Fuel injector leak or HPFP?

[wilmar13]

My car started to burn more fuel (about 20% reduction across board), used to get ~28mpg on highway at 75mph, now about 24mpg.

I had several issues, but short version is I have new PCV, new turbo, new O2 sensors, new cat, new MAF. These restored power (had a loose wastegate), but now I am getting P0420. Fuel usage did not improve. I pressure tested intake and found a small leak that on a clamp that didn't appear until 25psi. It didn't do anything to resolve as it is stock so I don't think it even hits this high of a boost. Vacuum at idle is 3in Hg at dipstick as expected. I see no exhaust leaks at manifold, turbo to cat, or cat to exhaust.

My short and long term fuel trims are decent at idle and under use (less than 10% always). I'm at a loss to figure out the problem.

Before I changed the cat, car would have a pretty strong unburnt fuel smell at startup that is not there now. There is no abnormal smell of gas in oil. But I checked the fuel rail pressure actual vs specified, and while its all good... I am under the impression that when engine is hot, and you turn off ignition you should see increase in actual pressure. This does not happen. It pretty much stays where it was when running. 4000Mpa, dropping just a teeny amount over a few minutes to like 3970Mpa. But it doesn't increase due to heat soak. Does this tell me anything? Is it indicative of a fuel injector not totally closing?

Either my new cat is bad and the mpg reduction is due to something else, or its running rich and for some reason the MAF and precat o2 sensor is not adjusting fuel trims. Other thoughts?

How do you test the injectors?

[Theiceman]

you test injectors by turning on ignition on a cold car without starting engine, then turn it off and get back out and check plugs, if one is soaked that injector is leaking... but if it was, the car is usually hard to start.

since you had no code before changing this parts it stands to reason its one of the parts you put in..... did you put in factory cat or generic aftermarket ? id try putting factory cat back in and see what happens.

[wilmar13]

you test injectors by turning on ignition on a cold car without starting engine, then turn it off and get back out and check plugs, if one is soaked hat injector is leaking... but if it was the car is usually hard to start.

since you had no coed before changing this parts it stands to reason its one of the parts you put in..... did you put in factory cat or generic aftermarket ? id try putting factory cat back in and see what happens.

Aftermarket cat... I am thinking same (putting factory cat back on)... but the fuel usage has me thinking there is something else going on. The fact that the fuel rail pressure is not rising (actually dropping slightly) after turning engine off due to heat soak seems to point to something... like an injector is partially open or something. It is definitely not totally failed as that is more overt. No issues starting car and plugs are dry, though I have not done the turn ignition on cold engine, then checking them. I would think fuel rail pressure would drop quickly if the injector was open, but a slight and very slow drop indicates I have a tiny leak somewhere (at least I think it does, haha).

[Theiceman]

Aftermarket cat... I am thinking same (putting factory cat back on)... but the fuel usage has me thinking there is something else going on. The fact that the fuel rail pressure is not rising (actually dropping slightly) after turning engine off due to heat soak seems to point to something... like an injector is partially open or something. It is definitely not totally failed as that is more overt. No issues starting car and plugs are dry, though I have not done the turn ignition on cold engine, then checking them. I would think fuel rail pressure would drop quickly if the injector was open, but a slight and very slow drop indicates I have a tiny leak somewhere (at least I think it does, haha).

yeah i have never done the fuel rail test thing you speak of but i have definitely heard of it before ... would be nice if you had a buddy with the same car you could test against

I never look at fuel economy as it simply isnt important to me in the least. But i would be careful using it to diagnose an issue . Gas formulations change, for seasons, car usage change for seasons ( Ac vs no Ac ) winter vs summer tires might change it, all these will affect fuel mileage .. There are just too many variables for me to try and use this data to determine if i have an issue.

I do understand your point however that it could be a clue to another issue ...
I think i would not worry about the fuel efficiency and concentrate on your OBD codes and clearing them up.

but something hypeerthetical to think about ( Nitroviper is going trough the same thing ) . If your pump seals were leaking, and it were leaking fuel into the oil , this would drop your rail pressure. it would also be drawing fuel laden vapours via your PCV into your intake. Would that all be burned off or would the o2 sensor pick it up as unburned hydrocarbons ?

Thee is also the tune to consider and i have no clue about that ...

But there is no way in hell i would replace the cat if i wasnt getting a code regardless of what it smells like . ( unless of course you put in a HFC for other reasons. ) . If this car even farts wrong it seems to throw a code, so if you dont get one , you can bet the cat is working as it should.

[Smac770]

The high pressure rail fuel pressure should go up a lot after engine off. I posted this in some thread some day, but it had risen to 170 bar (170 bar = 17.0 MPa = 17,000.0 kPA) before finally tapering off and dropping. Keep in mind my shortest drive is over 5 miles.

The system is a 200 bar design. The HPFP should hold to that pressure, before venting back to the low pressure side. Idle pressure is ~40 bar and WoT is 150 bar. So you should see 4000 kPA fuel pressure value at idle, then engine off, then see it rise to at least 10,000 kPA if not higher, depending on how long you've been running. If it's not rising from heat soak, either not enough fuel was actually in the rail, or the rail is leaking pressure somehow. Could be through an injector, could be through a seal, could be through the HPFP.

