Case Study : The importance of datalogging

Hello, fellow HPA members!

A couple weeks ago, my friend who own a reputable Subaru shop came accross a car (2010 STi) that was misfiring at low load and low rpm, while trying to maintain about 2000rpm. You know, when you have a loud exhaust and you don't want to piss of your neighborhood at night... I'm sure you get it. It was a vehicle that I previously street tuned via ECU Flash / RomRaider for some minor modifications and the car was running great since. There was no check engine light on, no ECU codes, nothing.

Subarus have a roughness monitor counter that registers even the slightest missfire. While driving the vehicle in the region I previously mentionned, you could feel the engine hesitating, if you opened the throttle a bit more it went away. Other than that, there was no noticeable problem. It was very weird. My friend had his Snap-On diagnostic tool connected and he can have access to some data, so he could see that the misfires were coming from the two cylinder of the same bank. So he checked the usuals, spark plugs, coils, fuel pressure, but everything checked out. I then offered him to make a quick datalog with RomRaider while driving. He thought we wouldn't find anything, but I insisted and he then agreed.

I connected my laptop, fired up RomRaider and started logging. I was using my template which I tune with, so I could see many parameters at once. Started the car, the AF learning were good, so we went for a drive. I was looking at the data while he was reproducing the problem, I could see the roughness monitor going up mostly on cyl 2, but didn't noticed anything out of the ordinary. I asked him to do it again, and then I noticed something strange. The max value for the left intake AVCS (Variable cam timing in Subaru lingo) was 50 deg of advance. The right side showed a max of 30 deg. I know I didn't target 50 deg in the tune, so I reset the min/max value and asked him to replicate the issue one more time. Boom, left side AVCS wasn't tracking the target at all. It was cycling from 0 to 50 and the misfires were happening when the cams were over 40 deg.

Here's a screenshot of the relevant data.

Notice the left intake AVCS cycling from 0 to 50, while the right side is pretty steady. Replacing the AVCS solenoid on the left side fixed the issue!

Key point of this story, use the tools you have to your disposition. A lot of techs replace parts instead of doing a proper troubleshooting. I call this "shotgun troubleshooting", which is replacing a bunch of parts until the issue is fixed!

Anyway, thank you for listening to my TED talk hahahaha!

Frank

Thumbs up!

Great post, thank you!