PDA

View Full Version : STFT still 25+....need help diagnosing



wclepse
03-22-2017, 02:24 PM
A while back I posted about having a P0171 (http://www.audizine.com/forum/showthread.php/753971-Fuel-trim-indicating-a-leak) which I assumed was related to a vacuum leak since I found one. So I started to tear into the system and found a bunch of leaks including the secondary O2 sensor nonfouler. I ordered replacement parts and basically replaced all the check valves the suction pump, breather (http://www.audizine.com/forum/showthread.php/756895-Replacing-breather-urgent-help-needed) and PCV.

I just went out for a ride and noticed the car seemed to be stumbling a little and when I checked the STFT it is still reading in the upper 20's but no code thrown or pending yet. Vacuum is still reporting at over 20mmHg via Torque at idle. So I am pretty sure I got all the leaks in the vac system, is there somewhere else I should look? Searches suggest logging MAF which I think I can do in torque but is that sufficient for someone to help diagnose? Any other sensors to log? Other places to check for leaks? Starting to think it isn't a leak since when I get on it the STFT doesn't go down as the intake should be positive and then leak out not in. Right?

Work done since initial code.
New air filter
Changed oil/filter
Replaced oil level sensor
Replaced elbow from SAI to hard pipe
Replaced all check valves
Replaced breather
Replaced PCV
Cleaned MAF
New Nonfouler (reinstalled old O2's)
New turbo to cat gasket

wclepse
03-22-2017, 04:45 PM
Thought of one other thing....I was getting a P0012 for quite a while and ended up replacing the solenoid on the CCT just before these other codes started coming up. I kind of mentioned it in the first thread I linked but didn't say why I changed it. Not sure if this code is somehow related even though changing that did eliminate that code, but then got the P0170.

Rodgman15
03-22-2017, 05:00 PM
Boost/vacuum gauge? Don't trust torque for your boost or vacuum readings.

wclepse
03-22-2017, 05:20 PM
No stand alone gauge. I thought Torques vac reading was supposed to be accurate but boost was questionable since it is calculated rather than measured. Is that not correct?

Rodgman15
03-22-2017, 06:34 PM
I've never even gotten it remotely close on my car, no matter what offset I put in. I only use torque to clear codes, never had good results in terms of logging.

Every B5 should have a boost/vacuum gauge, modded or not.

wclepse
03-22-2017, 06:40 PM
I do have an old laptop with VCDS Lite but the battery barely holds a charge so I have only used it in the garage, not sure if I could do any logs with it. And unfortunately my cable is serial and that is the only machine that has one. But if I get that working, is there a vacuum gauge that the computer sees? I can look to get a gauge but not sure how quick I can find one. Maybe I can snag something from work.

Rodgman15
03-23-2017, 11:52 AM
What year car? If you're wideband you should be able to get a number off of the map sensor. Narrowband, you're out of luck and need a gauge.

I recommend getting a gauge regardless.

wclepse
03-23-2017, 01:01 PM
2001.5 AWM engine code

Rodgman15
03-23-2017, 05:34 PM
Maybe the map doesn't read vacuum, only boost, I'm not sure as I'm DBC.

OverSpun
03-23-2017, 07:31 PM
You only replaced the solenoid on the CCT? Interesting... I replaced my CCT w/ a new OEM one when I had similar codes and it fixed the issue.

The aftermarket POS CCT that was in the car when I bought it and the screen fell off it and it threw the codes since the CCT was not working anymore.

wclepse
03-23-2017, 08:07 PM
Yeah just the solenoid, figured it was worth a shot. When I took the old one off I put 12V across it and while it actuated it was weak. The new one made a solid click on activation.

walky_talky20
03-23-2017, 09:29 PM
1.8T's do not have an actual vacuum (MAP) sensor at all. The 2000+ cars just have a charge air pressure sensor (AKA: boost sensor). The sensor is located directly on the intercooler, so it cannot possibly see manifold vacuum. Like...ever.

The only way you can get a vacuum reading is by connecting your own vacuum gauge to the manifold.

