Last month we had a 2012 Subaru Forester Touring (2.5-liter four-cylinder engine) come in with the check-engine light on. The customer said the warning light came on about a week earlier.
The only symptoms were the check-engine light and a burning smell.
We test-drove the car first to confirm the customer complaint. The car drove fine, but there was a smell of burning oil when we got it back to the shop.
We connected the car to the Subaru scanner and pulled the trouble codes. There were no pending codes, but there was a code P0171 (system lean bank 1) stored in the history.
I spend a lot of time focusing on the data we collect during diagnoses. One of the most important groups of data we analyze is the freeze-frame data. The freeze-frame data is the exact condition of the engine when the code is set.
We found that the freeze-frame data was stored when the car was fully warm at 60 mph. The short-term fuel trim was at 35.2% and the long-term fuel trim at 17.2%.
The fuel trims are particularly important because they show us how the fuel system is operating. The optimum fuel trim would be 0. If the fuel trims are at 0, it means the computer is not adding or subtracting fuel. Because both of the fuel trims were high, we assumed that the mass air flow sensor was giving an incorrect value or there may be a large vacuum leak.
Diagnosing the problem
We then test-drove the car again, but this time with the scanner connected to the car. We graphed the mass air flow sensor and monitored the fuel trims at the same time. We found that the mass air flow sensor was calculating correctly and the downstream oxygen was indicating adequate fuel trims.
Because the engine data was coming back correctly, we thought there may be a mechanical problem. We then performed a smoke test on the intake system, looking for vacuum leaks. There were no leaks.
While we performed the diagnostic, we also checked for the cause of the burning smell. We found that there was quite a large oil leak coming from the front right side of the engine. It was leaking from the right-side lower corner of the front timing cover and possibly the right-side valve cover.
After further investigation, we discovered that the oil was dripping straight into the front lean air fuel ratio sensor. This sensor is responsible for a large amount of the fuel control, directly out of the engine.
We then started looking at the data coming out of the front lean air fuel ratio sensor and it did not match with any of the actual engine data. The front air fuel ratio sensor was soaked with engine oil so badly that it was corrupted. This led to the sensor giving the engine computer incorrect data.
We do see timing cover oil leaks from the new Subaru chain motors, but we had never before seen one leak onto the front air fuel ratio sensor this severely.
I called the customer and told her we had to repair the oil leak in the front timing cover and replace the front lean air fuel ratio sensor and possibly the right-side valve cover as well.
After we repaired the leaks and replaced the sensor, we cleared codes and test-drove the car. All the monitors ran well and no codes returned.
This was an interesting challenge, because the powertrain never actually had a problem – except for the oil leak. This type of problem reminds me why it is so important to always be focused on finding clues in the data.