Subaru Outback Forums banner

2381 - 2400 of 2418 Posts

·
Registered
Joined
·
19 Posts
When it arrives, I'll use the borescope to confirm number and condition of cats.

Meanwhile I'm continuing to drive on premium (and combustion cleaner in the tank). P0420 hasn't returned yet. Which is weird given that the logs show the cat is doing nothing.
Ok, the latest:
  • the $20 ebay borescope arrived and I pulled out the oxygen sensors to have a look
  • found some wet oil on the oil sump which looks like it is coming from the gasket between the oil sump and the main engine block. Looks like the garage didn't seal it up properly after they fixed the leaking head gasket. Seems to be some sort of liquid gasket - grey goo of some kind. So hopefully the oil consumption mystery is solved and can be fixed fairly easily.
  • the borescope confirmed there is a second cat that doesn't have an O2 sensor downstream of it. No idea why Subaru would do that.
  • I've put two videos in the Google Drive (Public Access - Google Drive)
  • the resolution isn't great, but the upstream face of the cats look fine. No holes, no cracks, no sign of melting, they appear to be white in colour. There does seem to be some small particles on the substrate here and there, but I don't think that means anything?
  • Last two tanks of fuel were premium with subaru combustion cleaner added via the tank. Filled up with standard (91 octane) and the P0420 code came back while driving at 100 kph on the highway about a half hour after filling up. Coincidence?
Was hoping the borescope would give me definitive evidence. Sadly, not the case.

So, next step is the new aftermarket cat from the UK. I'm interested to hear if anyone sees anything on the videos that would change that plan...?

Paking89 - pdf page 775 of the fsm for my 2005 outback says the DTC will light up after 2 consecutive drive cycles that fail, and the light will go off after 3 consecutive drive cycles that pass. The prior page sets out the conditions for when a drive cycle occurs - they are pretty specific (set values for battery voltage, atmospheric pressure, speed, intake air, coolant temp etc etc) so it seems like it is hard to know when the test is performed. I'm no expert but it sounds like you've done everything and the only thing left to do is to replace the cat. If you want to be sure, have you tried doing some romraider logging? While I prefer to use OEM I can't justify it in my case - the cost of an OEM cat in Oz is almost the same as the value of my car. So, it might be worth considering non-genuine aftermarket ones? At least they are a lot cheaper in the US than anywhere else in the world!
 

·
Registered
Joined
·
19 Posts
Seeing as the leading edge of the cat looked fine on the borescope, I was wondering whether an exhaust leak might be affecting the cat....?

There is often a hint of an exhaust smell when driving with the front windows down. Had a look underneath - the join upstream of the cat has some white residue encrusted on it and the flanges are not parallel, which could be signs of a leak? See photos. I did hook up a vacuum cleaner to the tail pipe, but I couldn't hear any air being sucked in through this joint. But maybe the leak is slight and only happens under higher pressures and/or when it gets hot..?

Be great to hear from those more experienced than I am whether it looks like a leak....?
 

Attachments

·
Premium Member
2001 VDC/SC One of a Kind
Joined
·
14,318 Posts
Discussion Starter #2,383
I have a 2008 outback 2.5 base non California emissions. Check engine light has been on for 11 months now with the same p0420 code. I've read all the way through a bunch of forums on here including this one but I still haven't found what I'm looking for. So starting last November, I have replaced the a/f sensor, oxygen sensor, pcv valve, egr valve, driver side valve cover gasket with spark plug tube seals, brand new spark plugs, brand new ignition coil pack because one of the terminals fried, all 4 brand new NGK spark plug wires, brand new NGK Iridium IX spark plugs, cleaned the throttle plate and inside throttle body, I've run seafoam through the crank case and through the gas tank. The whole time I've been using a cheap code reader to monitor the i/m readiness monitors. Each time I put more new parts on more and more monitors have started to clear. Just 2 days ago for the very first time every single monitor said OK so they had completed the drive cycles. So I parked the car and let it sit for 20 minutes with the engine off. Turned it back on and the check engine light was still on. So every other time I do the driving cycle, the EGR monitor never completes, it always stays incomplete. I'm very sure my fuel injectors are working properly. Haven't had or noticed any problems with the fuel system. It was running rich because one the spark plug wires fried itself and the coil pack but after replacing them it seems to have fixed the fuel problem and leaned it back out. Also I have only used OEM parts from a Subaru dealer, no aftermarket parts at all. My question is, does anyone know how many driving cycles it takes consecutively to clear the EGR monitor? I haven't been able to get it passed for an emissions inspection, I live in Texas and it keeps failing inspection because of this **** code. I don't want to have to spend another $2,000-3,000 for a catalytic co verter if I can avoid it, but also I'm not sure that's the problem. Any help would be very greatly appreciated.
If you're in the Austin area, call me. Sent PM with number.