1. Can you find out your long term trims (LTFT) - both the additive and multiplicative (idle and part throttle)? That would be very helpful in determining where your problem may be.
2. Have you logged the MAF value? Does it run good enough to do a log on a 3rd gear pull, wide open throttle, from 2k-6k RPM?

wclepse
03-24-2017, 09:41 AM
I have a gauge I will hook up tonight and get some vac readings. Car does run but seems to be a bit rough, not sure if I would feel comfortable doing a wide open throttle run but certainly could do 2-6k run. Haven't logged anything and will have to see if my VCDS Lite laptop even works. I left it charging all day, hopefully I can get a few minutes to collect data. As for the LTFT, it was reporting basically the same as the STFT in Torque but I suspect that is because the battery was drained when I finished working on it. I probably left a door open or something. Anyway I figured that wiped old data out and that was why the values were very close.

wclepse
03-24-2017, 04:43 PM
well I underestimated the ID of the tube I planned to tap into, I think I got the clamps tight enough to make it leak proof but I think I want to get a better fit for driving it around. But since it seemed tight enough I did an idle test. Started it up and got a fair amount of smoke at first...don't think I have ever seen that before. While the SAI was on the vac was only 13-15"Hg, and after it shut off and the idle settled it went to about 18. So I guess I must still have a leak somewhere...thoughts?

After I get the larger T I hope to be able to log a 3rd gear pull and post up. But if you think I should do something else first let me know.

wclepse
03-25-2017, 09:01 AM
No one had the right size parts locally so I tried my luck. Idle vac was not as good this time, more like 15-16 but as I was engine braking while coasting down a small hill I saw a high of 26. During my drive I saw a max boost level of 12psi. After doing this pull I ended up with the following codes: 17536 - Fuel Trim; Bank 1 (Mult): System too Lean P1128 - 35-00; 17544 - P1136 - 35-00. Idle was horrendous and it smelled like crap, mostly WOT felt way smoother than expected but since I was on a rural road I couldn't quite complete the pull I thought I killed it around 5k but looks like the data stopped at 4k. For some reason I can't post attachements so what can I do with the log I have to get some feedback?

Let me know if this works: https://drive.google.com/open?id=0B5gjjbk24QoVTVdNb3BWejhPOGc

wclepse
03-25-2017, 03:45 PM
I went back out but this time made sure my APR tune was enabled. It felt pretty good, this time the log goes to ~5k and you can see I laid off and shifted. Let me know what the consensus is since I really have no idea what it means, other than the MAF looks like it stumbled at one point in this pull.

https://drive.google.com/open?id=0B5gjjbk24QoVTnZoMUtGaHBqOUk

ActiveMonkey
03-25-2017, 07:29 PM
Depending on how old your upstream O2 Sensor is I would replace it. Mine had ~128k miles and out of nowhere the car would start knocking under WOT, then under part throttle and then anytime I would create more the 4psi. Point of this is that your upstream O2 Sensor is also your A/F sensor and if it is old or dirty it can cause gremlins and make you chase ghosts. For a while no codes, then a code for "too lean" then P0133 and P0136. I swapped out my upstream O2 and POOF! runs flawlessly.

Best of luck.

walky_talky20
03-25-2017, 10:07 PM
I think I'm going to second the O2 sensor. The MAF numbers look great. 5k revs at full load shows 160 g/s. That equates to 200hp, which is exactly what it should be making on a chip tune.

Assuming your coolant temperature sensor is reading correctly (~90-95C in VCDS at operating temp), then I think your O2 sensor might be lying. Reporting a lean condition that doesn't exist and thus the ECU is dumping in way too much fuel. I would proceed as follows:

- Check that coolant temp reading is accurate (as reported by the ECU via VCDS or VCDS Lite)
- After you warm it up and do a good rip, shut the engine off quickly and see what the spark plugs look like. If it is actually lean, they will be light colored. If the O2 sensor is lying, they will be black.

