FINALLY solved driveability issue with my Subaru

Status
Not open for further replies.
Joined
Jul 2, 2007
Messages
5,293
After an exhaustive several months of chasing an intermittent idle quality problem that would also extend into slight driveability issue during warm-up. The whole thing mimicked a vacuum leak but never found any. Even went so far as to replace PCV valve in case it was the cause although it rattled around just fine like they're supposed to. Considered it to be possible flaky brake booster but it passed the usual test to reveal a bad booster and brakes never felt excessively hard or spongy.

Well I finally nailed it. It was the reference air hose for the fuel tank pressure sensor. It was jammed up and not sampling atmospheric pressure properly. Never tripped a CEL, although it would seem it would be prone to, due to being in the evap emissions suite of sensors. Once I freed it and corrected it, everything is back to normal and she's running great again. Idle is smooth again, low end torque is back where it should be, and warm-up behavior is back to normal.

As far as I can tell the lack of proper reference pressure to compare against in-tank pressure was causing it to run a pressure check cycle when it wasn't optimal and more often than normal. Only a theory, but there is no denying that since I resolved the hose issue the car driveability issues are gone. It is not a vacuum hose although it is the same size and type, just no vac sucking through it, it is there to provide a reference air pressure source for the fuel tank pressure sensor and ECU to calculate fuel tank pressure against atmospheric pressure of the environment around the car.

Pretty obscure problem, glad I found it.
 
Good to hear you got it!

Really odd though, that it didn't set an EVAP system code.
 
Wow!! Thanks for the update.
smile.gif
 
when you say it was 'jammed up' what are you describing? did it suck in some debris ? is there something here that might be a good thing for other Subies to inspect?
 
Found it because I decided to check over the evaporative system's hoses again to make sure secure, checked this one and found it kinked which is a better word than jammed I guess. It comes off the front of the sensor and has to do almost a 180 into a hole in the unibody/chassis that has no access any other way. I even removed the back seat to expose the panel that holds the fuel pump assemblies to see if I could access the body cavity that the end of that hose is poking into. No joy. So, it has to be in there to be protected from elements and sample atmospheric pressure.

The hose in question was unlabeled in this sketch from the service manual, so I added (C). Observed it to be kinked instead of just an acute angle like in the sketch; relieved said kink by pulling more of it out of the hole in the chassis cavity (not shown on sketch but would be off to the right if it were shown). I said "jammed up" earlier in this thread because basically it looks like it got too much length jammed into the chassis cavity and caused it to kink at the angle shown in the sketch.


300361d1449946330-question-about-fuel-tank-pressure-sensor-untitled.jpg
 
Don't have scanner. There is a magic handshake procedure that gets the codes to display on the odometer readout. I'll have a look.
 
Originally Posted By: LoneRanger
Don't have scanner.


Would it be cool to laugh in your face if there was a pending code this entire time...
wink.gif
 
I'm too tickled pink about having it running right again, to get uptight. So yeah it would be okay, LoL.
 
Status
Not open for further replies.
Back
Top Bottom