Another H6 3.0R misfire thread / mystery - Subaru Outback - Subaru Outback Forums
 7Likes
Reply
 
LinkBack Thread Tools Rate Thread
post #1 of 55 (permalink) Old 05-20-2017, 07:05 PM Thread Starter
Registered User
 
Dispatch20's Avatar
 
Join Date: Feb 2016
Location: Syracuse, NY
Car: '08 3.0R Limited Sedan; '09 Outback 3.0R
Posts: 151
Feedback Score: 0 reviews
Solved: Another H6 3.0R misfire thread / mystery

Solution Summary: The timing chain on Bank 1 had jumped 1 tooth on the crank sprocket. Ultimately, the giveaway was 30psi of reduced compression for Bank 1. (Bank 2 = 190psi, Bank 1 = 160 psi).

-----------------------------------

I've scoured every existing thread I could find, and I'm coming up short. I appreciate any help (or even just sympathy!). I just bought a 2009 Subaru Legacy 3.0R with 155k miles as a project to tinker with. The car wasn't well maintained, and it has CEL for Bank 1 misfires P0301, P0303, P0305. No other codes are thrown.

One advantage I have in diagnosing is that I also own 2 other 3.0R cars that run well. Swapping parts between cars to test functionality is pretty easy and economical.

Misfire Symptoms:
- Occasional, but persistent misfires on 3 cylinders of right side (passenger) bank: P0301, P0303, P0305. The misfires at idle never really go away.
- You can feel the misfires, so I think they are real. However, there are not enough misfires to ever get anywhere close to stalling (RPMs don't even really dip down)
- Worst misfiring on cold-startup, and it's enough to throw a blinking CEL with some minor RPM fluctuation
- There are fewer misfires after warm-up and driving. Few enough to not throw a CEL, but they still exist as you can see "roughness" or misfire counts continue to go up at avg rate of ~0.5 - 1 per second per cylinder (ie. 5% misfire rate)
- The misfires (timing and quantity) appear essentially random across the three cylinders.
- During driving, there tends to be no (or far fewer) misfires unless coming on and off the throttle lightly while coasting.

Observations:
- Car is very fast and runs smooth and hard on full throttle. Feels faster (likely lighter) than my other 2 cars. This makes me think I probably don't have a fuel pump issue, or a timing issue.
- The car never stumbles when you try to accelerate. Nor does it tend to stall when approaching a red light.
- Dash computer says 25.8mpg. That's way higher than my other two 3.0Rs
- Compression tested: Cylinder 1 - 195psi, 2 - 220, 3 - 200, 4 - 220, 5 - 200, 6 - (I didn't bother; too annoying) (Update: This symptom was a red flag I initially ignored. Gauge red high, but the delta for each bank was just too high even though it was '10%')
- No leakdown results as I couldn't figure out how to find TDC for each cylinder
- Short term fuel trim (STFT) tends to go to -14% at cold when misfiring is the worst. Once warm, it goes within spec (<10 for both STFT and LTFT). -14% STFT means the ECU is leaning out the fuel mixture. Why would it do this?
- Cold Vacuum is -19 inHg; Warm vacuum is -21 inHg (my other car runs -22 warm vacuum). MAP KPA reads ~30
- Vacuum drops / fluctuates by ~0.5 InHg during a bout of misfires. Unclear if it dropped simply because RPM dipped or because valve didn't close right?
- The VVT / AVCS angles for Bank 1 and Bank 2 appear to follow each other closely via FreeSSM
- There are no timing chain sounds, so its unlikely the chain jumped timing due to a failing tensioner
- The oil looked good (not metallic, not fuel smelling) when I changed it out after 5k miles
- The coolant expansion tank doesn't bubble at all, or look oily
- No oil had seemed into the spark plug holes
- No external seeping on headgaskets. No reason to suspect internal gasket leaks.
- Battery voltage and alternator charging voltages seemed normal
- Engine starts quickly. However, after sitting a week the car did crank much longer before it started.
- Fuel injectors do not appear to be leaking at the fuel rail o-rings
- No smoke coming out of the exhaust at any time
- Unclear what octane fuel the previous owner used from 62k miles to 155k miles.
* This may not mean anything, but unplugging the Bank 1 upstream O2 causes the car to stop misfiring (or stop reporting misfiring via OBDII). I cannot confirm whether this is a weird glitch, or whether the car isn't misfiring anymore!

