Concerning fuel trims

[ExtremeCaliber]

Year/Make/Model: 2007 Mazdaspeed3
Mileage: 230k (engine rebuild 30k~ miles ago, prev owner)
Location: ATL, GA
Concern: Fuel trims out of whack
DTC's: (please include descriptions) None
Modifications:
Forged Pistons (Unknown Details)
Cobb Cold Air Intake
Balance Shaft Delete
Upgraded High Pressure Fuel Pump Internals
Upgraded Timing chain
Spin on Oil Filter Conversion
Oil catch can Mishimoto
HKS Blow off Valve (VTA)
Turbosmart OPR V2
2.5 Swap


ECU/Tuning Software: Versatune Lite, Using torque pro for convenience and graphing (untuned)
Is the concern intermittent? No
Can you duplicate the concern? Yes
Recent Repairs: Worked on a/c compressor clutch a month ago. degreased some burnt grease from previously busted pass motor mount with a pressure washer and degreaser. Found one intake leak from a hose that goes from the intake manifold to the blow off valve. Spark plugs changed about 10k miles ago.

Basically, I'm trying to track down what's the deal with my fuel trims. It wasn't like this before, although I wish I had screenshotted what they were before to give an idea what was the norm as there are a few variables here with the 2.5 swap. I'm aware the car is due for a tune, but been procrastinating rebuilding a spare k04 to swap in; then would go out for tuning. I've pressurized the intake side with my lungs and found the leak as described previously.

I've attached a crude graph from torque pro, I know I'm missing throttle position but all I did was blip the throttle. Let me know what you'd need from me and I'll try my best to upload test results as soon as I can. I'd just like to know where to start since stft DROPS after a throttle blip and bogs the car down until it returns to normal, which is concerning to me. Have stayed off driving the car because of it just in-case.


Just let me know what graphs you'd need since I know this probably isn't enough.

 

[ExtremeCaliber]

After going out for a drive after resetting the ecu (battery off for a few secs) ltft is now at 12% and stft hovering 0 plus or minus 2% deviation. Better than 14 which I was getting before, will keep looking for leaks

 

[ExtremeCaliber]

 

[Crazycanadian]

Fuel trim issues shouldn't be looked at from a snap shot stand point... Fuel trims are used in relationship with codes, and drivability issues.. You also look at them for trends under different driving conditions... You aren't worried about AFR when you are in closed loop, The ECU is using fuel trims to correct the AFR to the commanded values...

A snap throttle doesn't tell you whats going on.. Torque lets you log and down load CSV files.. you'll need a 10 to 15 min drive cycle with some idling, stop and go traffic, variations in load and rpm.. You want smooth throttle movements slow throttle movements...

Versatuner logs would be better then Torque.. ForScan is even better then all of that...

 

[ExtremeCaliber]

Fuel trim issues shouldn't be looked at from a snap shot stand point... Fuel trims are used in relationship with codes, and drivability issues.. You also look at them for trends under different driving conditions... You aren't worried about AFR when you are in closed loop, The ECU is using fuel trims to correct the AFR to the commanded values...

A snap throttle doesn't tell you whats going on.. Torque lets you log and down load CSV files.. you'll need a 10 to 15 min drive cycle with some idling, stop and go traffic, variations in load and rpm.. You want smooth throttle movements slow throttle movements...

Versatuner logs would be better then Torque.. ForScan is even better then all of that...

I was mainly focused on idle values since I assume 14% on long term is pretty high than what it's supposed to be. (I assume in the +/- 7% range). The throttle blip was just a weird one since it plummets to -25% and that's just super weird to me, I wouldn't know what to look for first.

For example, a screengrab on idle from one of my old posts shows ltft is at 7%

 

[Awafrican]

Fuel trim issues shouldn't be looked at from a snap shot stand point... Fuel trims are used in relationship with codes, and drivability issues.. You also look at them for trends under different driving conditions... You aren't worried about AFR when you are in closed loop, The ECU is using fuel trims to correct the AFR to the commanded values...

A snap throttle doesn't tell you whats going on.. Torque lets you log and down load CSV files.. you'll need a 10 to 15 min drive cycle with some idling, stop and go traffic, variations in load and rpm.. You want smooth throttle movements slow throttle movements...
Versatuner logs would be better then Torque.. ForScan is even better then all of that...

you can absolutely snap shot fuel ltft while cruising and idling no need to review a 15min log. what's ltft at idle, AC fans on or off shouldn't change it but if it does probably dirty filter. what's ltft at cruise, then see how it's doing through acceleration (non WOT, gentle and harder).
ideal is ltft +/-5, good is -/+7, +/-10 you've got a leak and should find it +/-15 something's broken.
versta tune vs torque vs forscan vs cobb is all looking at the exact same data, sure perhaps the software and device has different polling rates from the obd port but for this purpose it's all fine.
@ExtremeCaliber have you done a boost leak test?

