I think all the coil packs are the same. He's just saying if you've got the harness flipped over the left side packs won't fire the right side cylinders, and vise-versa.

Any chance you've got like the #2 and #4 coilpacks swapped? That might cause #1 to fire, #3 to misfire, and #2 and #4 are just totally out of time? (I'm just guessing... I don't remember the firing order off hand to know if that's really probable.)

Given the wiring path along the valve cover and wire lengths I don't think you can swap the coil wires.

Well, I remember with my car I had to do a little research because there were a few ways to run the harness. I didn't take any notes when I had pulled it all off, so when it went back on I was working from a blank slate.

If you kept things relatively organized, you're right, you probably can't get them out of order.

Dean let my engine won't be back until after I get new valves made so feel free to ask for parts you might want to trouble shoot. I don't think you have bad packs but you never know.

I have spark!

Tried all 4 packs and couple extra and 9(1 stock, 8X1 step colder) or so plugs on all 4 leads. They all produce a very fine little spark. The one new one that the cylinder had fired sparked extra bright/large until I cleaned it and it went back to matching all the others.

I assume it should be a fine little hair thin spark not a burst of flame or firework looking spark, right?

So how the hell can I test injectors are working?

Is there enough fuel at start up that a Q-tip or piece of paper through the spark plug hole would get wet?

If you are turning it over and it is not firing you should be able to take a plug out and it should be wet. Your ecu has starter primer and starting enrichment cycles that put a bit of fuel in the motor to get it fired. Enough fuel where you should see it on any plug

Have you double and triple checked your crank sensors are plugged in?

Can't think of what it would do off the top of my head, but it is possible to switch the coil packs to run on the wrong cylinders. I put them on wrong on a friends car. We caught it a couple minutes later though and the engine was still out of car. They have the wiring clips that are dark or light colored. The dark/black goes toward the back of the car. the light is in the front. I doubt it's your problem but it could be worth checking. Also have you tested that you have good fuel pressure to the injectors? I ran the fuel lines backwards when I put my motor back in :( maybe you got just enough fuel for it to fire on one cylinder. I can't remember what order they actually get fuel. You can take your crank sensor out and measure the resistance to check it. It's easy to get to. I want to say it should be 1.2k ohm.

Jeff, thanks, but you are a bit behind. Wiring was fine, fuel lines were right and I had spark and no CELs so it was not cam or crank sensors.

Well, many, many diagnostic hours later, it is all back together and should start in the morning. I don't want to start it tonight since I am not ready to go do the initial warm up and break in drive.

I slightly fear I have already worn 1/2 the cylinder hone down cranking the starter 1000 times or so with and without plugs by now. :(

I am now an expert at all things spark and side feed fuel related without removing the manifold. I could probably give Cory a run for his money swapping injectors. ;)

After a bunch of other stuff and ensuring the injectors were getting signal, there was nothing left but the injectors themselves.

I have developed soon to be patented 1/2" Wide 8.5" Long specially folded strips of fuel sensor paper made from 100% recycled material that I will be marketing for testing if cylinders are getting fuel. :) Far easier then installing and removing plugs.

The first thing I would tell would be injector testers is that removing injectors dumps fuel into the cylinder which may give you a false positive. Make sure cylinders are dry before inserting the test strips and applying power to the injector. :)

I would also suggest having a fuel pressure tester in line during all testing. it is quite easy with side feed injectors for the injector to appear to be installed but the bottom seal not be tight and fuel leaks into the cylinder around it as soon as the pump comes on. The gauge immediately tells you this when it doesn't hold static pressure with the pump off.

So after much swapping, drying, testing, retesting, etc. I had 3 stuck injectors. Remember I said they had recently been cleaned and flow tested so they should/could not be the issue???

Apparently you are not supposed to store freshly cleaned and flow tested injectors for 4-6 months at least not without telling the people dong the cleaning that you will be. The normal after cleaning lube/rinse does not really hold up to being left dry and the injectors can get stuck closed. Ones pulled from use are susceptible to sticking as well if stored with just fuel in them for an extended period.

Bashing them in the ultrasonic cleaner again for a while frees them up so the open again. We will see in the AM.

Fired first spark of first crank! Warmed up and then went for a 30 minute drive. Runs fine!!! No leaks or anything obvious except burning off the oil on the manifolds from the blow up almost 2 years ago.

Put the snows on as the 615s have a definite flat spot and are basically slicks.

:banana:

Fucking Awesome! This is great to hear! Congrats on solving the issue and it not being expensive!

Congrats!

OK, I still have one weirdness I can't figure out.