Long term fuel trim is what? + or -? You're not rich if it's +10%.

You have a new cat and now have a P0420. This is one of those "hi flow" cats? Is this an Audi OE cat? If not, why do you presume it has the same performance signature as an Audi OE cat?

A leak in the intake at 25 psi boost is not really a leak, being that the intake track should never see 25 psi boost.

What exactly is your lambda doing during normal driving? Is it reading away from 1.0; is the long+short mix adjusting in accordance?

And poor fuel economy could just as much be a problem in the drivetrain or brakes or tires or whatever.

[silver_tt]

P0420 is a cat efficiency code so it is pretty common with aftermarket and especially high-flow cats; if you switched back to the new OE cat I'm assuming that would go away. Some people use spacers with the O2's to try to trick the ECU. In my cars when the cats have gone out I replaced it with the OE cat for this reason (cause I didn't want to have to play games with "defeats" and spacers). Fortunately this car only has 1, so it's not too bad......

[wilmar13]

I never look at fuel economy as it simply isnt important to me in the least. But i would be careful using it to diagnose an issue . Gas formulations change, for seasons, car usage change for seasons ( Ac vs no Ac ) winter vs summer tires might change it, all these will affect fuel mileage .. There are just too many variables for me to try and use this data to determine if i have an issue.

I do understand your point however that it could be a clue to another issue ...
I think i would not worry about the fuel efficiency and concentrate on your OBD codes and clearing them up.

but something hypeerthetical to think about ( Nitroviper is going trough the same thing ) . If your pump seals were leaking, and it were leaking fuel into the oil , this would drop your rail pressure. it would also be drawing fuel laden vapours via your PCV into your intake. Would that all be burned off or would the o2 sensor pick it up as unburned hydrocarbons ?

Thee is also the tune to consider and i have no clue about that ...

But there is no way in hell i would replace the cat if i wasnt getting a code regardless of what it smells like . ( unless of course you put in a HFC for other reasons. ) . If this car even farts wrong it seems to throw a code, so if you dont get one , you can bet the cat is working as it should.

Tons of variables to mpg on same vehicle, primarily atmospheric conditions, acceleration/deceleration, speed driven (it takes 8x the power to go 2x as fast), energy density of the fuel by season and supplier, losing alignment, brakes dragging, wheel bearings, and on and on. This is a real thing that I am taking all that into account on, isn't one tank, it was sudden and has occurred for more than 10k mi. For 90k mi of same usage car had 25mpg avg, now is down to 20mpg. Cold start and around town seem to have the biggest drop, but driving on freeway both ways is a better control. Its running rich. Its just like you say I was trying to ignore it, but I can't ignore it in big picture because it seems to be related.

I replaced the cat because I was getting p0420 occasionally, along with gas smell at start up, and a history of burning oil, it seemed sensible the cat was dead. The replacement was a Federal (non-CARB) cat from Rock Auto. It has same cell density as OEM and ceramic substrate, but volume of catalyst is ~.5" smaller diameter. I swapped back in OEM last night, it took me 90min to do, funny how it took me 4x that long first time. But P0420 is back almost immediately. A refurbished OEM cat is $1300! But it looks like I may be going that route. But I feel sure the car is running rich, the new non-performance cat failing efficiency tests right away seems to support that. New Bosch O2 sensors as well.

[wilmar13]

The high pressure rail fuel pressure should go up a lot after engine off. I posted this in some thread some day, but it had risen to 170 bar (170 bar = 17.0 MPa = 17,000.0 kPA) before finally tapering off and dropping. Keep in mind my shortest drive is over 5 miles.

The system is a 200 bar design. The HPFP should hold to that pressure, before venting back to the low pressure side. Idle pressure is ~40 bar and WoT is 150 bar. So you should see 4000 kPA fuel pressure value at idle, then engine off, then see it rise to at least 10,000 kPA if not higher, depending on how long you've been running. If it's not rising from heat soak, either not enough fuel was actually in the rail, or the rail is leaking pressure somehow. Could be through an injector, could be through a seal, could be through the HPFP.

I'm going to revisit this, I rebuilt the vacuum pump and removed the HPFP fitting, I'm going to re-tighten the fitting and see if that does anything. Though I would think you could smell gas under hood if it was leaking even slowly. If there is a leak it is slow, but yeah the fuel pressure does not build as it should, though it doesn't drop either.

Long term fuel trim is what? + or -? You're not rich if it's +10%.

Long term was + 6%... you assume that all the sensors are sensing correctly and reporting back accurate information. But someone isn't. That is what I am chasing here. Unless the new cat (not a low cell density cat) is a total fraud, the most obvious explanation is I am burning rich and it overwhelms the cat... yet the O2 sensors are not seeing it. I think I am going to try to put my old O2 sensors in one at a time and see if anything changes on lambda values and fuel trim.

There could be a minor fuel leakage issue into oil, but I know what gas in oil smells like, and this car doesn't have it at a level I would suspect that at all... still why would the wide-band not compensate for it if it is finding its way back into combustion chamber via PCV?

