Cold Misfire Subaru

Ronald Technical Support Specialist Trail, British Columbia Posted   Latest  
2005 Subaru Outback R 3.0L (EZ30D) 5-spd (TG5C7)
P0305 - Cylinder 5 Misfire Detected
Rough Idle

Starting at 250 000 km (156 000 mi) this vehicle would periodically set a P0305 misfire code. Freeze frame suggested to us the code set in the engine warm up process. Low speed (40 kmh (25 mph)), low load, 30 to 40 degrees Celsius coolant temp is a synopsis of the code set conditions.

The customer complaint was only the check engine light coming on. In our early testing we never experienced the misfire. Our initial diagnostic process was then switchtronics. We switched coils and plugs and then injectors. After each change the vehicle would be back with the customer for several days before it came back still setting the same P0305.

The possibility of a cold engine vacuum leak inspired us to change the intake manifold gaskets though no combination of vacuum leak checking revealed a problem. While the intake manifold was off we got to check for carbon build up on the intake valves. All looked good. Needless to say the misfire code P0305 came back.

The vehicle now has 254 000 km (159 000 mi) and a cold rough idle is now apparent on all cold starts and warmups. It won’t always set the code but the misfire is definitely present until warmed up.

Mechanical gauge compression and leak down testing was performed on cylinders 3 and 5. Leak down was less than 3% and compression was in spec. 

At this point I wanted to use my pressure transducer to do some tests and see what I could find. Cold cranking did seem to have a strange cadence so I started there.



More than one cylinder showed a lower cranking amp value than just cylinder 5.

This test was followed by a look at the intake manifold pressure waveform which shows one weak intake draw.



Next an exhaust pulse pressure test was performed. There were a couple anomalies on this waveform. I am not sure what to make of them.



A cranking compression test was performed on cylinder 5 (cylinder flagged for misfire)



Cylinder 1 cranking compression was performed to compare with cylinder 5.



These tests were then followed with running cylinder pressure tests from cylinder 5.



This is cylinder 1 running pressure.



I am not sure how to interpret all this data. I don’t see a clear problem. I am looking for advice and any further testing. 

I am also including scan data 

Cold rough idle.



Warm smooth idle.



Road test.



Thanks for any insight etc.

Randy Diagnostician
Glendora, California
Randy Default

Did you cheek for coolant intrusion

+1 Default Ð Bounty Awarded
Scott Technician
Chico, California
Scott Default

I have in the past had problem vehicles that would randomly do this. Most recent was a v6 engine. I was almost never able to reproduce a misfire. Customer had for a week before it would reset. Ended up being a leaking head gasket. Took it for a long hard drive right before closing up. Pulled into shop and pressurised cooling system to maximum pressure and let sit overnight. Next day pulled plug…

+4 Default Ð Bounty Awarded
Dmitriy Analyst
Toronto, Ontario
Dmitriy Default

Hi Ronald, 126 versus 140 psi in cranking compression peaks is quite a difference, something is causing that. It would be great to see coolant pressure pulse and crankcase pressure pulse waveforms added to your collection — may be they will reveal the issue?

+1 Default Ð Bounty Awarded
Paul Mobile Technician
Sault, Ontario
Paul Default

Hi Ron, awesome diag. Cylinder 6 has a weaker induction pull because it's companion #5 had weak power stroke. From the Running in-cylinder files it appears the Intake Valve is opening a bit too early for Cyl # 5 by about 17 degrees when compared to Cyl 1. So if the misfire goes away on HOT soak, it would not make sense that there is an issue on the entire bank 1, but a sticky intake valves are a…

0 Default Ð Bounty Awarded
Paul Mechanic
Harvey, Illinois
Paul Default

Looks like you have lower compression in cyl 5 compared to 1. I would capture more data on running compression on cyl 5, like a minimum 10 second total screen time to watch for varying compression-you can always zoom in. Usually indicates valve sealing issue if you get a roller coaster pattern when you collectively look at the compression peaks. The fact that it only happens cold though, could…