I would probably proceed with O2 sensor replacement if temp sensor is accurate and plugs are black. There is some further testing you can do on the sensor, but I think you need more than VDCS Lite. This is a 2001, so the primary O2 sensor is a wideband sensor. They have a lot of control over fuel mixture, so it makes good sense that this may be your problem. It will be more expensive than your average sensor, though. About $100. Looks like FCP has a good price on these:

https://www.fcpeuro.com/products/audi-volkswagen-oxygen-sensor-17351

If you don't feel confident, I believe I have a known good AWM primary sensor laying around. I changed mine when chasing an ignition timing issue. The new one didn't make a difference, so I'm pretty certain the old one still has some good life left in it. I could send it to you to try if you want.

wclepse
03-26-2017, 07:15 AM
Thanks again for all the help...it is much appreciated. First thing I did when backing out of the garage was to clear the code and immediately the idle went pretty smooth. At least as smooth as it can be with 280k. Not sure why when I previously cleared codes this never happened. Anyway, I wasn't sure how to confirm the temp readings but from what I recall there are two temp senders, one is for the dash the other is for the ECU. So what I did was confirm that when the dash was reading mid point the ECU reported temp was around 90C. I then got on it coming up a hill and it felt good to me, then had to drive about 1/2 mile through the neighborhood before I could shut her down. I pulled the plug and it is light colored (https://drive.google.com/open?id=0B5gjjbk24QoVc0hMWDRacHV3ZDQ) on the electrode dark by the threads as you can see in the picture. But when you mentioned dark deposits I remembered that when I removed the cat I noticed the primary O2 (https://drive.google.com/open?id=0B5gjjbk24QoVbm43VDhCV3ZHV1k) was dark. So I pulled that and took a picture of its current condition.

As a side note, are those vac readings normal? just wondering if something else is still leaking since the plug is light in color. Walky...appreciate the offer but I am anxious to get this thing running again. Thanks for the link though, found my local AAP has it n stock and with a code is slightly cheaper. So if we think it might be that I could throw it in and see. But plug color has me concerned now.

walky_talky20
03-27-2017, 07:45 AM
Plug color looks ok. The fact that clearing the codes (which clears learned fuel trims) made it run BETTER, means that its learned adjustments were wrong. This is a pretty good indicator of a faulty oxygen sensor.

Generally when you have a car with a bad MAF or a bunch of leaks, it is exactly the other way around. The car will be running sort of OK, then you clear the codes and it will barely run. This is because the heavy adjustments were the only thing keeping the mixture right.

I would feel pretty good about putting a fresh oxygen sensor in it. Even if that is not your entire issue, the car does have tons of miles on it, so it wouldn't be wasted money.

wclepse
03-27-2017, 07:57 AM
Txs for the advice...I will try to pick one up tonight. I hope this works because the wife is getting tired of me working on it and spending money on it. As you said it has lots of miles so she thinks I should just get something new. But I like the challenge, just wish I had your knowledge.

As for the vac readings, is 15-18 OK at idle and mid 20's with engine braking?

wclepse
03-27-2017, 02:22 PM
Well I was excited for about five minutes. I put the new sensor in started it up and idle was smooth. Went for a quick drive and it felt strong, I saw 15psi or so of boost. Then I get home and notice that the idle is stumbling again, used Torque to pull pending code P0170. Clearing the code evened out the idle again. One other oddity was a noise it made when I shut it off, almost like something was powering down or relieving pressure internally. Any thoughts?

walky_talky20
03-27-2017, 02:34 PM
Pressure sound is probably your Air conditioning. Normal.

I'm sorry the sensor didn't fix it for you. I'm surprised. But at least we can 100% trust the primary sensor data now.

I would like to see LTFT data from VCDS. And a datalog of Actual vs Target AFR on a 3rd gear pull.

That will help us make more sense of this. LTFT is block 032 which is not reachable by VCDS Lite unless registered or otherwise unlocked. Actual and Target AFR are block 031, also unreachable by Lite.

wclepse
03-27-2017, 04:47 PM
That presents a problem...I don't have a registered version. Never got it because a buddy had the full VCDS but he has since moved. I will have​ to check around and see if anyone local can help me out.

Any chance info from torque would be useful?

Rodgman15
03-27-2017, 05:12 PM
Actually, believe it or not, but I think Torque can measure LTFT and STFT. Don't quote me on that, can't remember.