[silver_tt]

The brand new OE cat from Audi is only $950 for my CPMB, so that refurbished price seems too high....

https://www.audiusaparts.com/oem-par...LWw0LWZsZXg%3D

Look at your pre and post cat o2's to see if it runs rich:
capture_739a7fd106810735bd560f111e436a16ce0349e5.jpg

[wilmar13]

The brand new OE cat from Audi is only $950 for my CPMB, so that refurbished price seems too high....

https://www.audiusaparts.com/oem-par...LWw0LWZsZXg%3D

Look at your pre and post cat o2's to see if it runs rich:
capture_739a7fd106810735bd560f111e436a16ce0349e5.jpg

Yeah that is a little bit better... but there is a $275 core charge on top of that, tax, and shipping both ways. Its a >$1k purchase either way.

[wilmar13]

And that link you sent is already up to $980 before core, tax and shipping. haha... was it seriously 950 and they raised price in 4hrs or was that a typo?

[silver_tt]

There is no tax because we don't live in Oregon.

$980 is the website price, the price I pay is slightly less than that due to my 20 year relationship with them -- $30 difference, close enough though. The $275 core charge is fully refundable. That place also only charges actual shipping costs and they get amazing rates due to the volume they ship. So it's probably right at $1,000 all-in.

I'm not saying it's cheap but cats never are........ on my track car they both went out at once and it was $4,500 ;)

[wilmar13]

unnamed.jpg
25 min drive... G071 is precat O2 sensor voltage, G073 is post cat O2 sensor voltage

I guess 2nd stays at .455v until it warms up? I don't know why it keeps throwing P0420 code. Maybe the deltas are too far apart. It appears the voltage settings are inverse of each other. As the S1 gets leaner (higher voltage), the S2 voltage goes to 0.

I logged several markers, one of which was "G228 Long term adaptation of oxygen sensor control"... it was mostly 0 but for several long sections it has value of -0.8, vacillating back to 0, then back to -0.8. then 0 for duration. No other values. I don't really know what to make of that.

[silver_tt]

I can see why. The red is your post-cat o2 sensor and it's spending most of it's time at the high end of its voltage range. An oxygen sensor will typically generate up to about 0.9V when the fuel mixture is rich and there is little unburned oxygen in the exhaust. When the mixture is lean, the sensor's output voltage will drop down to about 0.1V. Wide band sensors have different voltage ranges and should range between .7V and 4.0V, where .7 is very rich and 4.0 very lean.

[wilmar13]

I can see why. The red is your post-cat o2 sensor and it's spending most of it's time at the high end of its voltage range. An oxygen sensor will typically generate up to about 0.9V when the fuel mixture is rich and there is little unburned oxygen in the exhaust. When the mixture is lean, the sensor's output voltage will drop down to about 0.1V. Wide band sensors have different voltage ranges and should range between .7V and 4.0V, where .7 is very rich and 4.0 very lean.

oh, yeah OK only one is wideband. not apples to apples, duh. Thank you. So my cat is indeed bad. So before I spend $1100 on new cat, I should probably swap back in the aftermarket cat (not a HFC) and see if I get same result. At least this confirms the cat isn't working. But I don't really want to poison an expensive cat if I am running rich. I am an idiot when it comes to electrical stuff. But I am learning.

[silver_tt]

yes it looks like it to me, especially if you compare to my graph which is from my car now with a healthy cat that does not code

one thing I was wondering is, what kind of driving were you doing for those 25 mins? all highway? start and stop, just regular around the neighborhood up to the grocery store? Reason I ask is because I noticed your graph does have you running rich for most of the time samples... your pre cat is most of the time in the 1.2-1.6 range which is pretty rich... Look at my graph. The only time mine dipped and was really running rich was at time 400 where I punched it WOT on a wide-open straight. otherwise it's usually lean and only dips down rich probably when I was accelerating. My drive was a 5 min drive or so in a loop/square around the neighborhood... so it was a regular sample of driving with some straights, maybe a couple of stop lights, braking, accelerating, then big acceleration at time 400 as I said, etc.

[wilmar13]

It was leaving park, had some stop lights, a bit on freeway and then rolling two lane. I didn't accelerate hard at any point. Most of it was low throttle cruising, idling, and commuter zombie relaxed acceleration.

yeah, if 1.5v is lambda=1, it does show as running a lot below 1.5v. Which is rich... as to why the car is adding fuel to itself in its trims I don't know. The long term adaptation of mixture formation started at 0 after I reset the DTC and stayed positive getting to 12% at one point. So the ECU sees it is rich and adds more fuel. I would like to understand why. I wonder if there are adaptations that are deeper in there to reset, that occurred to me when I reset the transmission shifting adaptations. I drove this car (2011 Q5) with a sporadic p299 due to loose wastegate for probably close to 20k mi before really getting into it.

[silver_tt]

So ya, I agree with you. Your driving would have been roughly similar to mine except I never did freeway and mine was mostly cruising and a couple stops.... but close enough that I agree with you that yours appears to constantly be running rich. I am confused. One of the primary functions of the pre cat wideband o2 is to communicate back to the ECU and give it information to adjust the trims. Agree with you, questions is why does yours show running rich a lot but the ECU isn't leaning it out.