I recently had a 2002 Forester 2.5X with 208k miles and kept throwing a P0420 code. The shop the owner went to prior to coming to me had put in new plugs and the car started running better but still threw the code. The rear O2 sensor heater circuit was weak and the feedback to the ECM sucked.

As for the EGR monitor, it can take 20-150 miles depending on how you drive. In all the emission testing counties in Texas you should pass with one incomplete monitor, so if all the others are set and just the EGR is incomplete then it should pass as long as there are no codes set.

To get the EGR to set you would drive on the highway about 55-70 mph for a few minutes then let off the throttle and coast down for a bit. Then accelerate with 1/2 throttle back up to 55-70 and cruise a few more minutes. The ECM looks for changes in FT and air flow/vacuum with the EGR on and then off. You have to hit it just right for the ECM to run it's test.
 

·
Registered
2006 Subaru Outback 2.5i Limited wagon
Joined
·
110 Posts
My P0420 has been coming and going for awhile. Took it in and mechanic said to try Cat'a'Clean; that may have mostly fixed it for awhile, but it came back pretty strong during a recent roadtrip (lots of high-speed freeway driving); so I tried Cat'a'Clean again, with little luck. FWIW, I replaced rear O2 sensor a couple years ago, before this problem started (I forget what the code was, but it pointed directly to the O2 sensor).

I've gotten TorquePro working, so I'd like to generate some graphs to see what can be figured out. Main questions: what variables do I need for TorquePro to graph, and for how many minutes (starting from a cold start ?). Probably buried in these 120 pages, but ...
 

·
Premium Member
2001 VDC/SC One of a Kind
Joined
·
14,318 Posts
Discussion Starter #2,386
My P0420 has been coming and going for awhile. Took it in and mechanic said to try Cat'a'Clean; that may have mostly fixed it for awhile, but it came back pretty strong during a recent roadtrip (lots of high-speed freeway driving); so I tried Cat'a'Clean again, with little luck. FWIW, I replaced rear O2 sensor a couple years ago, before this problem started (I forget what the code was, but it pointed directly to the O2 sensor).

I've gotten TorquePro working, so I'd like to generate some graphs to see what can be figured out. Main questions: what variables do I need for TorquePro to graph, and for how many minutes (starting from a cold start ?). Probably buried in these 120 pages, but ...
Yeah, long thread with lots of info and data.

Cold start, you want to look at AF Correction (STFT), Rear O2 and ECT (should get to 190-200 and fluctuate in this range). The AF Correction won't start up until the system goes to closed loop fueling. This can be verified with the AF sensor resistance which should drop within less than a minute to a 2 digit value.

If the rear O2 is oscillating up and down with fuel correction at idle, you would then drive it down the road at a speed of 50 mph or greater for a few minutes and watch the rear O2 value with sustained speed and throttle. If it continues to oscillate the cat is bad. If not, then the rear O2 may not be heating up to operate properly.

To determine if the engine is performing as it should, you would look at the STFT, LTFT, Knock Correction, ECT, O2 and battery voltage to make sure the alternator isn't running oddball.

Also check to make sure the battery and cables are good. No corrosion, good conductance and the battery supplying sufficient ampere output.
 

·
Registered
2006 Subaru Outback 2.5i Limited wagon
Joined
·
110 Posts
Cold start, you want to look at AF Correction (STFT), Rear O2 and ECT (should get to 190-200 and fluctuate in this range). The AF Correction won't start up until the system goes to closed loop fueling. This can be verified with the AF sensor resistance which should drop within less than a minute to a 2 digit value.