 

[btstarcher]

Stock intake? Since you're not tuned, it wouldn't matter; trims should be close to 0. Between +/-10. If you have an aftermarket intake, you need a tune.

 

[ExtremeCaliber]

you can absolutely snap shot fuel ltft while cruising and idling no need to review a 15min log. what's ltft at idle, AC fans on or off shouldn't change it but if it does probably dirty filter. what's ltft at cruise, then see how it's doing through acceleration (non WOT, gentle and harder).
ideal is ltft +/-5, good is -/+7, +/-10 you've got a leak and should find it +/-15 something's broken.
versta tune vs torque vs forscan vs cobb is all looking at the exact same data, sure perhaps the software and device has different polling rates from the obd port but for this purpose it's all fine.
@ExtremeCaliber have you done a boost leak test?

Have not done a proper boost leak test since I'm waiting on some fitting to adapt my tire compressor since it's what I have on-hand. I did pressurize it with the power of my lungs to find the first leak which I described above, but anything at higher pressures, not yet and will give the news tomorrow when it arrives. Currently using one of those bladder adapters and fitting it after the MAF so it fits snug in the intake tube.

what's ltft at idle, AC fans on or off

The LTFT at idle was the 13.28% in the torque pro screenshot with short term targeting 0% pretty well. AC was off. I'll be out tonight and will observe cruising.

Stock intake? Since you're not tuned, it wouldn't matter; trims should be close to 0. Between +/-10. If you have an aftermarket intake, you need a tune.

Aftermarket intake so I'm expecting it to be a bit higher. Though previous numbers used to be at 7% for the longterm at idle, and currently is 13.28%

 

[btstarcher]

Might be normal then. Airflow will be different from stock, so your fuel trims will be way off. The only fix is a tune.

 

[ExtremeCaliber]

Might be normal then. Airflow will be different from stock, so your fuel trims will be way off. The only fix is a tune.

Don't think so, like I said values used to be 7%, the photo taken of my center display was from a month ago with the same setup as now.

One thing that's a mystery to me is why stft DROPS to -25%, bogging the car down whenever blipping the throttle. I maybe wrong but I feel that could narrow down what hose could be lose, I'm just spit balling though.

 

[Awafrican]

you can go over all hoses by hand but you need a boost leak check

 

[Crazycanadian]

you can absolutely snap shot fuel ltft while cruising and idling no need to review a 15min log. what's ltft at idle, AC fans on or off shouldn't change it but if it does probably dirty filter. what's ltft at cruise, then see how it's doing through acceleration (non WOT, gentle and harder).
ideal is ltft +/-5, good is -/+7, +/-10 you've got a leak and should find it +/-15 something's broken.
versta tune vs torque vs forscan vs cobb is all looking at the exact same data, sure perhaps the software and device has different polling rates from the obd port but for this purpose it's all fine.
@ExtremeCaliber have you done a boost leak test?

Snap shots don't give you trends.. The driving log gives you trends.. Trends give you a lot more data... Boost leaks, vacuum leaks, dirty Maf sensors can all be pinpointed from the drivers seat with a proper log... A snap shot is a piece of information, but it doesn't tell the story.. It can be miss leading..

+/- 5% is good, but +/- 10% with out a drivability complaint, or code can be difficult to pinpoint.. A snap shot is like looking at freeze frame data.. It may or may not show you the problem, seeing trends validates the data..

Torque uses generic OBD data unless you've paid for the extras.. even then its reeeeeally slow, Versatuner isn't much quicker, but it's usable.. Cobbs data rate gets closer to OEM which makes seeing trends easier.. ForScan data rate is comparable to OEM and much faster then Cobb... In the world of diagnostics you can miss a lot of info with slow data rate and it can send you down a rabbit hole..

 

[ExtremeCaliber]

Snap shots don't give you trends.. The driving log gives you trends.. Trends give you a lot more data... Boost leaks, vacuum leaks, dirty Maf sensors can all be pinpointed from the drivers seat with a proper log... A snap shot is a piece of information, but it doesn't tell the story.. It can be miss leading..

+/- 5% is good, but +/- 10% with out a drivability complaint, or code can be difficult to pinpoint.. A snap shot is like looking at freeze frame data.. It may or may not show you the problem, seeing trends validates the data..