Bordom
03-27-2017, 05:35 PM
Actually, believe it or not, but I think Torque can measure LTFT and STFT. Don't quote me on that, can't remember.
I have it and it can

Sent from my SM-G935W8 using Tapatalk

Rodgman15
03-28-2017, 05:15 AM
I have it and it can

Sent from my SM-G935W8 using Tapatalk

Good deal, I have it on the tablet that's integrated into my dash, but I literally only using it for clearing my EVAP code when the CEL comes on lol.

It may not be as in depth as VCDS (which you should still try to track down), but maybe it'll provide a bit of insight.

wclepse
03-28-2017, 04:46 PM
So I took it out after clearing the codes, today it didn't feel as smooth under power and just like before the idle got wonky after driving. The first pull I also logged RPM but since there is only one scale most data was lost. The second pull starts with me slowing down in third to around 2k RPM then WOT to about 5k. The graph shows a stumble near the end, maybe I laid off a bit because I could feel it miss. Anyway let me know if this has enough detail to be useful.

https://drive.google.com/file/d/0B5gjjbk24QoVWDctSGxWXzZ5MFU/view?usp=drivesdk

wclepse
03-28-2017, 05:42 PM
Sorry just realized I could log to a CSV file so I went out again. Log starts in first then I get to third and did a pull. It felt pretty good but in at least one pull it stumbled just under 5k.
https://drive.google.com/file/d/0B5gjjbk24QoVdjFMcGN1RzNmckE/view?usp=drivesdk

wclepse
03-28-2017, 05:43 PM
Other log
https://drive.google.com/file/d/0B5gjjbk24QoVRGNmUVJ5Y2RQTkU/view?usp=drivesdk

wclepse
03-29-2017, 05:17 AM
Sorry...I was on mobile last night and didn't realize those links wouldn't open in a spreadsheet, so here are new links. Really not sure what to think of this data, one shows the rpm going from 845 to 5016 in the next reading. Makes no sense, even at idle if I hit the throttle I don't think it would spin up that fast. Anyway if someone sees something telling please enlighten me.
https://drive.google.com/open?id=0B5gjjbk24QoVd0M3MV84V2thSmM
https://drive.google.com/open?id=0B5gjjbk24QoVZkswT04xZXVPMlU

walky_talky20
03-29-2017, 12:06 PM
These logs aren't showing both long term trims. There is an idle trim and a part throttle trim. It is hard to pinpoint the problem without those details.

To gather that data, you will need VCDS and you will need to put some miles on the car without clearing the codes so it can learn. Probably 2 trips where it warms all the way up would be sufficient. Trims are in block 032. Just read them after it learns. No need to log them.

From your logs, it looks like it's hitting pretty good AFR's. 14.7 and dipping 10.8 at WOT. Pretty much what my AWM does. Coolant temp looks good. Timing looks good. MAF values are wacky (not shown in g/s) so I'm not sure on those.

wclepse
03-29-2017, 03:21 PM
Ok txs I will run it tonight and hopefully get to to a full VCDS tomorrow night. From an earlier post you also suggested block 31 so I plan on grabbing that too since getting to the full version won't be as easy as using mine. So with that said anything else I should get?

walky_talky20
03-29-2017, 09:04 PM
Would like to see a log of the MAF with Throttle position and RPM. Then another log with block 031 along with RPM. If you can do all those together, that's great. I try not to log more than 2 blocks at a time because it slows down too much. Would also like to see timing retard (knock control).

