Recently I got a persistent "Service Engine Soon" light on with my Evo X. The car has 111k miles (179k km). The engine light went on after I stopped by Crappy Tire for microfiber towels. It was not flashing, just stayed on. Turning off and restarting the car did not make it go away.

So the next morning I visited a trusted mechanic shop, and the owner scanned the codes for me. The "Service Engine Soon" light had already turned off by itself, but the ECU did store a code P0171: system too lean. The shop owner did a quick check on various intake hoses, and found no leaks. So he asked me to schedule an appointment for further diagnosis and repair.

My Evo X has the stock tune on the following mechanical configuration:
　　



^Here, P0171 is the only concerning diagnostic trouble code (DTC). Other DTCs are not concerning. For example, B2102 simply means that the Evo X has a remote starter installed. C1911, C1921, C1931, and C1941 indicate that, as I have the TPMS turned off via ETACS, the rims don't have TPMS sensors installed.


^Failed fuel pump relay pulled out, with a new one ready to enter service.

▪ Fuel pump relay replaced

I did some research on the forums as I was waiting for the appointment date. Learning from other owners' experience, I started with the fuel pump relay. It failed the salt shaker test. So I bought a new relay (p/n: 8627A049) and had it installed. It made the idle noticeably smoother.

With the new relay, I did a few pulls. They neither caused the "Service Engine Soon" light to come back, nor had any noticeable improvement.
　

▪ Fuel pressure regulator hose checked

I checked the fuel pressure regulator (FPR) hose. The aftermarket silicone hose remains flexible and not leaking. The OEM rubber FPR hose did leak several years ago, and I had it replaced with a Turbosmart silicone hose (3mm ID, 18mm length).
　

▪ Fuel trim live data diagnosis: all healthy

After I brought in the Evo X to the trusted shop, the mechanic performed an extensive fuel trim live data diagnostic. It seems that both the fuel pump and fuel injectors are working normally; The ECU is injecting a proper amount of fuel for the target AFR at all RPM ranges; And there is no intake vacuum leak detected. Everything else is also healthy.

We also noticed that the steering wheel was shaking at highway speeds. It turned out that front tires simply needed to be rebalanced. Nothing from the engine was causing the shake.
　

▪ Spark plugs replaced

The spark plugs have also approached the end of their lifespan. These are NGK DILKR8B6's, installed when the odometer was at ~40k miles (~64k km). Now they have been in service for ~72k miles (~115k km). So it's the time for replacement anyway.

One of the spark plugs is dying due to high speed glazing, which is a common occurrence with chuki and kouki (MY 2010-2015) Evo X's with the hot plugs (p/n: DILKR8B6). (Zenki or MY 2008-2009 Evo X's use cold plugs ILKR8E6, which occationaly cause hesitation at cold start.) Both the hot plugs and cold plugs have a nominal heat range at 8.

I bought a set of four NGK Racing iridium's (p/n: R2556G-8) as the replacement. They have made the cold start quite fierce, the engine rev-happy and more eager, also made the engine more torquey at low RPM. What's more, the engine note has changed in a fancy way, sounding more WRC-like. In addition, of course, as new spark plugs, they have made the engine much, much smoother both when cruising and when accelerating.


^Old NGK DILKR8B6 (hot plugs) pulled out after ~72k miles (~115k km) of service. Note the right spark plug shows signs of high speed glazing, indicating that I need a new set of NGK iridium spark plugs.


^NGK Racing Competition R2556G-8, each comes with an iridium center electrode and a platinum ground electrode. Unlike DILKR8B6 which operates with a gap at 0.024" (0.6mm), the R2556G-8 should operate with a gap at 0.028" (0.7mm).


^A new OEM MAF sensor (p/n: MR985187) has just arrived. I wonder if this overpriced OEM sensor is the same as the NTK 74662 MG0050.
　

▪ Fuel cap replaced

The fuel cap seal (10 years old) is a bit loose and saggy, so I have it replaced with a new OEM fuel cap (p/n: 17114015). If the old fuel cap was sometimes leaking unmetered air, it could be a cause for the P0171 code.
　

▪ MAF sensor replaced

Evo X uses the hot wire type MAP sensor. Since we suspect the existing 10-year-old one is acting up, I had it replaced with a new OEM one (p/n: MR985187). I didn't want to prolong the headache by spraying the sensor with CRC and produce unreliable results.

The acceleration has become more sure and stable. My shop has test driven the car afterwards, and no faults were found. I have also been driving her for a couple of weeks. The "Service Engine Soon" light as well as the P0171 code has gone for good and never appeared again.
　

Conclusion

Ultimately, my shop and I were not able to pinpoint what exactly caused the P0171 code. The Evo X now runs noticeably better with all the repairs and part replacements. Fixing an "Service Engine Soon" code is like finding a way through a maze, and in this case I threw many possible things at the wall, and the "Service Engine Soon" light finally doesn't stick anymore.

 

Do you have the "Freeze Frame" data for this code? I get this code from time to time. My X is much more modded and tuned on e85. If the e85 changes slightly it'll set this code off because my LTFT is maxed.

 

No, I don't. But the fuel trim live data diagnosis should be the best way to locate the fault. The shop owner joked that it must be the fuel pump relay. My mechanic seems to think that, for an stock-tune Evo, the P0171 code is transient; But if the "Service Engine Soon" light comes on more often as the car ages more, the fuel trim data will eventually pinpoint the fault. I just think it's the MAF sensor. The hot wire type sensor has a finite lifespan, and at that age, if I had not replaced it, it would fail sooner or later.

Here is an informative booklet from Toyota about MAF sensors. It helped me when I was working on the fix:

 

My MAF sensor just died on Sunday in my 2011. Funny enough I found a good MAF from a 2004 Lancer Sportback that was in good shape..

 

I'm glad you did! My understanding is that the hot wire type MAF sensor has a lifespan pretty much like tire tread. Over a certain number of years, it will wear out and surely fail. Am I correct?

 

That seems to be the consensus. It makes sense based on my limited knowledge too. I notices something with it the failed unit vs the good unit too, I'll have to post a pic for yall to look at.