I don't think there should be any deeper adaptations or anything related to running with a loose wastegate even for 20k miles. Are you running a tune on your car or anything? I'm also confused here but interested in this problem.

Also just out of curiosity have you seen your spark plugs recently? Curious of their condition for starters.

Is it possible you could actually have a vacuum leak that the ECU is trying to overcompensate for by richening???

[silver_tt]

Sorry saw you said you tested with a pressure gauge and it was 3" Hg. That reading looks ok ballpark wise.

It's best to test with a manometer, this is my reading taken in " H20 (which is close to 3" Hg which is why I say yours looks ok ballpark)

IMG_7358.jpg

[RPMtech147]

I just noticed the exact same issue. Just so happens it started at exactly the same time I started running the AC as it's been in the 90's lately. In the morning i get 29-31 @70ish mph. On the way home, it's over 90 outside and I run the AC the entire 20 miles. My home trip mileage drops to around 24mpg.

[RPMtech147]

unnamed.jpg
25 min drive... G071 is precat O2 sensor voltage, G073 is post cat O2 sensor voltage

I guess 2nd stays at .455v until it warms up? I don't know why it keeps throwing P0420 code. Maybe the deltas are too far apart. It appears the voltage settings are inverse of each other. As the S1 gets leaner (higher voltage), the S2 voltage goes to 0.

I logged several markers, one of which was "G228 Long term adaptation of oxygen sensor control"... it was mostly 0 but for several long sections it has value of -0.8, vacillating back to 0, then back to -0.8. then 0 for duration. No other values. I don't really know what to make of that.

P0420's are pretty cut and dry. I suppose you could have a faulty O2, but 99% of the time it would throw a fault for the O2, not the cat. You can test the rear O2 by inducing a vac leak. The voltage should drop to near 0.

Just something to note: O2's don't "swing" by themselves. They just read the mixture. The swinging is actually the ECU slightly changing the fuel mixture from lean to rich. A TWC(the kind of Cat most modern cars have) needs that Lean-rich sweep to function properly. Reading raw O2 data evaded me until I learned that and once I figured it out, it made interpreting the data more clear.

That looks like a dying cat. It should sweep from near 0 to near 1 for just a bit when it's cold(The chemical reaction hasn't started taking place so it's seeing that lean-rich sweep the ECU is running) Once the cat lights off, the chemical reaction starts and the rear O2 can't see that rich-lean switch "through" the cat. It should settle around .5 and just barely swing. Any time you get on the throttle you'll get a good swing but cruising, it should stay pretty steady; that's how the ECU knows the cat's working. The Front O2 on the other hand should always be swinging high to low while cruising (Correct me if I'm wrong, but isn't the front O2 wideband, and you're watching an AMP swing apposed to the Voltage swing on the rear O2)

[wilmar13]

So ya, I agree with you. Your driving would have been roughly similar to mine except I never did freeway and mine was mostly cruising and a couple stops.... but close enough that I agree with you that yours appears to constantly be running rich. I am confused. One of the primary functions of the pre cat wideband o2 is to communicate back to the ECU and give it information to adjust the trims. Agree with you, questions is why does yours show running rich a lot but the ECU isn't leaning it out.

I don't think there should be any deeper adaptations or anything related to running with a loose wastegate even for 20k miles. Are you running a tune on your car or anything? I'm also confused here but interested in this problem.

Also just out of curiosity have you seen your spark plugs recently? Curious of their condition for starters.

Is it possible you could actually have a vacuum leak that the ECU is trying to overcompensate for by richening???

Stock tune. I am just glad the O2 voltage matches what my human senses tell me. It smells rich. The vacuum leak is a potential. I looked for boost leak, but not vacuum. There is another thread on here where a guy was chasing his tail and it turned out the problem was a cracked dipstick tube. I will try to do the WD-40 thing all around and see if I can find something. Any other way to detect a vacuum leak?

I had spark plugs out recently when I replaced PCV, or no it was when I installed turbo... I removed them to run starter and prime oil into turbo. They looked a bit like it was running rich, which I hoped would be resolved with the new turbo, edit: and new O2 sensors. The plugs and coils are 1yr old and have 10k mi on them.

[wilmar13]

Sorry saw you said you tested with a pressure gauge and it was 3" Hg. That reading looks ok ballpark wise.

It's best to test with a manometer, this is my reading taken in " H20 (which is close to 3" Hg which is why I say yours looks ok ballpark)

IMG_7358.jpg

Yeah you have a thread I think I remember from that photo looking for data points. Really I had 5cm of Hg which is more like 2.5" of Hg, but its same as my '17 Golf R (using vacuum gage on dipstick tube) at idle.

[wilmar13]

I just noticed the exact same issue. Just so happens it started at exactly the same time I started running the AC as it's been in the 90's lately. In the morning i get 29-31 @70ish mph. On the way home, it's over 90 outside and I run the AC the entire 20 miles. My home trip mileage drops to around 24mpg.