To determine if the engine is performing as it should, you would look at the STFT, LTFT, Knock Correction, ECT, O2 and battery voltage to make sure the alternator isn't running oddball.
In TorquePro, the definitions of variables don't quite jibe with the ones you've mentioned (I don't have a Windows machine so can't use RomRaider). So I'm not sure how to get the variables you suggest. I believe I selected every one that might apply, and here's what it gave me values for (the others were just blank in the dot-csv file):
  • air-fuel ratio (measured)
  • ECT
  • fuel trim bank 1 sensor 2 (%) - there's a bunch of other fuel trim variables, but only this one logged any values, I'll enable "bank 1 long-term" and "bank 1 short-term" next time, to give me what you call STFT and LTFT
  • mass airflow rate
  • O2 sensor1 equivalence ratio
  • O2 sensor1 equivalence ratio (alternate)
  • O2 sensor1 wide-range voltage
  • O2 volts bank1 sensor2
There's a whole slew of other O2-related variables, and I selected 'em all, but only these gave any values. So, when you say "rear O2", which of these O2 variables should I be looking at ? FWIW, the "O2 equivalence ratios" tend to hover around 1 plus/minus a few hundreths, but occasionally spike up to 1.38. "Wide-range voltage" is usually 2-point-something but occasionally spikes to 5+ (when equivalence ratio goes up to 1.38, seems to happen when I decelerate, maybe with downshift, should add RPM to list). "O2 volts" fluctuates wildly between 0 and 1; I'm guessing THAT's the one you're talking about, right ?

I don't see anything that looks like knock correction. Nor battery voltage oddly enough, but there's voltage of control module so I guess that will have to do. I don't see anything that looks like "AF sensor resistance", either. There's also something called "Horizontal Dilution of Precision" whatever the heck that is.

So if you could give me some guidance on how to get the right stuff logged, I'd be grateful, thanks.
 

·
Premium Member
2001 VDC/SC One of a Kind
Joined
·
14,318 Posts
Discussion Starter #2,388
@RustyShackleford
Sensor location is correspondent to bank number. NON CARB cars with a single header assembly and one cat in the manifold will always be referred to as sensor 1 since it monitors both banks simultaneously.

Sensor 1/1 is the front AF Sensor, prior to the cat. Sensor 1/2 is the second sensor, rear O2, behind the cat. This is the monitoring sensor to let the ECM know if the cat is functioning.

O2 sensor 1 equivalency ration would be the front AF sensor fuel trim feedback calculated by the ECM in LAMBDA. 1 Lambda is equivalent to stoichiometric value 14.7:1. Higher than 1 is lean, lower than 1 is rich. The value, along with the wide range voltage go high off throttle because the ECM cuts fuel when you decelerate in gear and momentarily during shifts. (Is your car a manual trans?). If your equivalency ratio is staying around 1 cruising, it will fluctuate some as this is how the ECM controls cat temperature, then the fueling is good as long as the STFT is not =/- 5% majority of the time. Negative STFT means the ECM is reducing fuel, positive means it's adding. If it's reducing, there's a problem with combustion and the engine isn't burning the fuel properly. If it's positive then the ECM is adding fuel and there's a fuel delivery problem or a vacuum leak.

Your 06 should have just a single manifold.
It will be an EJ253 with AVLS which controls the cam phasing and a MAF sensor for air volume.
Battery voltage and good conductance through both positive and negative cables is important, as well as the oil level. The AVLS system uses a high volume of oil to operate the valve lift system.

Timing belt alignment is also a factor in a proper running engine, so that's something to look at if all else checks out good; i.e. no vacuum leaks, sensors working, fuel flow is good, etc.
 

·
Registered
2006 Subaru Outback 2.5i Limited wagon
Joined
·
110 Posts
Yes, it's an EJ253 5-speed manual tranny. FWIW, the car seems to be running great.

So in post #2386 above where you talk about "rear O2", which of the variables that TorquePro seems to be allowing me to log is this referring to ? I describe what the various curves look like in post #2387. I guess it's not the equivalence ratio that you describe.

How do I test the battery cables' conductivity ? Try to measure the voltage between starter motor '+' terminal and engine block, while cranking ?
 

·
Registered
Joined
·
18,172 Posts
So in post #2386 above where you talk about "rear O2", which of the variables that TorquePro seems to be allowing me to log is this referring to ?
I'll help it along . . . O2 volts bank1 sensor2

How do I test the battery cables' conductivity ?
There are some battery conductance testers that can also be used to test for poor conductance in wiring related to the battery (e.g., Midtronics, and even Harbor Freight has one), but a multimeter can be used.

Basically, when current is flowing through a wire or a metal body, there will be some voltage drop between the ends. This is due to the natural electrical resistance of the material. So, some voltage is normal; but too much is not.