The car stumbles either transitioning from closed to open loop and/or it hits 9ish PSI boost, not sure which. It is pulling hard and then abruptly decreases the rate of acceleration.

Cleaned the MAF and checked all the vacuum lines. O2 numbers appear fine, but it also appears to happen just as A/F goes from near stoic to 12:1 which is why I also mentioned closed to open transition.

Stock base map with TGV CELS disabled. Tried Stage 2 real time with no change.

I think 9ish PSI is wastegate spring pressure. Right?

I was thinking boost solenoid, but it appears to work fine. Not sure it can maintain duty cycle, but it opens and closes fine. I flushed it and all the lines in case they ingested something. The pill is on the correct side of the T (High pressure side).

Anybody got a known good spare solenoid or one I could borrow to test with?

It is weird though, sometimes it goes right through 9psi, especially in I think 4th gear.

I still have my stock one you can borrow.

Dean you are already laying into it??? I like your style!!! Why don't you log it and post it up?

Really doubtful it is the boost solenoid.

You should be running on wastegate boost for the first 1000 miles anyway. You also really should not be on the stock map with the mods you have, especially if you are exceeding wastegate boost.

What do your fuel trims look like in closed loop before the transition? Post or email logs.

A post-MAF air leak is always a possibility with that Gimmick inlet. We can test that for you and log the car to figure it out on Monday.

I will probably log something today.

Yes, I am laying into it. After seeing the sides and skirts of the pistons that came out of my '04 motor I decided to more or less following the break it in hard theory here.

Prenote: I also decided to use Comp Cams break-in oil which has high levels of anti-galling agents but is still basically a Dino oil. I think I did the math and it was cheaper than dino oil + break in additives.

0. NOT RECCOMENDED: Turn it over countless times with 1 cylinder firing or none with and without spark plugs installed and occasionally fuel in the cylinders. :(

1. Warmed up to 150+ degree oil temp.
2. Go drive for 30 minutes on and off throttle a lot with increasing power levels.
3. Change oil/filter. I was surprised how cruddy the oil was after so little time and appeared to have some other film floating on it.
4. Go drive it enough to get all the OBD2 thingies ready for smog.

I also figured if I have a valve problem I may as well snap it off and destroy the motor as quick as possible rather than delaying it. :~:

Also log IAM (or whatever Cobb calls it), Feedback Knock Correction, and Fine Learning Knock Correction. The ECU may be pulling timing due to knock at the CL/OL transition. The stock 04 STI's do this often from the factory, and it is even worse with mods on a stock map. It is enough to feel it sometimes, and that would not be good for a new engine... or a even a broken-in engine.

I was thinking there was some stuff that did not get applied with just the real time Stage 2 map that I might need like open loop delays. Going to check that next.

Stage 2 real times should be handling wastegate duty cycles though to exceed wastegate spring levels which is why I am thinking solenoid.

I hate that Subaru believes almost all vacuum and low pressure air lines do not have clamps on them. I know it is to save money and increase service since they will likely hold through warranty and only get dry and crispy and leak after. When I put it back together, I put them on just about everything and all intake duct ports. Was going to spray some carb cleaner around this AM as well to check.

It idles fine with no hunting until it stumbles slightly when the fans kick on.

Does that guy's method apply equally well to turbo'd engines?

We also recommend boost during break-in, but just not mores than 7-8 psi.

Just confirmed Closed loop delays are not in the real time map. Switching to COBB Stage 2 base map.

In theory, higher combustion pressure = more ring pressure.

I am not going to 19+PSI but am willing to go beyond wastegate spring pressure at 9 PSI. I am also watching A/F, EGT, etc. Maybe I will break this motor in a new (for me) and wonderful way, but it is how I have decided to proceed.

I probably over/under torqued the head bolts and bent that intake valve anyway so WTF. :D

Sounds reasonable. I really think you should have a true break-in map though and have it not able to exceed the max boost you are willing to accept. Off-the-shelf stage 2 map (flashed, not real time) with recalibrated boost would be okay, as long as you checked thoroughly the AFR and knock.

Way to go Dean... Happy you have it running again.

Flashed and OBD2 connector pulled out and rerouted back to logging position. Why they put these connectors on the drivers side right in the knee zone and angled so they are in the way I will never understand. OK, I understand it is for the smog techs, but come on. Move it to one side, preferably the left and angle it any other way. And don't get me started on the plug shape/design and Cobb's choice of hosing and cable exit.

And it can't hurt to have a cleaned solenoid.

I know the Cobb Stage 2 maps pretty well. They or a rescaled version of them were what my car ran on since shortly after I got it and started using streettuner.