Nah, I am pretty analytical and from other life pursuits am very aware of the things that cause fuel economy variance. I was just using the fuel economy decrease as tangible evidence as it has been a macro shift controlling for other variables like air temps, wind speed/direction, elevation, barometric pressure, humidity, driving conditions, winter fuel etc. The exhaust smells rich, the plugs look like its rich. Now I can understand that the wide-band is showing it running rich under easy cruising situations. MPG is not the primary thing that has me thinking I have a rich condition that ruined my OEM cat and is bad enough I'm still triggering a P0420 with a new 48state aftermarket cat. I am pretty sure a new OEM cat would pass, but if I don't figure out why it is running rich, it will have a short life.

[old guy]

More info on O2 sensor voltages: Clicky click®

[silver_tt]

Any other way to detect a vacuum leak?

Yes. You can use a smoke tester at ambient (cold engine, same as outside temp) and pump smoke into the intake.

For vacuum leaks the manometer is the first line of defense because it's very cheap (I built that device for about $45) and easy (takes literally 1 minute). A manometer is just a very fine vacuum gauge but very effective. A lot of people think you can just use a vacuum gauge, which you can, but it doesn't offer the level of precision required. I have now gathered enough data points to know that the 35" H20 I'm pulling in that picture is a healthy value.

Another thing you can use is a non contact pyrometer, and infrared temp reading device that you use to look at each header tube temp; the over rich tube would be much cooler, the overly lean compensating tubes much hotter.

[wilmar13]

Yes. You can use a smoke tester at ambient (cold engine, same as outside temp) and pump smoke into the intake.

how does that find vacuum leak and better than pressurizing intake? Just visual? I have this: https://turboboostleaktesters.com/au...llet-aluminum/

And I pressurized it to 30psi, only finding a leak on the pressure side of turbo at 25psi, which wasn't really an issue as this car doesn't boost that high. If there is a leak that only shows on vacuum, wouldn't the smoke tester have same issue? I am just trying to find issue myself before taking it to a pro, who may or not be able to find it, and I don't have a smoke tester.

Another thing you can use is a non contact pyrometer, and infrared temp reading device that you use to look at each header tube temp; the over rich tube would be much cooler, the overly lean compensating tubes much hotter.

That's a good idea, but I don't seem to have an issue at idle and to rig up something for live driving seems a bit of a hassle. I do hear a whistle on full boost, but for all I know that is just the normal sound of a properly functioning turbo.

Just for sanity, I logged the MAF sensor, and at full throttle I am getting 170g/s which correlates to 211hp (rough conversion is .8g/s to HP).

I did the wd-40 spray all over and no change to idle.
MAF- good
O2 wideband- good
actual intake charge vs specified - good

Yet, engine is still adding fuel despite O2 seeing its rich. This is getting frustrating.

It's very possible the issue started when the thermostat was replaced by an Indy mech, and I just didn't notice it immediately. I guess I may need to remove the intake manifold, do the intake valve cleaning (hey why not?) and put everything back together carefully.

[silver_tt]

Did you notice the change in gas mileage after you changed the cat? One thing I wanted to point is that sometimes when people have a failing cat they actually get better gas mileage because the cat really isn't doing anything and the flow is not restricted. However when they put on a new cat, even a HFC, the gas mileage will be reduced as the new cat is doing more. When you have a bad cat the exhaust often will smell like you are running rich (it can also smell like rotten eggs etc).

I also want to revisit if your pre cat O2 is really showing that it's running rich all the time. If it really was I'm confused why your car isn't coding for running too rich, usually these cars are pretty sensitive in this regard. When I get time at some point I will try to reproduce my results using a longer highway drive like you had done rather than just a spin around the block (I will do it this weekend when I take my daughter to horse riding if I'm the only other one in the car). I did that test a few weeks ago to help someone else out and I remember the route I drove but I can't remember how many stop lights I hit. It was a "city drive" rather than a "highway drive".

One other thing I wanted to point out is that if you replace your cat with the OE cat, even if you were running rich it would be ok for a short or medium period of time -- I don't think it will kill your cat right away or anything.

Still thinking about all of this..........

[wilmar13]

No its a complicated past that was made worse by across country move. The thermostat was replaced because I was busy doing other stuff at about 85k mi. I had to drive it across country to move dogs and we were going to be in temp corp apt so to have some basic personal items. I didn't have my tools or VAGcom with me, they were held hostage by movers for several months until we found a house. It had problems at altitude, that got noticable at 4000' but real bad above 7000' but that was probably the wastegate mostly. I put in new air filter right before I left using audi toolkit but could not find any issues. When I got to east coast I ordered plugs, coils and new MAF as cleaning the one on car in hotel parking lots seemed to offer some help (I thought I may have gotten dirt in MAF at time, clueless as to what problem was). By the time I got my stuff back, we had decided to split which put fixing car lower on priority list... so a few weeks ago I finally got around to fixing issues a yr and 15k mi later (has just over 101k mi now). It ran OK, but was low on power as boost was intermittent and maintained 20% MPG drop. New PCV, new turbo, power restored and ran fine but the P0420 came as soon as readiness cycled. It may have been there before, but I just don't know. Then installed the new Federal cat and O2 sensors... but still instant P0420. Which brings me to point am now. Vacuum leak somewhere makes sense, but other than MAF, O2 sensor, charge(boost) sensor, IAT I don't know what the ECU is using to determine its getting unmetered air to add fuel when it is already rich and it sees that it is. Long term fuel trims are always positive.

