StkmltS

Why don't you just do a leakdown test? That's what I'd do at this point.

The last thing I wanted to do tonight was another leakdown test on this stinking car... but I need all the help I can get so I'll take every reasonable suggestion seriously.

Results from tonight's leakdown test:

Cyl 1: 27%

Cyl 2: 23%

Cyl 3: 24%

Cyl 4: 27%

I can't tell if I'm frustrated or encouraged. These #'s are different than from my first leakdown probably because of the new pressure regulator I have on my compressor. I fell much better about these #'s than the results from my first test. This time the engine was warm enough to be uncomfortable to the touch, but much lower than operating temp (sat for about an hour before the test).

Prelude: The glass is half full, I'm not giving up, and I'm fully committed to nickel-and-dime myself to death until I find a solution.

Assuming I try my best but never manage to get this issue resolved with something other than #INANSB, what would the harm be in flashing a stage 0 safety tune and bumping up my idle speed a little bit just to keep the car from stalling? I would then politely drive the car until something obvious happens.

Preventing the big b@#m is what everyone wants, but if I really do have a cracked ringland then hasn't my engine's countdown clock already started? If I take good care of everything else then why not just let failure run its course and replace the SB when it becomes necessary? Classic VF40 turbo failures excluded, I don't see any benefit in replacing the SB prior to failure as opposed to waiting for it to go belly up.

log #1 doesn't look too bad really. A tuner can make some small tweaks to your tune to get the actual boost closer to your target boost. But you'd probably want to get the issue with that cylinder resolved before tuning.

 

log #2, I have no idea. Was it surging during that log?

I couldn't feel any surging, and it definitely didn't feel like the log looks.

Logged a little last night and I'm curious about the difference between my target boost and actual boost. Bla bla bla target boost assumes barometric pressure of 14.7 psi... yesterday at the time of my logs the reported barometric pressure in Cincinnati was 14.67.

Chart 1 shows that my boost peaked at 10.5 psi, however on my mechanical boost gauge it looked like my actual boost was closer to 12 psi. My gauge isn't permanently mounted so I could have been looking at it wrong (it's pointed in a funky direction).

What the crap is up with the spikes in my second log? Could I be having MAP sensor problems?

A little bit of time is showing that my LVs aren't much different than what they were before I fixed the leak.

since your valve clearance job, did you do a leakdown test?

No.

But you still have a misfire?

Yes, no noticeable change.

I ordered a timing belt kit today so I may be able to verify my valve clearances this weekend.

I was hoping to wait and do my TB this winter with my #ynansb project, but who knows how long it'll be before I have enough 'proof' to justify the cost of a rebuild.

The rubber end are gandhi as of last night.

I'm going to find somewhere that has a professional smoke tester and pay them to find my leak.

I smoke tested last night with great results. My air compressor's built-in air regulator doesn't regulate well below ~20psi, so I upgraded to a digital regulator from harbor freight. Now my paint can 'system' doesn't pop open. With the help of a little seafoam and a steady air source at <5 psi I was able to get smoke for 5+ minutes at a time.

Video results:

It was leaking between the throttle body hose and in intercooler.

I removed the two soft rubber lip thing-a-majigs at either end of the hose, cleaned the inside of the hose, then reinstalled everything.

No more air leaks (for now)!

(pics added to post #1)

Bad news: Misfire still present.

Good news: My LVs seem to look a little better, but that might only be because I reset the ECU.

My drive to work this morning was super crappy because of heavy rain and heavier traffic and I didn't get above 4k rpm once.

On my drive home today I should see a few chances to log one or two light runs from 2k-5k.

How much of a difference should I be seeing between boost and target boost?

Hoping things will improve when you chase down that leak.

 

I didn't see "Fuel correct (%)" in your log, does that tend to get a large value?

I wasn't logging fuel correct %... but I am now. I'll post #s after my drive home.

The sound of my misfire -> Subaru Misfire EJ255 2005 Legacy GT:

First thing this morning I had a CEL and this is what cyl #2 was doing... crazy.

The spike in RPM was me seeing if higher RPMs would change anything.

Look at how much my fuel learn % was fluctuating.

After resetting the ECU to clear the code the misfire all but went away.

I feel like this is something I should be encouraged about. I'm gonna pick up a 'quality' air regulator from HF on my way home and try to smoke out my problem tonight.