Steps taken so far (all on just Bank 1) that didn't help:
- Previous owner tried putting in a new ECU (after other debugging, and obviously never found the root cause)
- Fresh oil change with my favorite: 5W30 Castrol Edge Synthetic with Subaru OE filter
- New Upstream/front Denso O2 sensor for Bank 1
- Swapped Downstream Bank 1 sensor from functional car
- New plugs
- Swapped coil packs from functional car
- Swapped fuel injectors from functional car
- Swapped VVT AVCS OCVs from functional car
- New ground straps to each valve cover
- Sprayed carb cleaner all around engine/manifold and RPMs never fluctuated (i.e. no vacuum leak around manifold and airbox)
- Swapped knock sensor from functional car

The classic H6 half-engine misfires! It seems like a strange coincidence for the right 3 cylinders to have a random misfiring issue, unless every cylinder had something in common. I'm having a hard time finding the common issue here.

Possible Issues:

- Are the valves on half the engine built up with carbon? Are they not sealing well? Worth trying seafoam and techron? I realize a leakdown test would have shown a minor valve seating or clearance issue better than a compression test. Is it worth buying a borescope to try and visually look for carbon buildup?

- Do the valves need adjusting (i.e. too little clearance after 150k)? Maybe they are already burnt? What are the odds all three Bank 1 cylinders developed a valve issue while Bank 2 runs fine?

- I think there is a single fuel regulator on the 09 H6, and two fuel rails had their own damper. Is the right fuel rail damper not working right? I guess I could swap the fuel rail from my other vehicle over to this one, but that seems like a real long shot.

- Am I missing something with the VVT? The advance angles / etc all seem fine, and the car drives hard on full throttle.

- Would an EGR issue affect both banks equally? Same thing with an IACV and PCV issue? Would they ever cause misfires?

- Is Bank 1 timing off by 1 tooth? The compression was about 20psi less on Bank 1 than Bank 2. That's only 10%. Maybe the intake or exhaust cam is slightly off? (Update - This turned out to be the issue!)

Any other ideas? I really don't want to spring for a $1400 JDM engine. It's just a ton of work replacing the motor or pulling the head off my existing one. But maybe I'll need to in order to make this car pass NY inspection.

Thanks!

Useful H6 Misfire Reference links:
https://www.subaruoutback.org/forums/...-llbean-3.html
https://community.cartalk.com/t/misf...lbean/73243/37
H6 3.0 dead bank 1,3,5. - 1990 to Present Legacy, Impreza, Outback, Forester, Baja, WRX&WrxSTI, SVX - Ultimate Subaru Message Board
https://www.losaltosonline.com/speci...he-road/52795-
https://www.subaruoutback.org/forums/...misfiring.html
2003 Subaru Outback H6 random misfires all cylinders - 1990 to Present Legacy, Impreza, Outback, Forester, Baja, WRX&WrxSTI, SVX - Ultimate Subaru Message Board
Help needed!!06 B9T misfiring in Cylinder 2,4,6 - SB9T.com


Last edited by Dispatch20; 06-09-2017 at 08:07 PM. Reason: Adding solution
Dispatch20 is offline  
Sponsored Links
Advertisement
 
post #2 of 55 (permalink) Old 05-20-2017, 11:17 PM
Registered User
 
F-Bobby's Avatar
 
Join Date: Apr 2017
Location: Watertown
Car: 2003 Outback Sedan 3.0 2004 Fozzie X
Posts: 103
Feedback Score: 0 reviews
Does the Car have an EGR system? And does that EGR tube go onto the bank where the misfires are?.
Also have you checked for vacuum leaks?