The only other thing I can think of is to reset throttle body alignment which should be done if its removed. It is simple to do, going to go do it now. I don't have much faith it will do anything, but I won't be mad if it solves issue.
http://wiki.ross-tech.com/wiki/index...lignment_(TBA)

[silver_tt]

Ya I can understand being busy with other stuff. Sounds like it's been a rough year with a lot going on, I hope it gets better for you, just keep grinding. I went on a ~30 min trip today to take my daughter to horse riding and here is the log:
Capture.jpg

This was started with a cold engine (note post cat has constant voltage at the beginning like yours) and the drive was almost all highway as I live very close to a parkway and it's about a 30 minute drive or so. First ~20 minutes is highway going 70 or 80 (my kid is in the car) then the last 10-15 is on a 2 or 1 lane road doing 40 or 50 mph. A few stops in there but not too many. Just to give you a flavor for the drive style.....

You know your car better than anyone because I can't see it but I'm not 100% convinced from the data that the car is broken beyond the cat. If it was my car I would replace the cat with the OE if I got the P0420 and cleared it a couple of times and it keeps coming back. When I got it in my track car I cleared it but it would come back pretty quick.

[silver_tt]

Just something to note: O2's don't "swing" by themselves. They just read the mixture. The swinging is actually the ECU slightly changing the fuel mixture from lean to rich. A TWC(the kind of Cat most modern cars have) needs that Lean-rich sweep to function properly. Reading raw O2 data evaded me until I learned that and once I figured it out, it made interpreting the data more clear.

That looks like a dying cat. It should sweep from near 0 to near 1 for just a bit when it's cold(The chemical reaction hasn't started taking place so it's seeing that lean-rich sweep the ECU is running) Once the cat lights off, the chemical reaction starts and the rear O2 can't see that rich-lean switch "through" the cat. It should settle around .5 and just barely swing. Any time you get on the throttle you'll get a good swing but cruising, it should stay pretty steady; that's how the ECU knows the cat's working. The Front O2 on the other hand should always be swinging high to low while cruising (Correct me if I'm wrong, but isn't the front O2 wideband, and you're watching an AMP swing apposed to the Voltage swing on the rear O2)

Front is wide band and the rear is a narrow band (traditional) O2. My other car was port injected and the pre-cat and post-cat are interchangeable (both narrow band); with that car the pre-cat O2 voltage literally oscillates, or sweeps as you put it, between .1 and .8. It's a sine wave with a short period with floor=.1 and ceiling=.8. The 2.0T has a pre-cat that's a wide band (I'm pretty sure it switched to wide band with the introduction of DFI, same with other brands), I notice the car is able to control the fuel trims much tighter. It "wobbles" more than full oscillation and obviously does so with a much smaller/tighter amplitude.

[wilmar13]

You know your car better than anyone because I can't see it but I'm not 100% convinced from the data that the car is broken beyond the cat. If it was my car I would replace the cat with the OE if I got the P0420 and cleared it a couple of times and it keeps coming back. When I got it in my track car I cleared it but it would come back pretty quick.

The brand new 49 state catalytic converter not able to pass readiness checks is what really had me going back to a rich A/F as cause. Its a model specific cat (for B8 A4/5/Q5) not a universal. Its possible the substrate is defective in the new cat.

Yeah I can see yours goes rich a lot but time spent is different. I ordered a new OEM cat. Going to just stop looking for problem until then. I do still suspect something is off with fuel injector, like when they changed my water pump, the manifold was removed and something may not be sealed perfectly, but will table that for now. I looked at my data and by getting rid of all values >1.6v (which is compression braking mostly), the average of remaining values was like 1.47V, so its not really running rich.

[silver_tt]

Cool, I think the OE cat was the smart move here because you can eliminate issues that can be caused by a HFC or non-OE cat that are "noise" in trying to get to your real issue(s) here. I agree with you... get that in there and then level set and see where you stand then. These cars are, in general, pretty sensitive to things like vacuum leaks and AFR's being off so I'm thinking maybe just the new OE cat is going to have you sorted. We will see..........

I know you said you just recently moved. Not sure where you moved from or what Fletcher, NC is like but one thing to note is that the quality of gas and where you buy it definitely matters. I always only buy gas at large name brand stations (BP, Shell, whatever) with known detergent packages and never at generic brands like "Bob's Fill-N-Stop" because you don't know what kind of additives or off-brand detergent packages they are using which can mess up fuel injectors and other things. You could also try running a bottle of something like this... not saying it's going to change your world but it's also only $10. The guy that owns LN Engineering lives in a smaller town outside of Chicago and said all the gas there is like Circle K now, nothing he can do, so he runs a bottle of this all the time:

https://drivenracingoil.com/i-304978...oz-bottle.html

[Smac770]

P0420 on the EA888 is strictly measured OSC / modeled marginal OSC < 1. The car will test the cat OSC when conditions suit and divide that by the lookup table value of a marginal cat OSC, the lookup table based on air mass flow and modeled exhaust gas temp (yet another thing the ECM needs but has no actual measurement of the real engine for; speculates using yet another lookup table). You can query this test result in measuring values: IDE01873, Test of catalytic converter: measured value. What I don't know is if the result is the result of the actual measurement stage or the division. I only know that on my car, after several minutes of operation, I get a value. For me it was 1.3318 on the trip to dinner and 1.3679 on the trip home.