BtSsm_20150805_062340.csv

just referencing this post since it has some good pics of the intake manifold. wonder what could be leaking? could u post a video of another smoke test?

I'm gonna have some time to mess around with my car on Friday evening so I'll give it my best shot.

My DIY "system" is a metal paint can with a hose going in and a hose going out. Last time I used an old cotton rag & motor oil and it didn't work very well. Would using a regular smoke bomb cause any damage? This time I'll try wetting a rag with a little Seafoam. It's obviously save for the motor and anyone who's ever used it knows it smokes like a mofo.

My throttle body hose is good to go. Smoke from the leak I couldn't locate was exiting the motor from somewhere underneath the manifold and it was creaping out where the manifold splits for cyls 2/4 (diagram in post #1). I've checked and rechecked the T fitting and surrounding hoses under the IC.

Everyone says our plastic manifolds don't crack, but if there was a crack I imagine that smoke would leak out just like it's doing on mine.

Riderdude125... I'll look him up.

My Christmas bonus can't come soon enough.

Sent from inner space.

Screw cylinders #1-4, screw turbos, screw misfires, screw banjo bolts, screw Porsche, screw the internet as a whole, screw Tris (RIP), screw #ynansb, screw me, screw BtSsm (with all due respect), screw Subaru of Ameria, screw the state of Indiana, screw Henry Ford, screw the new Corvette, and especially screw my car.

Why me, whyyyyyyyyyyyy!

Sent from inner space.

2 seconds after starting my car after work.

Great, now even BtSsm is telling me "ynansb".

http://images.tapatalk-cdn.com/15/08/03/e3fefb39936db62bc6d648875eb5f1d9.jpg

Sent from inner space.

Engine temp has zero affect on the misfire. With a fully warmed motor I've driven to point A whilst fighting a terrible idle, parked at point A for 5 minutes without turning the engine off, and then drove to point B with the idle mysteriously perfect.

Ain't no way I'm taking off the valve cover again (yet). I sheared off one of the bolts when I put it back on after my valve adjust-a-roo and I don't want to push my luck asking it to seal correctly again without extracting/replacing the bolt.

I DIY smoke tested it once and found a leak coming from under the intake manifold on the driver's side. I didn't fix it because I couldn't get to it and I couldn't see exactly where it was leaking. Now that I have a borescope I may smoke it again and use the camera to find the exact location of the leak.

fyi and from my experience, the misfire CEL will be thrown if the roughness misfire counts goes like 35+ several times in a row. Below that, the CEL is usually not thrown.

 

Also I can't remember if you mentioned this before, but did you recheck clearance once the timing belt was in place right before you put the valve cover back? Maybe it is still a bit off?

 

What are you LT fuel trims btw?

I forgot to check the valves after putting the belt back on (shaming begins... now).

See LV in post above (I added it as you replied).

Glad to hear there is some improvement. So these days, what sort of misfire counts are you typically seeing?

 

When you say they are audibly present, does it have a constant beat? Or is it an occasional muffled pop sound?

 

Regarding tuning, I use RR to do all of the editing. And a Tactrix OP 2.0 cable (and corresponding ecuflash software) to actually flash the ECU. Tuning is a vast subject, but you can start with a few things. For example, I would suggest removing the Closed Loop to Open Loop delays. That's often what people are referring to when they say the stock tune burns exhaust valves. See the first section "Closed to Open Loop Delay" in this post for example.

That stock-ish map is exactly what I've been looking for, thanks for the link!

My misfire sounds like a muffled pop.

It's sporadic and without a noticeable pattern, but it's consistent enough to be heard once or twice every second or two. The frequency could be compared to an automatic gun set to fire 2- or 3-round bursts. I'll try to take a video of it tonight. My small exhaust leak makes it pretty easy to hear.

My typical misfire count for cyl#2 is anywhere between 4 and 8, but sometimes far less.

Strange things is, it only occurs below 3k RPM. Is that a limitation of my ECU/sensors, or does it really stop >3k? .

Link here for the csv file.

In this particular log I have knock, most of my other logs show knock at zero.

Real numbers from today (15,267 data points):

Cyl#1 is misfiring 0.000% of the time (0/1,5267)

Cyl#2 is misfiring 9.792% (1,495/15,267)

Cyl#3 is misfiring 1.827% (279/15,267)

Cyl#4 is misfiring 0.210% (32/15,267)

So I took my car on a 600 mile trip two weeks ago, and I've been driving it daily since then. Prior to my valve adjustment I was getting a CEL every other day, sometimes multiple times per day. Since I last reset my ECU 750 miles ago my CEL hasn't come on. My last 150 miles have been strangely problem-free*.