Auto Tech
F-Body Fanatic
Scooby Noob.
F-Bobby is offline  
post #3 of 55 (permalink) Old 05-21-2017, 02:05 AM
Registered User
 
Greenmamba's Avatar
 
Join Date: Feb 2011
Location: Chigwell Essex, England
Car: Outback 2006 3.0Rn 5EAT Wagon.
Posts: 421
Feedback Score: 0 reviews
Knock sensor?

Nikinikitembonosorembooohmamuchigummagummaguchi
Greenmamba is offline  
post #4 of 55 (permalink) Old 05-21-2017, 09:37 AM Thread Starter
Registered User
 
Dispatch20's Avatar
 
Join Date: Feb 2016
Location: Syracuse, NY
Car: '08 3.0R Limited Sedan; '09 Outback 3.0R
Posts: 151
Feedback Score: 0 reviews
Quote:
Originally Posted by F-Bobby View Post
Does the Car have an EGR system? And does that EGR tube go onto the bank where the misfires are?.
Also have you checked for vacuum leaks?
It does have an EGR. It appears to take exhaust from the Bank 2 side of the block and has a valve that runs a metal pipe onto the bottom of the intake manifold. This seems like it would affect both banks equally, right? In theory, it should.

I've looked for vacuum leaks and sprayed carb cleaner all over the engine. I have not tried to pressurize the manifold and then spray soap/water yet. I am pulling pretty good vacuum at idle, and the O2 sensors aren't telling the ECU to add extra fuel (+STFT) to compensate for the unmetered air. Although I'm not sure if that is how the H6 works (i.e. MAP vs MAF).

Quote:
Originally Posted by Greenmamba View Post
Knock sensor?
After your suggestion, I went out and swapped the Bank 1 sensor between my cars. No change in behavior, but it was a good idea.

Thanks for the ideas / questions!
Attached Thumbnails
Click image for larger version

Name:	egr.JPG
Views:	1035
Size:	48.3 KB
ID:	350017  
Dispatch20 is offline  
post #5 of 55 (permalink) Old 05-21-2017, 11:34 AM Thread Starter
Registered User
 
Dispatch20's Avatar
 
Join Date: Feb 2016
Location: Syracuse, NY
Car: '08 3.0R Limited Sedan; '09 Outback 3.0R
Posts: 151
Feedback Score: 0 reviews
Another observation:

My healthy 3.0R Outback runs 15 degrees of total timing at idle and stays pretty steady at that. My misfiring 3.0R starts at 15 degrees, and then slowly and choppily goes up to 22-23 degrees (over course of 30 seconds). It seems like the ECU see's misfires and perhaps advances the timing to try and clear them up?
Dispatch20 is offline  
post #6 of 55 (permalink) Old 05-21-2017, 06:11 PM Thread Starter
Registered User
 
Dispatch20's Avatar
 
Join Date: Feb 2016
Location: Syracuse, NY
Car: '08 3.0R Limited Sedan; '09 Outback 3.0R
Posts: 151
Feedback Score: 0 reviews
I attached a graph using data from RomRaider logger. Does this give any clues?

The car was running in open loop (i.e. not using upstream O2 sensors for AFR trimming) for the first 15 seconds. And then the AFR corrections (i.e. short term fuel trims) kicked in and brought both banks AFR to 14.7. The misfires/roughness counts started 32 seconds in, although it isn't clear whether they may have been happening prior to that or not.
Attached Thumbnails
Click image for larger version

Name:	loggindata.gif
Views:	331
Size:	36.6 KB
ID:	350049  
Dispatch20 is offline  
post #7 of 55 (permalink) Old 05-21-2017, 08:25 PM
Registered User
 
Join Date: Jun 2006
Location: Eastern ON Canada
Car: 07 OBW 2.5i D-4AT
Posts: 15,457
Feedback Score: 11 reviews
When first started and in OL, the fueling is based, in part, on the last A/F Learning (LTFT). There would be one for each bank. When in CL, the LTFTs can vary, and will be set for ranges of air flow. The A/F Learning values can be a lead into the AFR levels in OL, but also should be considered in CL, especially if the A/F Correction (STFT) is significantly off. In other words, both sets of data might inform what's driving the significant negative trim in Bank 1. (If the Bank 1 STFT is -15 and the LTFT is -10, that might be why the right side is misfiring.)