0 Default Ð Bounty Awarded
Leo Technician
Rhinebeck, New York
Leo Default

Nice case so far! I’ll just share a tidbit that I’ve seen on these engines quite a few times: diag​.​net/file/ftfdauq89… ​ Within the red dye you’ll see fine cracks. In these cases, the misfires would only occur during cold start, 5 seconds or so (coolant leaking into cylinder overnight). How about a pulse sensor on the radiator? Wouldn’t hurt to do a pulse reading on the crankcase…

+4 Default Ð Bounty Awarded
Albin Diagnostician
Leavenworth, Washington
Albin Default

Your scope captures show a compression loss in several cylinders, the question is, "where is the compression going"? Stick your FLS in the crank case dipstick and give the engine a crank. I didn't see anything sticking out in your intake & exhaust pressure waveforms that would lead me to a leaking valve. The compression is either going past the rings, or out into the cooling system.

0 Default Ð Bounty Awarded
Gil Technician
Fresno, California
Gil Default

Nice data Ronald, your running compression exhaust pocket and intake pockets should be almost the same , but neither cylinder is , And don’t think coolant in cylinder is going to lower compression, I know some Subaru engines had exhaust seat issues, check the cranking exhaust pocket, should be less than 3 psi, compare it the good cylinder, cold and warm, I cant open waveform at this time

0 Default Ð Bounty Awarded
Victor Owner/Technician
Doral, Florida
Victor Default

I dealt with a similar problem just last week​.​The cure was a bottle of head gasket sealant.(temporary fix)

0 Default Ð Bounty Awarded
Hollis Technician
Boulder, Colorado
Hollis Default

Excellent work so far, Thank You ! You're almost there. As Leo G. mentioned, get a pulse capture from the cooling system while the misfire is present to hopefully seal the deal. Some typical symptoms that the customer may have experienced and forgotten prior to the Check Engine Light illuminating, are an occasional or intermittent loss of cabin heat or a short period of engine overheating…

+1 Default Ð Bounty Awarded
Niall Owner/Technician
Stirling, United Kingdom
Niall Default

If (big if) we carry out an LPG conversion on these engines its a must to fit valve lubrication, they still get valve regression I always assume that wear will also occur on petrol but more slowly.

0 Default Ð Bounty Awarded
Ronald Technical Support Specialist
Trail, British Columbia
Ronald Update

Took me a while to get back to this job. As many suggested I performed an idle crankcase pressure test and a cooling system pressure test at idle as well. I am not sure what to make of my results. The waveforms do not resemble the pico examples. I am not sure if I had something wrong but I don't think there is any significant leakage in both cases. I am thinking it is time to check valve…

0 Update Ð Bounty Awarded
Albin Diagnostician
Leavenworth, Washington
Albin Default

When using a Pico to capture this type of waveform, you need to use a long time base, such at 10 MS per division, you can always zoom in on your capture. When your timebase is too short, there is no way you can scrunch up your waveform to make it usable. Now, if you are using a snapOn scope, you need to use the short timebase like you have used. Also, to make your waveform mean more it is always…

+1 Default Ð Bounty Awarded
Gil Technician
Fresno, California
Gil Default

Hey Ronald, I’m not sure that the valve adjustment is going to be the cure cause I’ve never seen tight valves get loose as it warms up, I know the wps500 pressure transducer is awesome, but the first look transducer (or a similar transducer) is even more sensitive In the intake or exhaust when performing a relative compression test, or even the in cylinder transducer in the suspect cylinder and…

0 Default Ð Bounty Awarded
Ronald Technical Support Specialist
Trail, British Columbia
Ronald Delay

Problem is not solved yet. Gave vehicle back to customer. Problem is most definitely mechanical but not bad enough yet. Testing procedures to date have not produced a definitive answer. When I get vehicle back I will try to improve some of my testing procedures as suggested.

0 Delay Ð Bounty Awarded