For the grounds, engine running, turn on electrical loads such as headlight and running lights, heater fan, perhaps the rear defogger or a seat heater. Raise the engine rpm to around 1000, then measure the voltage between the battery negative post (not the cable clamp) and the alternator case, engine block, transmission bell housing, and a good ground on the body. Each of these are, ideally, less than 0.1 V (100 mv) but 0.2 V isn't terrible. Higher than that suggests that ground connections or the wires themselves are not great.

Similarly, measure the voltage between the battery positive post and the large positive output post on the alternator.

Try to measure the voltage between starter motor '+' terminal and engine block, while cranking ?
To test the cable between the battery positive post an the starter, use the same approach as above, i.e., measure between the battary positive post and the starter motor + terminal. (I would imagine that would be the terminal on the starter solenoid, as the motor + terminal is difficult to access, especially while there' voltage there and the engine is being cranked.) That way, it's looking at the voltage drop in the cable when the heavy starter current is flowing. The voltage could be somewhat higher than the other tests (above) due to the high starter current. However, that cable is used only for starting; a bad positive cable to the starter will affect starting, but not engine, transmission etc., operation once the engine is running. I would pass on that for the time being if starting is fine.
 

·
Registered
2006 Subaru Outback 2.5i Limited wagon
Joined
·
110 Posts
I'll help it along . . . O2 volts bank1 sensor2
Thanks, that's what I was afraid of. It seems to oscillate rapidly (more than once per second) between about 100-900mv.

Seems to point to bad cat, but this was only about a 10min run; I need to do it again, heading out to the freeway as suggested. The weird thing is, I believe (it's been so intermittent that it's hard to say with much certainty) that this is exactly when the MIL tends to come on - after going down the freeway for awhile. It is possible the ECM waits for the sensor to heat up before it decides the cat is bad ?

If "rear O2" continues to oscillate on the freeway, does this definitely point to a bad cat, and rule out the other things (such as battery cables, PCV, etc) that cardoc mentioned way back on page 1as the primary culprit ?
 

·
Premium Member
2001 VDC/SC One of a Kind
Joined
·
14,318 Posts
Discussion Starter #2,392
@RustyShackleford
Oscillation frequency such as you describe is indicative of a bad cat. If it were occasional oscillation then it would point to a problem with engine performance or electrical flow.

The caveat is that the cat doesn't wear out. It is affected greatly over time by the several factors listed at the beginning of this thread.

Type and grade of the gasoline you use over time causes high amounts of carbon to build up in the combustion chamber. This carbon reduces the available space in the combustion chamber and affects air flow. After the carbon is heated during combustion it retains heat and causes detonation/knock which is an explosion in the chamber caused by preignition of the fuel. When this occurs the ECM adjusts the fuel and timing to compensate and stop the detonation. In doing this it contributes to more carbon build up in the engine and exhaust because most times the fueling is run rich to stop the knock. High octane gas resist detonation and assist greatly in keeping the engine running clean and efficient. Less knock means less fuel trim adjustment. Running clean also improves your gas mileage.

Stay clear of Murphy gas stations. Worse gas available for any car. Except maybe a 1976 Pinto.

Carbon build up on valves affect air flow in to the combustion chamber. Since this is a port injection engine, the ability to atomize the fuel during the intake stroke is greatly affected and the swirling of the air becomes limited. This creates a rich condition and the fuel doesn't burn efficiently which creates a rich exhaust mixture and the ECM compensates by reducing fuel. Another effect is it can create misfires. You may not notice a rich fuel mix because rich fueling creates more power. This engine will also run leaned out and provide power. Old school hot rods run prime power outputs at 10:1, sometimes lower toward 9:1. (My VDC puts out more power leaned out under boost than enriched, but that would be detrimental to the engine.)

This is where the STFT values help you determine performance issues.

Clogged PCV valves allow for build up of gases created by the heating of the engine oil and pressure in the crankcase builds up. This also contributes to higher oil temperatures and increased sludge/carbon build up in the oil journals and decking. Over time the journals are restricted limiting oil flow. The PCV system also affects how the engine breaths during combustion. Clogged PCV valves will also create a higher amount of carbon within the intake manifold which effects air flow. The pressure will eventually find a way out and with increased crankcase pressure you get increased oil particulates in the air stream.