Also consider the voltage data from the two O2 sensor (rear sensors). The rear sensors can affect fueling, and, they can provide verification of the fuel trim levels.

Your 09 H6 should have a MAF sensor, and there should be a Romraider parameter for it.

If your Romraider log has more parameters, perhaps upload it as an attachment to a post here (has to be converted to a .zip file).
plain OM is online now  
post #8 of 55 (permalink) Old 05-21-2017, 10:21 PM Thread Starter
Registered User
 
Dispatch20's Avatar
 
Join Date: Feb 2016
Location: Syracuse, NY
Car: '08 3.0R Limited Sedan; '09 Outback 3.0R
Posts: 151
Feedback Score: 0 reviews
Quote:
Originally Posted by plain OM View Post
When first started and in OL, the fueling is based, in part, on the last A/F Learning (LTFT). There would be one for each bank. When in CL, the LTFTs can vary, and will be set for ranges of air flow. The A/F Learning values can be a lead into the AFR levels in OL, but also should be considered in CL, especially if the A/F Correction (STFT) is significantly off. In other words, both sets of data might inform what's driving the significant negative trim in Bank 1. (If the Bank 1 STFT is -15 and the LTFT is -10, that might be why the right side is misfiring.)

Also consider the voltage data from the two O2 sensor (rear sensors). The rear sensors can affect fueling, and, they can provide verification of the fuel trim levels.

Your 09 H6 should have a MAF sensor, and there should be a Romraider parameter for it.

If your Romraider log has more parameters, perhaps upload it as an attachment to a post here (has to be converted to a .zip file).
Good info, thanks! I've attached a new log from a cold start. I added a ton of parameters so the refresh rate was only about 10 times per second. The LTFT was actually at 0 at idle. I'm not sure if I've had enough driving time for it to register a real LTFT, as the car isn't in great condition for driving. I do recall after a long drive (prior to clearing codes today) I was only seeing a -7 LTFT and a very small STFT on Bank 1. The misfires are quite a bit better after a long drive (i.e. vacuum leak is less at warm, or EGR or IACV is working slightly better when hot, etc?)

I also tried to find unique (and obvious) parameters for the rear O2 sensors voltage readings. I think RomRaider may have an incorrect definition. "Front O2 Sensor #1" and "Front O2 Sensor #2" actually look like a typical front and rear O2 voltage (.8V and .1V) of one of the banks. I took a second log with just those two at a faster 50Hz refresh, but it didn't really show anything interesting.
Attached Files
File Type: zip romraiderlog_20170521_215415.zip (40.3 KB, 84 views)
Dispatch20 is offline  
post #9 of 55 (permalink) Old 05-22-2017, 09:19 AM
Registered User
 
Join Date: Jun 2006
Location: Eastern ON Canada
Car: 07 OBW 2.5i D-4AT
Posts: 15,457
Feedback Score: 11 reviews
Quote:
It seems like a strange coincidence for the right 3 cylinders to have a random misfiring issue, unless every cylinder had something in common. I'm having a hard time finding the common issue here.
Each side has it's own fuel trim control system (i.e., control "loop"), based on feedback from the exhaust sensors. On your 2009 H6, according to the FSM, there's a cat on each side, with an A/F Sensor before each cat, and an O2 Sensor after each cat. (Romraider is probably identifying the latter sensors as "Front O2" sensors.) Consequently, if there's a problem in one control loop (in Bank 1 or in Bank 2), then it's quite possible for that bank to exhibit an issue, such as misfiring, while the other doesn't.