P0420 is not going to reflect that you're endlessly running rich or lean, as all it cares about is the duration of the test. And it can validate the exhaust comp using the lambda sensor. The live lambda (minus 1) is part of the integration formula.

That said, if there are problems such as incorrect MAF readings leading to incorrect air mass (yet another integral you can monitor in measuring values), or other problems leading to different actual EGT than what the ECM is predicting (you can see EGTs in measuring values, and they are all presumed computed lookup values, there is no actual EGT measurement on our engine), these would lead to incorrect lookup value for a marginal cat OSC.

The lambda sensor voltage output in measuring values is made up by the ECM to give a value for old diagnostic techniques. But the lambda sensor is a wide band that produces a variable current flow to reflect the lambda, and that measuring value is more informative anyway. For the voltage value, you'll see it around 1.5 for lambda 1, walk down to 1.0 as you hold WOT, and jitter around 2.6 off throttle. You should ignore the voltage value and monitor the lambda value instead.

The jump sensor is a normal voltage measurement, at 0 idling, .4-.6 cruising, around .8 WOT. The ECM considers there a problem if the jump sensor voltage is above .2 when not on throttle.

[Smac770]

I do like this video on the P0420 and OSC (oxygen storage capability) subject:
www.youtube.com/watch?v=vyVnhCIMDnw

[wilmar13]

P0420 on the EA888 is strictly measured OSC / modeled marginal OSC < 1. The car will test the cat OSC when conditions suit and divide that by the lookup table value of a marginal cat OSC, the lookup table based on air mass flow and modeled exhaust gas temp (yet another thing the ECM needs but has no actual measurement of the real engine for; speculates using yet another lookup table). You can query this test result in measuring values: IDE01873, Test of catalytic converter: measured value. What I don't know is if the result is the result of the actual measurement stage or the division. I only know that on my car, after several minutes of operation, I get a value. For me it was 1.3318 on the trip to dinner and 1.3679 on the trip home.

P0420 is not going to reflect that you're endlessly running rich or lean, as all it cares about is the duration of the test. And it can validate the exhaust comp using the lambda sensor. The live lambda (minus 1) is part of the integration formula.

That said, if there are problems such as incorrect MAF readings leading to incorrect air mass (yet another integral you can monitor in measuring values), or other problems leading to different actual EGT than what the ECM is predicting (you can see EGTs in measuring values, and they are all presumed computed lookup values, there is no actual EGT measurement on our engine), these would lead to incorrect lookup value for a marginal cat OSC.

The lambda sensor voltage output in measuring values is made up by the ECM to give a value for old diagnostic techniques. But the lambda sensor is a wide band that produces a variable current flow to reflect the lambda, and that measuring value is more informative anyway. For the voltage value, you'll see it around 1.5 for lambda 1, walk down to 1.0 as you hold WOT, and jitter around 2.6 off throttle. You should ignore the voltage value and monitor the lambda value instead.

The jump sensor is a normal voltage measurement, at 0 idling, .4-.6 cruising, around .8 WOT. The ECM considers there a problem if the jump sensor voltage is above .2 when not on throttle.

When I get the new cat, I will evaluate it vs the current one. I will play with the OSC values in vagcom... I have seen several tests but not familiar with them. It will be nice if IDE01873 is same across platforms, but it seems they are all just a little different. I question why you would want to ignore the voltage. Are you sure the voltage is extrapolated from the ECM? I thought it was the other way around, that the lambda values were extrapolated by the voltage signal. haha If it produces a variable current flow, then there should be a variable amperage field at least. For sure the O2 sensor does not output lambda... that is an equivalency created by whatever the signal is that varies (either voltage, amperage, or something else) Right?

My MAF sensor is another culprit I would like to eliminate. I did the max flow test and got what the engine should be making at max power, but I am not confident that tells me much about how reliable it is at lower flow rates.
My symptoms:
- smell of gas on cold startup
- fuel rail does not build pressure from heat soak
- got P0420 after installing new turbo which if it was there before, it was very intermittent, P0420 repeated after new O2 sensors and new 49state cat installed, cleared and returned several times, could not get through readiness checks.
- noticeable drop in fuel economy drop controlling for all variables except one major thing: It is comparing southern CA to rest of country. Perhaps there is more joules/gal in CA formulation 91 octane vs 93 octane which has "up to 10% ethanol"... I am going to ignore fuel economy as there are just too many unknowns... the smell of gas on cold start up (way above normal cold car start, it smells like its running super rich) and fuel rail not building pressure from heat soak after engine stops, but not leaking it either makes me suspect a very tiny leak in injectors, but I don't know enough to know if this is possible and if so if that really explains it.

