2014 Mazda 3: P0171 with high LTFT, no vac leak

Hi,

I have a 2014 Mazda 3 with a P0171 code. I have tried a number of diagnostic options and spent $$$ on trying to get it fixed. My current options involve replacing the exhaust manifold, I'd pay the $1200 in parts and labor but the part is also on backorder.

Situation: LTFT is 20.5% at idle, and STFT oscillates between -1% and 14%. At higher RPMs, LTFT goes down to 17.5%, and STFT stays < 10%. At idle, RPMs oscillate between 550 and 700, this corresponds to the STFT oscillation.

Things tried:

1. Replace the upstream O2 sensor (wideband) at Mazda's recommendation
2. Replace the MAF sensor (borrowed from a friend's car to test it out)
3. Get exhaust leaks between the upstream and the downstream O2 sensor (flex pipe area) fixed
4. Get a mechanic to visually inspect for vacuum leaks (none found, no smoke test was done though)

The Mazda techs are not happy with the quality of the welds on the flex pipe. It's probably not perfect, but I've gone to a couple of welders and I'm pretty sure that whatever leaks remain are pretty minor. I'm not sure if they would explain the trims I'm seeing, but Mazda won't proceed until I get the exhaust situation 100% fixed, which will require me to buy a new exhaust manifold, which I can't find in stock + is also expensive.

Are there some simpler/obvious things I could try before spending more time and effort figuring out exhaust parts? I guess a stuck open PCV valve is one thing that can be ruled out quickly? The car is used for short errands, < 5000 miles/year if that's relevant. I have access to a super cheap OBD2 scanner and Torque Pro, but can consider getting something better if that might help. Thank you.

The exhaust wouldn't exactly be top of my list for p0171, unless there's a fairly significant leak close to the o2 itself. In my experience, exhaust leaks will cause post catalyst fuel trim codes or slow response codes, not the kind of fuel trims you are seeing.
Have you considered fuel quality (old, or incorrect fuel like e85), or fuel pressure?

(Thanks for the heads up on the double post too)

Noah wrote:

Have you considered fuel quality (old, or incorrect fuel like e85)

Fuel quality: I'm guessing you mean that I might currently be on poor/incorrect fuel? I don't think that's likely as I fill the standard fuel at Costco.

I'm not sure if a bad top up at some point caused any damage though - I've had this car for 2 years and don't remember filling up anywhere shady.

Fuel pressure: if my trims drop at higher RPMs, I thought those could mostly be ruled out? I can get Forscan for the FUEL_PRES PID and check it though.

They are a little less positive at higher RPM, if I'm reading your post correctly, but LTFT is still 17 while STFT is 10. That's still adding 27% total fuel trim. Not the kind of improvement I would expect if there was say a vacuum leak.

ankush.jain wrote:

Noah wrote:

Fuel pressure: if my trims drop at higher RPMs, I thought those could mostly be ruled out? I can get Forscan for the FUEL_PRES PID and check it though.

A fuel delivery/volume problem will worsen with increased rpm and load. A fuel pressure problem will be pretty consistent across the board. Which your car is.

Take a look to see how fuel pressure is regulated before reading too much into the fuel pressure PIDs. If it's electronically controlled a skewed sensor could be your problem.

An excellent point Matt

Noah wrote:

They are a little less positive at higher RPM, if I'm reading your post correctly, but LTFT is still 17 while STFT is 10. That's still adding 27% total fuel trim. Not the kind of improvement I would expect if there was say a vacuum leak.

Matt T wrote:

A fuel delivery/volume problem will worsen with increased rpm and load. A fuel pressure problem will be pretty consistent across the board. Which your car is.

I have some more datapoints that I think are helpful, but I'm not sure what conclusion to draw from them. Based on my limited understanding, I'm leaning on this being an airside issue given these data points.

I got a new OBD scanner that can read a lot more Mazda PIDs than the one I had before. These are the things that seem to be off.

1. Total fuel trims go from 35% at idle to 0% as I floor the throttle (when MAF reads 100g/s).
2. TP_REL (throttle valve opening angle) is 17% at idle/accelerator released. Manual says that it should be 12%.
3. On visual inspection, the throttle body valve thing is not fully closed with the car turned off. It seems to have a slight angle to it.
4. EVAPCP (Purge solenoid valve control duty value) is 15% at idle. Manual says that it should be 0%.
5. The following PCV valve test seems to fail: a vacuum line going into the fat air hose between the air filter and the throttle body should have some suction at idle, according to the manual. It does not seem to have any measurable suction at idle, it feels like it's developing some suction when accelerator is pressed, but it's not consistent.

I seem to have some evidence for something being wrong with all of - the throttle body, the PCV valve, and the purge solenoid valve. Is there a way to make sense from these metrics, short of removing and checking the valves? Would anything explain the TP_REL deviation? Thanks again!

I came across this post while looking into the solution for the exact same issue.
2014 Mazda Axela (the Japan version of the Mada 3)
LTFT is always at +20 with STFT jumping from +3 - +8 while Idle to 0 at high RPM.

Did a smoke test and no leak, changes the PVC valve, clean MAF, still no improverment.

LTFT is always at +20 and I would sometimes get the P0171 code.

Did you ever manage to figure out the cause?

Took to a mechanic but he is just guessing it might be a O2 sensor issue but sadly where I live, a replacement O2 sensor is at least $300 + labour and I'm scared to just randomly throw parts at the car and hope for the best