Quote:
Steps taken so far (all on just Bank 1) that didn't help:
. . .
- New Denso O2 sensor
As noted, each bank has two sensors, one before, and one after, the cat. Which one was replaced?

I ask because there was a thread about misfires on one side of a 2005 H4 that the final report indicates was a defective rear O2 sensor. (https://www.subaruoutback.org/forums/...ml#post3512601) In that case, it appears an early clue might have been that the post-cat O2 sensor voltage remained low and the A/F Correction #3 was fixed at -.31% when the FSM suggested it should be positive. In your log, near the end of the 38 seconds, the Bank 1 Front O2 sensor voltage seems to trend down to a low voltage ("low" essentially being below 0.4 V) and the A/F Correction #3 seems to stabilize at -0.31 V. Again, here, the FSM suggests that Correction #3 would be in the 0-1 Volt (positive) range when the engine is fully warmed up and idling. (In your log, the engine doesn't reach full operating temperature.) Might be interesting to see if after your H6 is fully warmed up, those two parameters are similar to the ones in the 2005 H4 case.

Also, when the engine is fully warmed up and idling, or the car is being driven at a constant speed/throttle, the post cat O2 sensor voltages should be relatively stable in the range of 0.6-0.9 V. During the 38 seconds of the log, both rear sensor signals are varying widely, but near the end when the coolant temperature is around 140 F, the bank 1 sensor voltage appears to be steadying on the low side while, the bank 2 sensor is trending toward the higher side. The low bank 1 reading could be because the ECU is pulling fuel at around -12 % on that side to maintain stoich according to the A/F Sensor. If the AFR is indeed at stoic (sensor is reading correctly), and the rear sensor is reading low, then it seems to me that the cat could be bad, the sensor could be bad, or there could be a leak in the exhaust that's skewing the sensor readings.

Incidentally, when codes are cleared, the ECU is reset, and that also clears the "learned" fuel trims (A/F Learning/LTFT). Once the engine is restarted and allowed to run, the ECU will relearn the trims based on the then current operating conditions.
plain OM is online now  
post #10 of 55 (permalink) Old 05-22-2017, 10:43 AM Thread Starter
Registered User
 
Dispatch20's Avatar
 
Join Date: Feb 2016
Location: Syracuse, NY
Car: '08 3.0R Limited Sedan; '09 Outback 3.0R
Posts: 151
Feedback Score: 0 reviews
Quote:
Originally Posted by plain OM View Post
Each side has it's own fuel trim control system (i.e., control "loop"), based on feedback from the exhaust sensors. On your 2009 H6, according to the FSM, there's a cat on each side, with an A/F Sensor before each cat, and an O2 Sensor after each cat. (Romraider is probably identifying the latter sensors as "Front O2" sensors.) Consequently, if there's a problem in one control loop (in Bank 1 or in Bank 2), then it's quite possible for that bank to exhibit an issue, such as misfiring, while the other doesn't.



As noted, each bank has two sensors, one before, and one after, the cat. Which one was replaced?

I ask because there was a thread about misfires on one side of a 2005 H4 that the final report indicates was a defective rear O2 sensor. (https://www.subaruoutback.org/forums/...ml#post3512601) In that case, it appears an early clue might have been that the post-cat O2 sensor voltage remained low and the A/F Correction #3 was fixed at -.31% when the FSM suggested it should be positive. In your log, near the end of the 38 seconds, the Bank 1 Front O2 sensor voltage seems to trend down to a low voltage ("low" essentially being below 0.4 V) and the A/F Correction #3 seems to stabilize at -0.31 V. Again, here, the FSM suggests that Correction #3 would be in the 0-1 Volt (positive) range when the engine is fully warmed up and idling. (In your log, the engine doesn't reach full operating temperature.) Might be interesting to see if after your H6 is fully warmed up, those two parameters are similar to the ones in the 2005 H4 case.