Then grab the trims from 032. Engine does not need to be running for that. But codes must not be cleared for several miles/trips before you view them.

wclepse
03-30-2017, 07:48 PM
So tonight on the way over to do some logging the CEL came on. Before we started logging the code showing was:
17536 - Fuel Trim; Bank 1 (Mult)
P1128 - 35-00 - System too Lean

Then after doing some pulls and idling for about 15 minutes the following codes were showing:
17536 - Fuel Trim; Bank 1 (Mult)
P1128 - 35-00 - System too Lean
16684 - Random/Multiple Cylinder Misfire Detected
P0300 - 35-00 -
16687 - Cylinder 3
P0303 - 35-00 - Misfire Detected
17519 - Oxygen (Lambda) Sensor Regulation: Bank 1
P1111 - 35-00 - System too Lean
16688 - Cylinder 4
P0304 - 35-00 - Misfire Detected

Again the idle felt like crap but the pulls felt relatively smooth though tonight it seemed to feel a little weaker than some other recent pulls. Log can be found here (https://drive.google.com/open?id=0B5gjjbk24QoVMlRYUTh2U21fSmc) and here (https://drive.google.com/open?id=0B5gjjbk24QoVNDRXYTNkVUdTUXM). Any thoughts?

walky_talky20
03-30-2017, 08:26 PM
Wow. The idle trim is too high, but the part throttle is pinned. One would expect the MAF to be shot, but the numbers look fine for stock power. The numbers from the Wideband o2 sensor don't look normal, though. An AWM should be pretty good at keeping the actual matching the request. Yours is running richer than target almost all the time, and it seems oddly locked at .727 much of the time.

Some questions:
- When did it start doing this?
- Has the car been modified (chip tune of any kind, any go-faster parts in the engine bay, etc)?
- What is the deal with the "non fouler", exactly? All the details on that, please.

Possibilities I'm thinking of:
- Injector size is not correct
- Fuel pressure is not correct
- O2 sensor(s) are not getting an accurate read for some reason
- Software has been modified.

wclepse
03-30-2017, 08:48 PM
Car has 280k on it and has an APR chip since about 10k, so yes programming is different but this is not a new change. At around 150k the stock cat died and we installed a high flow cat, the only way to get the readiness to come up clear (ie not below threshold error) was to install the non fouler. So while I changed the non fouler recently this was because the old one rusted out and I had a leak. Is the secondary O2 only for cat effectiveness?

So recent history is this...I needed to get the vehicle inspected so I finally did something about the CEL I was getting for the CCT (P0012). I picked up a new CCT but only swapped in the solenoid, this eliminated that code, but when I was trying to get my readiness to come up I got P0170 which I figured was a vac leak. Poking around in there I found several leaks and decided to replace all the parts mentioned above. Prior to this I don't recall having this rough idle but I do remember thinking the car stunk and a buddy driving behind me said it was obnoxious. After changing the parts that is when the idle was most pronounced. Could I have gotten a bad check valve or inadvertantly put one in backwards? I did check them by blowing into them but I can't generate the pressure or vacuum the engine can. I did keep the stock ones since they passed the blow test, would it make sense to put those back in? One other question, my MAF sensor has a spot for a resistor but none is connected or in the housing...are all sensors supposed to have this?

walky_talky20
03-31-2017, 08:30 PM
Thanks for the information.
- Secondary o2 sensor is "mostly" for monitoring the cat. It does have a slight "fine tuning" effect on fuel mixture in some Audi software versions.
- A check valve is not likely to be your issue. With a 25% multiplicative trim, we are talking a huge correction for tons of air. Not a little vacuum leak.

Questions:
- The non fouler: it is installed just on the rear o2 sensor, right?
- Can you be 100% certain the o2 sensors are in their correct positions? This is critical.

wclepse
04-01-2017, 10:26 AM
Non fouler is only on the rear sensor. Ans I am fairly certain they are in the right spots. The sensor just before the cat has more pins than the rear one. I think it is like 6 vs 3.

Since this code wasn't coming up before and you think this is a big leak any chance I got a bad part or something that could cause this...like the suction pump? What about the breather I put in,,I had a hell of a time getting the clip in since the o-ring was so tight. Could a damaged o-ring there potentially cause this? Just trying to think what would have caused this shift since the car ran fine before.

wclepse
04-03-2017, 05:51 AM
Not sure what to think here, this morning was the first time I really drove the car for a while. This morning I cleared the codes and started my 40 mile commute, I noticed that the idle was smooth the entire time so when I got to work I checked the codes and there is a pending P0171 but the STFT is reading 0.0 with the LTFT saying 5.5. To further confuse things I noticed my display is saying I averaged around 20.5mpg, normally this would be around 27. So it is definitely dumping more fuel but now that the STFT is 0.0 and a relatively low 5.5 for LTFT I am not sure why it would be dumping that much.