My misfire is still audibly present (BtSsm confirms) but my low idle problem is completely gone. As best as I can remember it hasn't tried to stall even once since I got back from my 600 mile trip. Until recently, it was bad enough that I almost dreaded driving the car because I was fed up with it stalling in stop-and-go traffic, at stop signs, traffic lights, intersections, etc.

- My LVs indicate that I probably have a vacuum/boost leak somewhere... or a crappy cylinder.

- Per my borescope pics my cylinder #2 looks pretty healthy.

Could it be that my engine has just now become familiar enough with the adjusted valves to keep away the low idle? I kinda want to reset my ECU to see if the idle problem will come back but then driving the car will be unpleasant again.

If my low idle doesn't come back I'm just going to get the car off the stock tune and be cautious and drive it as-is.

Any open source tuning recommendations? For now I'm going to be very conservative and do it all myself as I learn along on the way. My plan is to use an already proven map that doesn't require any supporting mods, and slightly tweak it for my car. I'm not looking for more power, I just want to take everyone's advice and get off the stock map.

*Oil is still leaking from my OCV and smoking just enough to stink up the area whenever the car isn't moving, but that's totally unrelated and nothing's fault but mine.

I took the car to the mechanic shop who suggested change the catalytic converter, after doing that... the car running normal with no missfire...i will drive the this week to make sure there's no more missfire.

 

Sent from my SM-T230NU using Tapatalk

That'll be awesome if it fixes your problem!

Which cat got changed? Did your mechanic say why it needed changed? That seems like an unusual diagnosis for a misfire.

wow, your missing cylinder came back? awesome news! did you do anything else to it, or did it just come back? I'll bet your valve clearance adjustment did it, maybe it didn't come back right away because there was a little carbon buildup between the valve and seat, and it got hammered away eventually?

 

looks like (and sounds like) now you just have a classic post-turbo boost leak. the lean low load cells, and the rich high load cell, and the whistling noise all point to that. are you running the stock tmic? maybe it's separating at the end tank (is the tmic oily near the turbo outlet?). Or perhaps the throttle body hose isn't sealing?

I don't think it came back [fixed] because you can still hear and feel it misfiring. I guess the valve job and several hundred miles of steady highway driving was enough to clean things up a bit and keep the CEL from coming on. My misfire and low idle problems are still very much present.

Yes I have the stock TMIC.

Yeah, do check the throttle body hose. It is notorious to come apart after a while. See here http://legacygt.com/forums/showthread.php/pegged-negative-af-learning-all-ranges-no-knock-iam-1-225843.html

Although I would note that when it happened to me, all values were getting negative, probably because the leak was very obvious in my case.

Bummer, it was only a lose hose clamp on the throttle body hose. Maybe I didn't tighten it all the way when I put thinks back together after adjusting my valves. Back to the drawing board.

Despite the motor issues I've been dealing with I must admit... after a wash, clay bar, and wax, it's a treat having a 10-year old car that looks this good.

Any thoughts about the appearance of my cylinder #2?

I drove the car about 630 miles since Monday (trip to Pittsburgh) and about 100 miles into the trip the car developed a loud boost leak. If you listen very closely you can hear it hissing at >0 psi, and anything over 5 psi makes a super loud "pssssweeeee" sound. I just stayed out of boost the whole time and hoped nothing popped (it didn't). I didn't get one single CEL the whole trip, which was very surprising and fortunate because I forgot to bring my bluetooth OBD adapter and my OTG adapter. 600+ miles without cruise control would have sucked.

I haven't found the leak yet but I'm pretty excited because it's another potential fix that's less time/money than #Inansb. I know I'm grasping at straws, but a guy's gotta have dreams.

Here's my fun LV from this morning.

My clinder #2.

The 90° attachment for my new borescope doesn't work well enough to see the cylinder wall. Despite that the camera is still easily worth the $20. It works great with my OTG cable on the two android devices i tried, and the free CameraFi app is excellent.

[ame=http://smile.amazon.com/gp/product/B00EQNCMTY?psc=1&redirect=true&ref_=oh_aui_detailpage_o01_s00]This one[/ame]

It'll be awesome if it works on my phone using the OTG cable I bought to use with BtSsm.