Also, when the engine is fully warmed up and idling, or the car is being driven at a constant speed/throttle, the post cat O2 sensor voltages should be relatively stable in the range of 0.6-0.9 V. During the 38 seconds of the log, both rear sensor signals are varying widely, but near the end when the coolant temperature is around 140 F, the bank 1 sensor voltage appears to be steadying on the low side while, the bank 2 sensor is trending toward the higher side. The low bank 1 reading could be because the ECU is pulling fuel at around -12 % on that side to maintain stoich according to the A/F Sensor. If the AFR is indeed at stoic (sensor is reading correctly), and the rear sensor is reading low, then it seems to me that the cat could be bad, the sensor could be bad, or there could be a leak in the exhaust that's skewing the sensor readings.

Incidentally, when codes are cleared, the ECU is reset, and that also clears the "learned" fuel trims (A/F Learning/LTFT). Once the engine is restarted and allowed to run, the ECU will relearn the trims based on the then current operating conditions.
Interesting! And thanks for the link and thoughts. I had only changed the upstream O2 on Bank 1, as it is the one that typically goes bad and messes up AFR in closed loop. Now that you mention it, I do recall seeing other cases where downstream O2s in Subaru's could do more than examine the health of the catalyst.

I'll try to do two things after work tonight:
1) Swap in a functional downstream O2 for Bank 1. This is a 10 minute job thanks to my other vehicle.
2) Get the car fully warmed up and take a new data log at cruising speed to look for AF Learning #3 and rear O2 voltage.

During WOT, I assume the cars runs in open loop and ignores both O2 sensors. I'll try to take a WOT log as well to monitor for roughness/misfires. My concern with this is that the roughness counters reset as load changes, so accelerating through a lot of RPM may mask actual misfires. I don't recall if the roughness resets on delta of RPM, requested load, MAF air flow, throttle position, etc.

Here's another quirk I don't understand. If I unhook the front O2 sensor to force open loop, then the misfire/roughness counters never increment. I can't tell if the actual random misfires have gone away, though, due to the somewhat subtle and random misfires. In fact, if I unhook a front O2 and also unhook a coil pack, the misfire count doesn't go up, either. I wasn't sure if the ECU was smart enough to not count a disconnected coil pack as a misfire, though. (I did not try to monitor for misfire counts with the O2 connected and a coil pack disconnected. But maybe I should?). If disconnecting the O2 sensors and running open loop operation actually made the car run fine, it would be a huge clue (i.e. timing chain is fine, compression is fine). The car certainly accelerates healthy in open loop (and closed loop). Zero hesitation or stumbling.

Lastly - RomRaider gives me different logging options (PIDs, I guess) for my 2009 3.0R Sedan compared to my 2009 3.0R Outback. It's like it thinks the ECUs are different. I know my misfiring 09 3.0R sedan does have a different ECU as the previous owner had the car at the local Subaru dealer, and they recommended a new ECU as a potential fix. I'd love to identically log my 09 Outback to my 09 Sedan, but the slight differences made it not as straightforward as you'd like to think.

The previous owner had brought this car to 4 garages but I don't have the receipts. I noticed that somebody had messed with the front O2 and right side fuel injectors. In fact, the car was missing the metal fuel rail cover on the right side, and the engine cover was in the trunk. She said the previous garages thought she needed a new engine, or that she had a wiring harness issue. I think they are wrong, though.

Dispatch20 is offline  
Reply

Quick Reply
Message:
Options

Register Now



In order to be able to post messages on the Subaru Outback - Subaru Outback Forums forums, you must first register.
Please enter your desired user name, your email address and other required details in the form below.

User Name:
Password
Please enter a password for your user account. Note that passwords are case-sensitive.

Password:


Confirm Password:
Email Address
Please enter a valid email address for yourself.

Email Address:
OR

Log-in










Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page
Rate This Thread
Rate This Thread:



Posting Rules  
You may post new threads
You may post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On

 
For the best viewing experience please update your browser to Google Chrome