Subaru FB25 - misfires and O2 sensor?

[D60]

I'm fighting an '11 Forester normally aspirated that had a no-start/very hard start. Now it's starting relatively easily but idle sucks sometimes and it's recording misfires in 1 and 3. Common problem on all the Subaru engines from what I've seen.

There's often talk of the O2 sensor causing misfires but how would this affect cold start? As I understand it the ECM isn't looking at HEGO for starting purposes, at least cold?

I'm only curious about O2 because if I unplug the upstream no more misfires are recorded, but I'm not sure it actually runs or idles much better. Will the ECM even try to record misfire counts if the O2 is disconn'ed?

Basically I'm wondering if I'm creating a red herring by looking at misfire counts with and without O2 sensor?

I've gone through a lot of other troubleshooting but in the interest of brevity (too late!) I'd rather not get into that here. Just wondering if it's significant that misfire counts stop with O2 unplugged?

 

[Trav]

On initial cold start it is in open loop and as you said should not looking at the O2 input, I have no idea if the ecm continues to monitor misfires with the O2 unplugged. With heated O2 they do not stay in open loop very long and unplugging the O2 will force into closed loop so I would guess it does not monitor misfire, just a guess.
Have you looked at fuel trims and when the O2 starts switching? On any Subaru check the grounds, the ones from the cylinder heads to frame are very thin and prone to breaking.

 

[twX]

The O2 sensor will affect AF learning, which will cause fuel corrections even on cold start before the O2 sensor is active. If the problem goes away on the first cold start after an ECU reset, maybe the O2 sensor is the cause.

On a cold start, the fuelling is mainly controlled by the MAF sensor. If you have an ODB2 scanner, check the MAF reading at warm idle. It should be reading ~2.5 g/s.

Subaru has a TSB for the FA engines for carbon deposits causing rough idle or misfire codes. It doesn't apply to your model, but yours is still direct injected, right? I'd run a bottle of Subaru PEA cleaner through the fuel tank to clean the fuel injectors.

Cylinders 1 and 3 are on the same bank. If there are misfires on these cylinders, but never any on 2 and 4, I'd lean more towards something that affects only those cylinders, like the variable valve timing, though I'm not sure if improper valve timing would actually cause misfires.

 

[CarbonToast]

Check your fuel injectors.

I fixed a outback once that had a misfire due to a bad fuel injector. I was able to confirm it was dead by unplugging it and checking the resistance and comparing it to a good cylinder. 1 injector later it was back on the road.

 

[meep]

I had an ej22 with misfires at idle, ended up being a stretched timing belt. Belt had 30k on it but was aftermarket. I installed an oem belt and problem was solved. I did NOT know it was the TB… mechanic suggested wiring harness and I said “no thank you.” Found out by replacing the water pump and noticing the belt stretch.

 

[D60]

I had an ej22 with misfires at idle, ended up being a stretched timing belt. Belt had 30k on it but was aftermarket. I installed an oem belt and problem was solved. I did NOT know it was the TB… mechanic suggested wiring harness and I said “no thank you.” Found out by replacing the water pump and noticing the belt stretch.

Yeah this vehicle has only 88k on it. I did find a short vid where he claimed to have found the belt jumped by 3-4 teeth, but it was an EJ and I don't think this is all that common?

 

[D60]

The O2 sensor will affect AF learning, which will cause fuel corrections even on cold start before the O2 sensor is active. If the problem goes away on the first cold start after an ECU reset, maybe the O2 sensor is the cause.

On a cold start, the fuelling is mainly controlled by the MAF sensor. If you have an ODB2 scanner, check the MAF reading at warm idle. It should be reading ~2.5 g/s.

Subaru has a TSB for the FA engines for carbon deposits causing rough idle or misfire codes. It doesn't apply to your model, but yours is still direct injected, right? I'd run a bottle of Subaru PEA cleaner through the fuel tank to clean the fuel injectors.

Cylinders 1 and 3 are on the same bank. If there are misfires on these cylinders, but never any on 2 and 4, I'd lean more towards something that affects only those cylinders, like the variable valve timing, though I'm not sure if improper valve timing would actually cause misfires.