walky_talky20
04-03-2017, 06:41 AM
Talked to my father about this (he's an actual mechanic). Basically, with that high of a part throttle trim, it's almost always a faulty MAF. But since your MAF values look perfect, we are thinking that it has to be incorrect fuel delivery. Injectors are being pulsed properly, but for some reason the incorrect amount of fuel is being delivered. It would be ideal if the ECU were a known stock item instead of something modified, but it's unlikely to be software if this thing worked fine before.

Recommendation: check the fuel pressure. Make sure to check static pressure (should be 4 bar / 58 psi) and should go down with with idle vacuum and up 1 to 1 under boost pressure.

If fuel pressure is good, next would be to check injector flow. There is also a chance for a wiring issue where the injectors aren't getting enough voltage. Easy enough to check that with a meter.

wclepse
04-03-2017, 10:03 AM
For my own education; how do you know the injectors are being pulsed properly? I do agree that it is unlikely to be the tune since it hasn't changed in 270k miles but would it help to change the tune to "stock."

I will check some local auto parts places to see if they have fuel pressure kit, but do I need any strange fittings that the kits don't usually have? You also mention the voltage to the injector, I assume the voltage is only supplied when it is to open which happens super quick so how do you get a good reading on that?

thanks again for the continued tech support.

walky_talky20
04-03-2017, 10:52 AM
- Well, to be clear, we don't exactly know for certain they are being pulsed properly yet. There could be trouble with the ECU output driver circuitry such that they aren't. But we basically know the ECU is seeing the correct amount of air, and the correct temperatures. So it should be giving plenty of injector pulse. For some reason it's adding a lot more than the map calls for to obtain the AFR it wants.

- I don't think the fittings are anything special. The kit I used on mine was pretty standard.

- The voltage supplied is constant. Whenever the fuel pump is running - the injectors have battery voltage applied. It is the ground that is pulsed by the ECU to control them. So you can easily check the voltage supply while it's running to make sure that side is good, anyway.

wclepse
04-03-2017, 03:19 PM
Was watching the STFT and LTFT on the way home and noticed that the reason it was stuck at zero was the car was running open loop. Clearing fault codes brought back closed loop and then the STFT was bouncing around, saw it go negative and as high as 10.5. I'll see what it does tomorrow on the way to work. Don't think I would have time through the week to test the fuel pressure anyway, so I will just keep monitoring this stuff for now.

wclepse
04-04-2017, 04:48 AM
Today the LTFT ended up around 10 after 40 miles. The STFT was bouncing around, the highest I saw was under heavy acceleration and it was just into double digits. Most of the time in partial throttle it was reading -5 to 5. I know the LTFT is a timed average of the STFT, but is the STFT always centered around zero or is it shifted by the LTFT? What I mean by that is my current reading of 10 for LTFT means that a -10 STFT is actually 0 not -10 f things were running right? Also worth mentioning the mpg went up to 24 from 22 yesterday.

wclepse
04-04-2017, 02:36 PM
Afternoon update: still no codes, LTFT is under 10, STFT is usually between -5 & 5 and mpg avg is up to 25. Also worth noting is O2 sensor, evap system and cat all clear readiness. The only hold out is SAI. Could this be a wiring issue or something?

walky_talky20
04-04-2017, 06:36 PM
Looking better. You sure you didn't accidentally fill up with E85 one time or something?

On the SAI, output test mode in VCDS will check the pump, solenoid and combi valve very quickly.


Sent from my iPhone using Tapatalk

wclepse
04-04-2017, 07:59 PM
Always go to the same spot and pump it myself despite local laws so I doubt it but maybe I got some bad gas??

ben2401
04-04-2017, 11:40 PM
A while back I posted about having a P0171 (http://www.audizine.com/forum/showthread.php/753971-Fuel-trim-indicating-a-leak) which I assumed was related to a vacuum leak since I found one. So I started to tear into the system and found a bunch of leaks including the secondary O2 sensor nonfouler. I ordered replacement parts and basically replaced all the check valves the suction pump, breather (http://www.audizine.com/forum/showthread.php/756895-Replacing-breather-urgent-help-needed) and PCV.