Torque uses generic OBD data unless you've paid for the extras.. even then its reeeeeally slow, Versatuner isn't much quicker, but it's usable.. Cobbs data rate gets closer to OEM which makes seeing trends easier.. ForScan data rate is comparable to OEM and much faster then Cobb... In the world of diagnostics you can miss a lot of info with slow data rate and it can send you down a rabbit hole..

I'll figure out how to log with forscan tomorrow, will first wait for the adapter I need to pressurize the intake and hunt for leaks there and go for a drive after.

I don't know how much more data is needed for just idle stuff though, felt like the snapshot was enough

 

[ExtremeCaliber]

Just a bit of an update, haven't done anything but went to go for a cold start and got a cel being p0300. Startup had a little hiccup but after rpms leveled, sounded fine. Fuel trims at operating temp on idle is now 8.59% long term. When it started, was 11%.


After a short drive, came home with ltft at 11.72%. Waiting on my valve to come in to further find a vacuum leak from the intake. Next update will be after that, then a drive with forscan

 

[ExtremeCaliber]

Alrighty so far found one, but still can hear pressure being relieved elsewhere. Pressurizing up to 6psi, goes down to around 0.
Anyone know what seal I should purchase for this vacuum line to the brake booster?

 

[ExtremeCaliber]

Maybe my testing aparatus making a tiny leak which is causing the sound. Will retest with a flange instead, though in the meantime I'd like to order that brake booster seal then retest after. I'm not really familiar with whatever seal those specific lines use, but I assume someone out there done them before and could help out

 

[ExtremeCaliber]

Had let the car warm to operating temp out of curiosity to check whats the situation. Will go for a quick drive to see if values change. I wonder if it's because I was messing with the brake booster hose bit.

 

[Crazycanadian]

A vacuum leak gets smaller as the manifold pressure raises, this turns into a boost leak when it goes above atmospheric.. You're trims will show positive at idle and light driving, they'll go down and possibly dip into the negative as you climb into boost...

Between the throttle body and turbo out let the charge pipes have a slight bit of pressure in them, No leaks in this area will cause positive fuel trims... A leak in this area will cause negative trims...

A leak in the piping between the throttle body inlet and MAF sensor can cause positive trims.. This will cause positive trims even as you climb into boost, unless the air flow sucks the leak closed in this area..

A dirty MAF sensor causes more negative trims at idle, but go more positive as load increases.. The response rate of the sensor goes down which will cause a higher level of knock when boost is coming on/throttle transition..

These trim patterns are assuming your tune was dialed in before the problem began.. If your tune wasn't then there's other things you'd see.. If you graph out the total trims of a test drive, against your MAF voltage you'll seen a hump form.. Higher then expected waste gate duty cycles.. with high air flow, but low boost levels indicate a boost leak..

Since you found those bubbles.. Idle the car when it's warm and your fuel trims are stable.. Take brake clean( Non Chlorinated! VERY important) or carb cleaner and spray that fitting.. You could use propane, but it doesn't work as well... watch your trims live and see what they do.. If that's the leak you are chasing, your short term trims will drop and quickly go negative... This way you'll know fixing that leak will fix your problem..

Also Mode 6 data shows you individual cylinder misfire data... So with that P0300 code, even though it hasn't flagged what cylinder is misfiring you can still see if it's only 1 or 2 or maybe all 4 cylinders that were misfiring.. Torque should have mode 6 data, but it might not be listed... ForScan has it clearly listed..

 

[ExtremeCaliber]

Update, catastrophic issue. What I was up to today was pressurizing the intake again, had closed up the pcv going into the valve cover but opened it up again by accident under 5 psi (clamp opened by the slightest touch), didn't think much of it. That escapade didn't result in anything substantial as I couldn't find any other leaks

Car has thrown a code saying o2 sensor voltage was high prior to all this but I'm not sure if thats what's giving me issues now.

After I didn't find anything, I gave the car a start and white smoke came out the exhaust, no tint to it and smelled a bit like water. Was alarmed so I shut the car off, waited a few seconds and gave it another go with the same result. And this time left it running a little more until I heard bogging, checked afr and saw it plummeted. After the 3rd try did not much come from the exhaust, and did notice a black watery spot on the floor.

What happens now is the car will start fine. But after a few seconds, afr's plummet, and fuel trims tank to -25. I shut the car off before anything else happens. Let me know what I should do going forward.

 

[ExtremeCaliber]

Let me know what PID's would be essential, here's a first log file since I'm not really sure what Pid's help here. 30 ish seconds on purpose since I'm not sure if I should run the car for longer or not.