[wilmar13]

The new cat (49 state) being completely incapable of passing readiness checks plays an annoyingly confusing part in this. I think its unlikely that the cat was totally defective, but possible. I do wonder since this car is a CA car, if the ECU programs are just a bit tighter than what goes to other states. That seems very unlikely, but when they were updating software for PCV, I suppose its not impossible that CA mandated its own nanny features into it since at the time VAG was in deep doodoo over the TDI emissions scandal. It would explain some of why I got a P0420 with new O2 sensors and 49 state cat.

Perhaps the fuel trim adaptations were crazy lean due to long term running of loose wastegate, so then with new turbo it over compensated and is still looking to hone in on middle ground. Since there is no EGT sensor, perhaps it tries to richen things up to get the catalyst hot too when it sees no result at S2. I don't know. It runs great now, and I have kind of exhausted my interest in this car since its my dog hauler/beater/utility car. When I get the new cat, if everything clears, I will probably just consider it good.

[morris39]

P0420 on the EA888 is strictly measured OSC / modeled marginal OSC < 1. The car will test the cat OSC when conditions suit and divide that by the lookup table value of a marginal cat OSC, the lookup table based on air mass flow and modeled exhaust gas temp (yet another thing the ECM needs but has no actual measurement of the real engine for; speculates using yet another lookup table). You can query this test result in measuring values: IDE01873, Test of catalytic converter: measured value. What I don't know is if the result is the result of the actual measurement stage or the division. I only know that on my car, after several minutes of operation, I get a value. For me it was 1.3318 on the trip to dinner and 1.3679 on the trip home.

P0420 is not going to reflect that you're endlessly running rich or lean, as all it cares about is the duration of the test. And it can validate the exhaust comp using the lambda sensor. The live lambda (minus 1) is part of the integration formula.

That said, if there are problems such as incorrect MAF readings leading to incorrect air mass (yet another integral you can monitor in measuring values), or other problems leading to different actual EGT than what the ECM is predicting (you can see EGTs in measuring values, and they are all presumed computed lookup values, there is no actual EGT measurement on our engine), these would lead to incorrect lookup value for a marginal cat OSC.

The lambda sensor voltage output in measuring values is made up by the ECM to give a value for old diagnostic techniques. But the lambda sensor is a wide band that produces a variable current flow to reflect the lambda, and that measuring value is more informative anyway. For the voltage value, you'll see it around 1.5 for lambda 1, walk down to 1.0 as you hold WOT, and jitter around 2.6 off throttle. You should ignore the voltage value and monitor the lambda value instead.

The jump sensor is a normal voltage measurement, at 0 idling, .4-.6 cruising, around .8 WOT. The ECM considers there a problem if the jump sensor voltage is above .2 when not on throttle.

Your take on OSC motivated me to measure (8V motor, MQB platform though) but I get values I don't understand.From Module 01 via VCDS
IDE01873 Test of catalytic converter: measured value: 0.078
IDE01874 Test of catalytic converter: measured value 1.078
IDE01874 Test of catalytic converter: result Test not started
Can you share your source of info? Emissions and readiness all good BTW.I want to establish a baseline for future ref.

[Smac770]

Morris, as 1874 just say test off, I assumed the alternative value would be test on or test running or something, not an actual numerical value. I'll have to log that value too and see what comes up.
It could be then the 1873 is the integration sum computing the current cat OSC, and 1874 is the result of the division by the OSC value of a "marginal cat" from the lookup table based on modeled EGT and measured air mass integral during the test.

I start with the Quick Reference Specification Book for the model and year. I find these from google searches, often as documents at NHTSB. These provide a decent scope of what Audi actually uses on that engine as a trigger for that P code. Though sometimes, like with the P0420, you look at it and think "saaaay whaaat?"
You can also review the Audi and VW EPA certification applications on the EPA site. The Audi docs tend to just say "see confidential section", but the VW docs tend to have the OBD summary chart of P codes.
There's also the Bosch ME7 engine strategy document on ERWIN; it might be out there on the net too. As it's ME7 and not MED17, it's a nice starting point, but you can't take it as authoritative.
Then there's the various Audi and VW SSPs out there.

A US 2015 S3 would be engine code 215kW CYFB. Is that what you have? A CYFB is a Gen3 engine, so something we never got in the NAR market B8 vehicles. I can't find a quick ref spec book for a Mk3 A3 on google. Maybe someone with an active ERWIN access and pull it.
As such, I can't say what the P0420 trip point would be for that car, though division result of <1.0 seems logical assuming cat efficiency is testing in the same manner still.

[Smac770]

There is not a separate CA emissions program in the CAEB (this thread is about the 2011 Q5 2.0T?). The programming is 50 state as is.
The transverse version in the Mk2 A3/TT had a separate CA compliant version. But that Gen1 engine didn't have a number of improvements of the Gen2 engine in the B8.

That you don't build pressure in the high pressure rail after engine off says to me you probably have a HPFP that's leaking pressure prematurely back into the low pressure circuit. You should be able to tell that by throwing a fuel pressure gauge on the soft line going into the HPFP, but you'd need a counter measurement from a "working" install to know how the movement of the value should be different.