Yeah it's flowing right around 2.5g/s at hot idle.

I get occasional misfires on 4 but thus far none on 2. I found another thread somewhere in cyberland with these EXACT symptoms, but every thread on the Subaru misfire debacle always just trails off with no posted resolution.

I understand the bug eyes had a TSB for a software update to essentially "desensitize" the monitors, but that was well before the '11 FB25

For giggles I swapped coils 2 and 3, but no change. Plugs look beautiful and are NGK ‐- I suspect they were done recently as the current owner has only had it for ~18 mos

I thoroughly cleaned EGR valve and MAF and removed TB and cleaned butterfly (dirty because this is the typical 1qt per 1000 oil drinker)

Owner reports he was near empty, filled up at local indy station and it wouldn't start the next morning. He wanted to blame "bad gas", although I use the same station almost exclusively so I'm skeptical.

I did dump a bottle of the red HEET in the tank which also allegedly has an injector cleaner, although I only chose it because it's what I had on hand from a clearance sale somewhere sometime.

Did the HEET help? Eh, I don't think so. The problem now is that the problem is so vague and somewhat inconsistent. I hate these because you can never conclusively say if anything you did made a difference.

 

[D60]

Also when it's acting up and idling stupid there's a slight delay in throttle response just sitting idling. It's a fraction of a second and a lot of drivers may not notice.

I'm not sure it means anything, but it's one way I can confirm if it's in stupid mode or good behavior mode. When in good behavior mode the response to tipping into the throttle (at idle in P) is smooth and immediate.

I tried looking at throttle percentages in live data and that all seems fine -- it increases immediately and linearly (is that a word?) even in stupid mode, so whatever comprises the TPS in TBW (throttle by wire) seems to be working.

 

[meep]

IIRC the misfire on mine was primarily on 4 which is why he suggested a wiring harness, and maybe a smaller number on 1.

 

[Trav]

Any codes except the misfire? Fuel trims?

 

[twX]

I get occasional misfires on 4 but thus far none on 2.

In that case, I'd say whatever is going on probably relates to all cylinders, even if the one bank is more prone to misfiring. So, maybe not related to injectors, plugs, coils, or AVCS.

Owner reports he was near empty, filled up at local indy station and it wouldn't start the next morning. He wanted to blame "bad gas", although I use the same station almost exclusively so I'm skeptical.

Maybe sediment at the bottom of the fuel tank got sucked into the pump when the fuel level was really low and clogged up the fuel filter? See if you can monitor fuel pressure and targeted fuel pressure. A bad LPFP or HPFP could also cause low fuel pressure and misfires.

If the issue is intermittent, any data monitoring you do while the engine is running well may not be representative, so try to read the MAF, fuel pressure, fuel trims, etc, while the problem is occurring. Maybe look at AVCS readings too just in case. Check coolant temperature too to make sure the sensor is accurate, since the enrichment on cold start is based on coolant temp.

I had an issue on my FA20 where the MAF reading spiked to weird values a few times while cruising in really cold weather. Disconnected and reconnected the MAF sensor connector, and it hasn't happened since. Maybe check the connectors for the MAF and O2, and clean them with some contact cleaner.

 

[tiger862]

Go to site like Aldata and pay 20.00 for one month of information including wiring schematics and technical service bulletins. It is a good resource for you to understand your particular vehicle as well as what other techs have found for same problem. I just got through with an 012 Impala that kicked my butt until I got right information. It was misfire on cylinder 4 . Need a good scanner to do a balance test and compression, otherwise you're guessing.

 

[D60]

Any codes except the misfire? Fuel trims?

Inconclusive only due to drive time. Owner said he had a friend pull codes before it was towed to me and only P0171. I don't know how long that had been there, though.

I smoked it like crazy last night and can't find a vac leak.

Once I had pending codes for 1 and 3 misfires. It hasn't actually set hard codes for me yet -- and in the absence of that I went looking at the misfire counters.

