau tables? I'm not sure that I've heard of them.

 

Also I know you have the 5eat, but consider test flashing A2WC522N.

 

I have flashed a manual rom successfully on my 2010 and others on the romraider forum have done so for other cars/years also.

 

That rom would have better support and likely more tables.

 

Example :

http://www.romraider.com/forum/viewtopic.php?f=32&t=7892

https://github.com/LegacyNsfw/EcuHacks

For some reason I didn't see the second part of your post.

 

So you're saying that you flashed the manual transmission rom to the 5EAT? Were there any negative effects?

 

I know the transmission has a separate TCU but I'm not too familiar on how the ECU communicates with it other than requested torque and throttle position.

 

 

 

 

 

Sent from my VS501 using Tapatalk

[moral hazard]

I did that, yes. Could not notice any real difference.

 

Here is an example of another person doing it also:

http://www.romraider.com/forum/viewtopic.php?f=15&t=8687

 

It's just a thought, if you wanted better rom support.

It's obviously a risky move and you should have a way to recover from a bricked state before you try something like that.

 

If you succeed, then you can mess around with speed density etc:

https://www.romraider.com/forum/viewtopic.php?f=37&t=7818

Edited October 9, 2019 by moral hazard

[dwmccauley01]

I did that, yes. Could not notice any real difference.

 

Here is an example of another person doing it also:

http://www.romraider.com/forum/viewtopic.php?f=15&t=8687

 

It's just a thought, if you wanted better rom support.

It's obviously a risky move and you should have a way to recover from a bricked state before you try something like that.

 

If you succeed, then you can mess around with speed density etc:

https://www.romraider.com/forum/viewtopic.php?f=37&t=7818

 

Thats really cool, I never even thought about that. It would be nice to have some updated logic.

 

I got a copy of the Rom you mentioned earlier and it shows up with the transient fueling and everything in ecuflash so Ill try to test write it later and see what happens.

[moral hazard]

The other thing worth doing if you are able, use something like HxD hex editor to compare the first 0x02000 bits of the stock rom and the new rom. This will ensure the crossflash is safe.

 

More information can be found here :

https://www.romraider.com/forum/viewtopic.php?f=40&t=10263

[dwmccauley01]

All lines except 0x02000 are the same, could I alter this line or does that mean the roms are not compatible?

 

Sent from my VS501 using Tapatalk

[dwmccauley01]

The other thing worth doing if you are able, use something like HxD hex editor to compare the first 0x02000 bits of the stock rom and the new rom. This will ensure the crossflash is safe.

 

More information can be found here :

https://www.romraider.com/forum/viewtopic.php?f=40&t=10263

 

I may have misunderstood, is the 0X02000 line required to be the same or is it all lines going up it?

 

attached is the hex compare

[dwmccauley01]

I sat down on Excel today to see what kind of comparisons and assumptions I could make to fix the OL MAF scaling.

 

As it turns out, from examination of current logs, it looks like im running lean instead of rich in the top end. This conclusion makes sense because Ive had to increase pretty much every value from the base scaling I used.

 

Assuming the CL scaling is fairly dialed in with values remaining +/- 1%, the OL should also be increased by a relatively similar ratio. The (very) rough average of scaling increase in the CL was about 10%. I multiplied the 2.3v+ region by 1.1 and adjusted as necessary to retain a similar curve to stock.

 

To attempt to check my work, I compared the percent difference from one value to another in both the stock and altered scaling. The altered MAF scaling seems to still hold relatively the same rate of percent change as stock with varying fluctuations most likely from differing airflow patterns with the CAI.

 

An immediate concern is that the scaling is in some places 40% higher than stock. However, it may be acceptable given that the housing of the MAF sensor is about 9.7% larger, the cone filter will be much less restrictive than stock, and my setup has eliminated the curve of the stock intake which furthur decreases restriction. However, 40% still seems rather high.

 

Ill add the spreadsheet to my drive folder I used to calculate and assume the majority of what I said. If anyone sees anything let me know.

 

Drive -> https://drive.google.com/open?id=1uVMxRv46ocVvOHsVg02LtXrqI-XEnmii

Edited October 9, 2019 by dwmccauley01

[dwmccauley01]

Also I test wrote A2WC522N with no errors so I will attmept to write it tomorrow.

[moral hazard]

I may have misunderstood, is the 0X02000 line required to be the same or is it all lines going up it?

 

The lines up until that point, that's my understanding.

[dwmccauley01]

The lines up until that point, that's my understanding.

So got the map to flash but after a couple of minures I got P0852 Neutral switch input high. Is there a table to alter to resolve this issue or is it just a cel I need to disable from the tune? The oil temperature was around 190 so I dont think it's inputting too much power before the torque converter locks up

 

Sent from my VS501 using Tapatalk

Edited October 10, 2019 by dwmccauley01

[moral hazard]

Yeah try to disable that CEL, unaware of any related table.

[dwmccauley01]

Something isn't right, I drove around and my average mpg went from around 20 to 14.5 . Ill look more into it in the morning.

 

Sent from my VS501 using Tapatalk

[dwmccauley01]

Alright so im guessing the increased gas use is because I scaled up the maf too high. I noticed some tables were different in the A2WC522N such as idle and injector latency so i copied them over but the car ran awful. Everytime I stopped it would stumble real bad so I switched it back to the original and it ran fine.

 

Next im going to try and put the old scaling for the MAF back to see if that's what is causing the problem.

 

A side note, for some reason my learning view on romraider quit working so I'm not sure what is going on with that.

 

Sent from my VS501 using Tapatalk

Edited October 13, 2019 by dwmccauley01

[dwmccauley01]

Scaled down the MAF in the upper range. The car was not happy. Going to switch it back and try to scale it up slightly to see if that fixes anything. just guessing at this point.

 

It looks like im underboosting a bit also which would definitley cause some of the problem.

romraiderlog_20191012_214542.csv

Edited October 13, 2019 by dwmccauley01

[dwmccauley01]

I returned back to the 10% scaled up MAF Scaling I created using the Excel spreadsheet andd added an additional 3% to all values.

 

The car is running quite a bit better as far as knock goes. I did a series of 2nd gear pulls and I couldnt get any kind of feedback correction so I think im in a fairly safe area. As far as performance goes, my MPG after the pulls was around 10.5 which is pretty awful. Next couple flashes I'll dedicate to fine tuning the MAF scaling.