I thought going to Stock would be ubber safe, but realize now it really wasn't I only really needed the stock intake map/box as I fear my custom intake map for the Injen might have contributed to the last blow up.

Time for a test drive and some logging.

Ditto.

Much happier. Hesitation gone, boost mostly controlled or over controlled still a bit, but better.

Anybody have the cobb stage 2 '07 STI wastegate table for comparison? I remember I am using a VF43 with an '07 pill so the '04 VF39 larger pill tables are not quite in control. Still sticking at 9PSI at times but getting through it in 3rd and 4th and 5th mostly fine. 6th hits 9PSI like a wall. or was it 5th and 6th?

I think the problem was a combination which is why it was so weird. Closed loop delays were probably a big part of it and wastegate teh rest with maybe the solenoid cleaning helping a little.

It was really weird having the car cease to accelerate for a second or what seamed like an eternity.

Thanks to those who offered their stock solenoids but looks like I don't need them.

The 07 STI has a small 0.8mm pill that usually needs to be changed out to the earlier size (1.2mm) in order to control boost properly with a aftermarket downpipe. I think Cobb even recommends or requires this with their 07 Stage maps. You are playing with fire using the 04 wastegate tables with an 0.8mm pill.

OK, will grab the '04 pill. Was thinking that might be easier. Will probably just swap hoses. I think I have them.

Looking at logs, the knock sensor appears to be wonky... It throws out random knock under little or no load and even under deceleration negative rel. pressure, teens throttle position. WTF???

Do forged pistons cause enough noise for the knock sensor to go off? Dropped my damn multiplier all the way to 0. Do they go bad?

You sure you don't want to run a MBC, Dean? Everyone who tries the Hallman Pro with an appropriate tune loves it. I've run one for years.

I have a turboXS MBC lying around and may end up with a MBC tune at some point but not now.

Changing to that will make no difference if I am getting random knock signals stealing my timing.

False knock is a PITA. Never heard of forged pistons doing it (and those KB's are pretty quiet anyway). Usually it is something rattling like a loose or broken heat shield, or a loose part tapping against the block/heads/manifold. Sometimes raising the knock correction thresholds above the offending load or RPM can help.

Knock sensor is unlikely the culprit but I have one you can try out to see for sure.

This may be why you are running into the hesitation. When the IAM drops below .35 (or whatever it is set to) or feedback knock correction exceeds a set threshold the ECU will go to wastegate boost momentarily or until the knock goes away for a while. Also you can feel it when a lot of timing is pulled.

This morning I am going to make sure the knock sensor is oriented correctly and properly torqued.

Yes, apparently that makes a difference.

The harness should leave the sensor at a 60 degree angle to the right rear off a front to rear line and torqued to 17.4 ft-lbs.

I wonder if the knock sensor basically being completely loose would do it? :rolleyes:

Not even finger tight. What idiot put that on? Oh, yeah.

I wanted to put on the turbo heat shield and big intercooler anyway, right?

let's go see...

Well, that did it.

I do still need an 0.043" '05-'06 STI restrictor pill/hose. I guess the one from the '04 went with the cracked VF39 Turbo.

Stop by. We have a bin full of the 1.2mm pills.

:lol:

If you were to take a microphone and pound it on the floor, what do you think the audio recorded it would sound like? A knock sensor is really just a mic bolted to the block, right? Bouncing it all around would totally make the ECU go crazy trying to figure out what it's listening to!

Congrats Dean on your success. I just swapped out my own block too (ralispec). I just passed 1200 miles. Still have my paranoid ears on. Did the hard break in thing this time.

Time to update your mods list linked in your sig. :)

How'd you like the Cobb Stage II motor?

Woops. updated.

The cobb block was good. I got about 60K miles on it. It was noisy from day one but ran great and didn't burn excessive oil. It started eating oil bad (1 qt/500 miles) after about 50K miles. Consumption had worsened to 1 qt/200 miles by 60K. Compression was low but even. I suspect oil rings are toast. No smoke from oil fill. Exhaust smoked from hard accelleration. Didn't smoke on boost alone with artificial load via brakes. I thought it might be the stock turbo (I was near 100K miles and a 19psi tune) so that got upgraded :D. Still smoked so I pulled the motor and took my time gathering parts and started assembling. New block is quieter than the cobb. My cobb block was louder than other cobb blocks. If they still made them I would have gone with them again.

50K seems to be the limit for built blocks with few making it that far.

If I have to build another one I'll adjust my goals towards longevity. Perhaps longer skirts and tighter than 'normal' P/W clearances.