On the battery and cabling; the ampere output of the battery is the most important and conductance through the grounds is at the top of the list followed by the positive cable from the alternator to the fuse box and battery to fuse box. Lack of electrical flow prevents proper operation of all the electrical appliances on the car. All the engine management suffers. Inputs and outputs work slow or not at all. In cars with automatic transmissions the transmission will burn up clutches due to slow reacting shift solenoids. The more appliances running, lights, AC, stereo, the greater the probability for engine performance to drop when the electrical sources are weak. The AC draws a lot of current and when you run it with a weak battery or cabling the sensors and computer will run slow. Like your phone with a low battery. Things tend to shut down. Operation of all the actuators is reliant on the grounds. The ECM controls ground flow to operate the injectors, relays, solenoids, etc.

Your car can seemingly run great and still have issues, or have things that are deteriorating. Something damaged the cat and prior to replacing the cat I suggest you insure the engine is performing proper. It may be that the engine is running proper now, without data I can't tell you for sure.

If Torque Pro can't data log over time and create a file that you can post for us to see, look in to Blue Driver. It will create a csv file that you can share.
 

·
Registered
Joined
·
18,172 Posts
The weird thing is, I believe (it's been so intermittent that it's hard to say with much certainty) that this is exactly when the MIL tends to come on - after going down the freeway for awhile.
Normal. The ECM runs a test for the cat converter once per drive cycle when preconditions are met, one of which is the car is being driven steadily at 44 mph or more, for a period of 30 - 55 seconds. Also, the fault has to be detected in two consecutive drive cycles before the code is set and the CEL is turned on.

If "rear O2" continues to oscillate on the freeway, does this definitely point to a bad cat,
In a few cases here, constant oscillating seemed to be due to a faulty rear O2 sensor, or at least replacing the sensor cleared that symptom. But as @cardoc mentioned above, there's a wide range of factors that can interplay; replacing the sensor in those cases might not have addressed the core issue(s) longer term.
 

·
Registered
2006 Subaru Outback 2.5i Limited wagon
Joined
·
110 Posts
Ok, thanks for the patient explanation, esp. you @cardoc. I'm going to do another run, including some freeway driving and adding STFT and LTFT to the logged variables in TorquePro, and post some graphs (I assume you mean sharing graphs, not actual csv files).
 

·
Registered
Joined
·
18,172 Posts
(I assume you mean sharing graphs, not actual csv files).
I think, preferably, the actual file. That allows others to examine parameters in different combinations that might not have been included in another's graphs.

That said, unfortunately, .csv files can't be attached to posts here, but they can be uploaded to a web storage site, and a link posted here so that others can download it.
 

·
Registered
2006 Subaru Outback 2.5i Limited wagon
Joined
·
110 Posts
I think, preferably, the actual file. That allows others to examine parameters in different combinations that might not have been included in another's graphs.

That said, unfortunately, .csv files can't be attached to posts here, but they can be uploaded to a web storage site, and a link posted here so that others can download it.
Google-drive should work nicely.

As far as what PIDs to log, I see cardoc's list in post #2344. Many have different names in Torque Pro, but I think I can get most of them.
 

·
Registered
2006 Subaru Outback 2.5i Limited wagon
Joined
·
110 Posts
I'll defer to @cardoc on the list.

I'm not familiar with Torque Pro. Can the sampling be faster? 1/sec seems rather slow.
I thought I'd seen Torque Pro mentioned here as apparently the best option for non-Windows folk (can't use Rom Raider). Yeah, I believe the sampling can be faster, but depends on the capability of your OBD2-Bleutooth adapter.
 

·
Registered
1993 Legacy Wagon Auto trans
Joined
·
4 Posts
If I can throw a word of caution? Recorded PIDs can be helpful, but remember you are observing computed data not actual data so do not fall into the trap of thinking you are seeing "gospel last word" information. As an example: You can receive a "Bank lean" code and long/short fuel trims to support it. However, this is determined by an "Oxygen sensor" not a "fuel" sensor. What does that mean? If you have a cylinder with even "perfect" mix of 14.7:1 air/fuel ratio and that cylinder fails to fire the engine computer will "see" all that unused oxygen as a lean condition even though raw unburnt fuel is coming from that bank (bank being used for non-inline engines). Too much O2=Lean, too little O2=Rich.No actual, affordable, road reliable measurement of fuel can be done with current technology.
 

·
Registered
Joined
·
119 Posts
I'm wondering how beneficial it would be to catch the CEL tripping when logging? I'm thinking of logging all the time until I catch it, a lot of work but if helpful may try. Mine throws the code sporadically, I went 1200 miles between them a week ago, then only 93.
 
2381 - 2400 of 2418 Posts
Top