BlackWagon GT

The pressure switch on the back of the power steering pump on my '05 OBXT is leaking. Is it available separately from the pump, even unofficially? In the past I've been told "thAt ParT DoeSn't ExIst" for a specific year/model of vehicle but expanding the search to similar others yields what I need. I'm hoping this is similar, but internet searching has only found other people asking the same question. Aftermarket pumps don't seem to come with one, and even the best price I've seen for an OEM pump is more than I want to spend on this endeavor right now.

On a similar note, will the 34433-AG020 pressure switch harness fit my '05? It's officially for '07-up, but that doesn't mean it won't work. The wire insulation on the one in the car is deteriorated and flaking off due to being soaked in fluid, so it's not worth replacing unless I do the pressure switch at the same time, or replace the pump with an OEM one that includes the harness.

The pump itself works fine (for now?), so if the switch doesn't exist separately I might just tolerate the leak for now and keep an eye on the fluid level.

83 like near Randhurst? Was over there yesterday, could've been me.

North of Randhurst. If anything, it was between Hicks and Half Day.

If you have an Accesport, I would check the condition of your voltage with the car running. When my alternator was going bad, my car ran horribly and I was seeing multiple misfires on all 4 cylinders. On a good alternator at idle, you should see at least 14 volts.

Thanks for the input. I thought of that too, and at least at the battery it's around 14-14.5 volts. I don't remember if I checked it at the ECU with AP or not; I'll definitely double back on that.

I also had the thought that a bad rectifier diode might be causing a sine-wave output from the alternator, but my (cheap and basic) oscilloscope showed clean voltage.

Swapped #2 and #4 injectors on Sunday and changed the AP datalog list to include the cylinder roughness values. The first log I ran with the RPM steady around 2100 showed intermittent roughness on all four cylinders, not just #4 like I thought I saw before.

What are the odds that I have multiple bad injectors?

I finally have some direction with my car.

Yesterday I was able to witness and log Cylinder 4 Roughness incrementing between 2100-2300RPM with no load, at one point reaching as high as 9. I swapped #2 and #4 coils to see if the roughness would follow, but it did not. Next step is going to be swapping injectors in the same fashion. If that also is inconclusive, I'm not sure if I should focus on mechanical (compression & leakdown) or electrical (ECU, harness, etc).

rlefevre, you have an '09, correct?

I have attached a few logs I made today. I found out that my AP will only log up to 16 points, so things I wanted to include like cylinder roughness didn't make the cut. I can always change the point list and re-log if it will help to figure out what has been so elusive so far.

datalog46.csv

datalog48.csv

datalog50.csv

Obviously, manual transmissioned cars have a greater sense of engine performance because they have a direct connection from the engine to the wheels. And just as obviously, the automatic transmissioned cars have a slushy connection.

You raise one point and give credence to mine at the same time.

If you are correct, then the problem, or I should say a variation of the symptom, is nonspecific to either type of transmission but is simply buffered by the torque converter so it appears that manual transmission cars are more symptomatic.

However, if there is a symptom variation that is specific to MT cars, then that would give more weight to the possibility that it could be caused by a faulty DMFW since it is the only factory-installed directly-coupled component that is not present with an AT.

Cruising these threads has led me to believe there are multiple causes and therefore multiple solutions to similar symptoms. The problem is, without directly experiencing the symptoms side-by-side, we are all attempting to interpret other people's descriptions to make armchair diagnoses. I know I certainly fall into that category.

Since the data that has come out of my ECU has yielded no clues, I'm starting to wonder if I am looking in the wrong place to find the problem. I did some thinking tonight, but what if it's actually a mechanical issue rather than tuning/software? The dual-mass flywheel specifically came to mind. It's a shot in the dark, yes, but taking its construction into consideration, I would posit that a worn or damaged DMFW coupled with some driveline slack might just be enough to cause oscillation.

I could be way off, but when the obvious answer is wrong, it's time to start thinking outside the box.

Are you pro-tuned? Is it AP or ECUFlash?

AP, sort of. I don't know what map my AP put in my ECU when I divorced it since I think I had a custom map when I married it, so I want to get a verified stock map/rom back into it before I go any further.

I did the fuel line mod, and it made zero difference in the running characteristics. Next step is to flash my ECU.

I'm frustrated and running out of ideas.

I seem to have solved the problem for the most part on my legacy. I was having the problem where it would stutter at certain RPMs when cruising or going up small hills. I found this thread http://legacygt.com/forums/showthread.php/stutter-studder-hesitation-etc-198070.html?t=198070

 

Basically the theory is that a resonance builds up at certain RPMs in the fuel injector line which causes one of the injectors to get starved of fuel and then that cylinder runs extremely lean. I followed the modification suggested in the thread above and replaced one fuel line with a longer line.

 

It seems to have worked well for me. I can still get the car to stutter, but it is much less common now. I have driven about 100 miles since replacing the tubing, and have only had it stutter maybe 3 times. Before the replacement, my car would stutter more than 5 times every time I drove.

 

As a side note, there are some compensation tables that subaru tried to use to "fix" this issue which you are supposed to zero out when you do this mod. I have not got a chance to do this yet, so hopefully my stutter will further improve when I do that.

 

It may not be the fix for everyone, but it cost me about $20 in fuel line and clamps, so it is worth a shot.

At this point, I'll try anything, but I have my doubts that this will fix my problem. According to the links in that thread, the resonance peaks at about 3,000 RPM. Some back of envelope maths on that give a natural frequency of about 100Hz for an STi's fuel system. In my case, the engine runs the worst at about 2100 RPM, which would be somewhere around 70Hz. I suspect that the LGT's fuel line is longer than an STi's but I don't think it's enough to drop the resonance by some 30%. In fact, I went back and looked at some of my old logs, and my correction 1 is consistently 5-10% negative.

I really hope this works for some people, but with how defiant my car has been so far, I admit I'm cynical.

Two in two days!

Tonight, WB I-90. Black or dark blue wagon with exhaust and firm suspension. You got away from me before I had a chance to get a good look, and since I was in my company car (silver Fusion) I wasn't about to give chase through traffic.

I'd be up for a meetup, too. Even more so if it's for sushi.

but did you check a learning view? what k&n are you using and do you confirm it's not causing fuel trim issues?

Me? I have a stock air filter.

Cobb OTS maps do knock at low loads, if you have some learned knock there, the stutter you feel is the car moving in/out of load cells that have a jump or dip in timing from one to the next. Solution is to get a better tune. Another possibility is fueling. Best bet is to check a learning view to see if either of those two things are a reasonable cause for your particular car. Another option is that the Cobb timing tables (the base and advance) are not smooth, this is the fault of cobb themselves and can very well cause a stutter on it's own, even if you aren't knocking. This is why open source, or a custom pro tune is beneficial for driveability.

If mine still does it after being flashed with a proper ROM (see my post on page 5), Cobb is not what is causing my issue.