I just went out for a ride and noticed the car seemed to be stumbling a little and when I checked the STFT it is still reading in the upper 20's but no code thrown or pending yet. Vacuum is still reporting at over 20mmHg via Torque at idle. So I am pretty sure I got all the leaks in the vac system, is there somewhere else I should look? Searches suggest logging MAF which I think I can do in torque but is that sufficient for someone to help diagnose? Any other sensors to log? Other places to check for leaks? Starting to think it isn't a leak since when I get on it the STFT doesn't go down as the intake should be positive and then leak out not in. Right?

Work done since initial code.
New air filter
Changed oil/filter
Replaced oil level sensor
Replaced elbow from SAI to hard pipe
Replaced all check valves
Replaced breather
Replaced PCV
Cleaned MAF
New Nonfouler (reinstalled old O2's)
New turbo to cat gasket

High fluctuating stft around 25% that goes away when you accelerate and gets worse as you slow down/stop (sometimes with rough running or misfires) in addition to being a vacuum leak could be a broken or stuck N80. The ecu uses the N80 like a 5th injector and after you clear codes or refuel it sometimes can take a while before the self-test or actual venting cause it to run lean (meaning the ecu is expecting the extra gas vapours but instead sees a lean condition while venting the tank). Usually the N80 gets vented on easy driving or gradual decel. coming to a stop or while idling at a stop. You can also observe fuel trims change in real-time after clearing faults and going on a drive. Some codes will cause the self-test cycle sequence to interrupt or make the car run open loop (like a P0171 Fuel System too lean caused from a combination of vac leaks in pcv system or intake/charge piping, bad MAF, leaking N80, clogged fuel filter) which make these problems harder to diagnose. You may not even see a P0411 for the Secondary Air system if you have it deleted if the sequence gets interrupted with another code beforehand.

I would make sure that you:
1. have no leaks in the S-hose (from the PCV to the S-hose) and that it is not collapsing under vacuum, check brake booster hoses and SJP, check F-hose, check DV line
2. your replacement PCV valve is OEM or at least flows the same when you blow through it (had problems with URO parts PCV valve and noticed they don't carry that one on Rockauto anymore)
3. make sure your DV is not ruptured if a diaphragm-type or lubed and sealing if piston-type and leaking air internally, this will cause a positive adjustment as well
4. check for leaks around the injector cups and seals and the throttle body and intake manifold, check FPR line for leaks or gas smell (gas in vac line from FPR to manifold means failing FPR)
5. Do a throttle body alignment - key on don't touch pedal for 1-3min - do it with hood open if you want to hear the little motor work
6. Reset all codes and drive until you get a code -- what comes up first?
7. take off the N80 valve and spray it out with brake clean or carb cleaner
8. if you have the SAI system installed a stuck combi valve will cause a vacuum leak
9. log your MAF at WOT 3rd gear to see if you get proper max hp. reading (peak hp x 0.8 for g/s) it should idle at 2.5-5g/s MAFs read lower as they age or get dirty

wclepse
04-06-2017, 10:01 AM
Thanks for the other suggestions. So far the car still seems to be driving pretty good and looking at the other things you mentioned I am thinking I might take a look at the combi valve. Items 1,2,3,4 & 5 I would think would be permanent issues and not fix themselves to allow me to drive normally now. Don't know enough about N80 valve but would think the leak path would be limited to volume of gas tank. But the combi valve could be sticking or something and since my SAI hasn't cleared maybe this is the reason? I do have 280k miles and that valve has never been replaced, so it is likely it is pretty nasty. Maybe it is worth pulling it out and cleaning it.

walky_talky20
04-06-2017, 11:23 AM
Usually the combi valves rust themselves closed. You can check it with vacuum or VCDS output test mode. If it doesn't make the nice plunger noise - it needs replaced. Cleaning probably won't do anything.