At cold start STFT is pegged at 35 (is this normal?). But once warm it tends to hover around +2

 

[D60]

In that case, I'd say whatever is going on probably relates to all cylinders, even if the one bank is more prone to misfiring. So, maybe not related to injectors, plugs, coils, or AVCS.


Maybe sediment at the bottom of the fuel tank got sucked into the pump when the fuel level was really low and clogged up the fuel filter? See if you can monitor fuel pressure and targeted fuel pressure. A bad LPFP or HPFP could also cause low fuel pressure and misfires.

If the issue is intermittent, any data monitoring you do while the engine is running well may not be representative, so try to read the MAF, fuel pressure, fuel trims, etc, while the problem is occurring. Maybe look at AVCS readings too just in case. Check coolant temperature too to make sure the sensor is accurate, since the enrichment on cold start is based on coolant temp.

I had an issue on my FA20 where the MAF reading spiked to weird values a few times while cruising in really cold weather. Disconnected and reconnected the MAF sensor connector, and it hasn't happened since. Maybe check the connectors for the MAF and O2, and clean them with some contact cleaner.

FP sits around 50 all the time.

I will say if idle stumbles significantly (which isn't very common) the needle on my gauge jumps around in both directions a lot, but I don't think this is abnormal.

I don't think '11 got a standalone fuel filter ‐- I think there's only one in the tank. I may be mistaken?

 

[2EHA]

HEGO?

that’s a planar uego it’s closed loop in 10 seconds

 

[D60]

My next thing is to look at actual voltages and graph the O2 switching ‐- although this still doesn't seem relevant to cold start issues.

My little Xtool doesn't graph but I have an Autel that will. I looked at O2 current when warm and I swear my Xtool said it's giving me millivolts. But it was late and now I'm questioning the scale because I was getting readings as low as an occasional 5 up to 126. So, I was either seeing a max of .126V or 1.26, which seems wonky either way.

Based upon trying to track the numbers in my head I didn't feel like it was switching hi/low very consistently, but that's a losing game for my brain. I need to graph it.

 

[D60]

Ok I do apologize -- as upstream is an AFR the values I was seeing may be good. Need to dig around further tomorrow.

Idling in the shop for about 15-20 min the vehicle did finally set a P0303 but that's it. LTFT was showing 9.4 but STFT was 0.8. No time to drive it tonight

 

[Trav]

It sounds like coil or plug on #3. A sticking, leaking or clogged injector would have much different numbers. These are old school Denso injectors, they are very reliable and rarely have any issues. No DI or HPFP on this one, it could even be defective plug or plug that is failing when it gets hot. Do you have a coil tester? I have had this one for some years now and it does a good job diagnosing coils.

https://www.amazon.com/General-Technologies-Corp-GTC505-Ignition/dp/B001TD5JYE/ref=asc_df_B001TD5JYE/?tag=hyprod-20&linkCode=df0&hvadid=312065537890&hvpos=&hvnetw=g&hvrand=14465413733628912825&hvpone=&hvptwo=&hvqmt=&hvdev=c&hvdvcmdl=&hvlocint=&hvlocphy=9001771&hvtargid=pla-569205885826&psc=1&mcid=0142ac602bfb365da0dbdbf91d4bb4aa&gclid=EAIaIQobChMIiNee-ayPhAMVu19HAR3u-w9ZEAQYASABEgJVRfD_BwE

 

[twX]

I will say if idle stumbles significantly (which isn't very common) the needle on my gauge jumps around in both directions a lot, but I don't think this is abnormal.

Sounds suspicious to me, especially since the issue popped up after the owner ran it low on fuel and refilled the tank. It could just be that the revs drop low enough when it stumbles for the alternator voltage to drop, affecting the pump and fuel pressure, but it could be the fuel pressure causing the misfires.

I'd monitor the alternator voltage, and make sure the battery posts have a good connection. Low voltage could affect both the fuel pump and the required injector pulse width. If that all checks out, I'd investigate the pump itself and the fuel filter.

 

[wtd]

Does the compression seem fine? You may need to scope the fuel injector and ignition system on the cylinder that is having the misfire so you can see in real time if any of those things are dropping out.