Jump to content
LegacyGT.com

Fueling issues: the "I've completely surrendered" edition


Recommended Posts

I'm fairly certain in AP terms, the parameter NSFW is describing is called "Target AFR" or something similar. "Commanded Fuel Final." From the Cobb Subaru Monitor List. If it's not in your gauge list, make sure your AP is fully updated. I believe I remember seeing these parameters when I used to have my old APv3:

 

Commanded Fuel Final (Comm Fuel Final) -> Final commanded fueling before the injector pulse width is calculated. This includes all corrections to fueling, including short-term and long-term fuel trims.

Commanded Fuel Primary OL Map (Comm Fuel Map) -> Commanded open loop fueling as determined by the "Primary Open Loop Fueling" table(s) with all direct compensations applied.

Link to comment
Share on other sites

  • Replies 56
  • Created
  • Last Reply

Troubleshooting a car over the interweb is like shooting a .338 Lapua at 1000yds blind folded.

 

You can ask for advice and general examples of similar situations and results. But in situations like this it is always best to get it in the hands of someone who has the proper tools and experience (which might take more than one person/shop).

Link to comment
Share on other sites

Given what he's describing I think there's a pretty good chance that would turn into having a tuner troubleshoot the car. Not sure if good investment. :)

 

"Final Fueling Base" is the name chosen by the RomRaider reverse-engineering guys, but Cobb probably does have the same parameter. It is the value from the fueling table plus all of the compensations. You can mostly think of it as AFR that the ECU is trying to achieve - that's not strictly accurate but it's close.

 

Anyway, if that parameter is moving around in ways that match your actual AFR then that would point to a problem with the tune rather than with something under the hood. And if that parameter is stable while your actual AFR wanders around a lot, then that would suggest that the tune is just fine but you've got a hardware problem of some kind.

 

In a steady cruise it will oscillate +/- 0.5 or so about once per second. That's just an emissions thing, don't worry about it, but if you see that oscillation then you've definitely found the right parameter.

 

The devil is in the details! I've attached some logs, and comparing Commanded Final Fueling to the front AF sensor shows a LOT of inconsistencies in the exact areas I'm having issues with. I didn't get a log of it, but during cold start the actual AFR wanders a between 13.0-16.0, whereas Commanded Fueling consistently leans as coolant temp rises from 10.0-14.6. However, in situations where there is no throttle position delta, it all lines up. This, to me, says fuel pressure issue or lazy AF sensor... no?

 

I'm fairly certain in AP terms, the parameter NSFW is describing is called "Target AFR" or something similar. "Commanded Fuel Final." From the Cobb Subaru Monitor List. If it's not in your gauge list, make sure your AP is fully updated. I believe I remember seeing these parameters when I used to have my old APv3:

 

Many thanks! I was about to look up COBB's table definitions but you saved me the trouble.

 

NSFW, I'm just thinking its been two weeks and the issue is still not fixed. May be its time to call in the experts.

 

If I install the regulator and there's still no improvement, I'll most likely do that. It's just hard to stomach the local shops' $150/hour diagnostic fees. Not to mention that honestly, it's a pride thing as well. I built this motor, I should be able to diagnose fueling issues on my own.

 

Troubleshooting a car over the interweb is like shooting a .338 Lapua at 1000yds blind folded.

 

You can ask for advice and general examples of similar situations and results. But in situations like this it is always best to get it in the hands of someone who has the proper tools and experience (which might take more than one person/shop).

 

I'll 100% agree with you. I come here for the knowledge of others, so that I might improve my understnding and ability to troubleshoot. For example, NSFW's suggestion to log Final Fueling Base was a fantastic idea, as it's led to a lot of discovery regarding my issue.

 

So, in short thus far:

1) Regulator should be delivered tomorrow, the last piece of my puzzle.

2) Actual AFR values do not line up with the ECU's commanded fueling values. They are always leaner, and in the exact spots I've been having issues with: rapid throttle change, and 2800rpm fuel harmonic area

Acceleration.csv

Cruise.csv

MTBwrench's Stage 3 5EAT #racewagon 266awhp/255awtq @17.5psi, Tuned By Graham of Boosted Performance

 

Everyone knows what I taste like.
Link to comment
Share on other sites

Let us know how the FPR works. IIRC you already replaced the front O2 sensor (which tend to read lean as they fail). I know you already verified the MAF (just make sure your volts are not swinging at the MAF, a swing of 0.2-0.4 is nothing to worry about). After that it could very well be a base map issue.
Link to comment
Share on other sites

If I install the regulator and there's still no improvement, I'll most likely do that. It's just hard to stomach the local shops' $150/hour diagnostic fees. Not to mention that honestly, it's a pride thing as well. I built this motor, I should be able to diagnose fueling issues on my own.

 

Beyond pride, putting in the time and effort to do this yourself (with the wonderful help of the forum and the very generous folks, from "regular" enthusiasts to professional tuners like Mike, Mike and Dave) means you'll be better off in the future with regards to taking care of your car. Not to mention, your steps in learning will help future visitors and likely your own generosity and time spent re-teaching what you've learned to others in person or on the forums will continue to bring more value to the community.

Link to comment
Share on other sites

Report after new FPR install:

 

Oh my god, that whole little rat's nest of diaphragms and hoses and clamps just sucks. Especially when it's corroded. I installed the OEM regulator onto the bypass assembly, relocated the reference line to the BPV as I had before, and started it up.

 

No improvement. On cold start it still hunts around cycling between 15ish and 13ish AFR. It starts up rich, as usual, then engine speed will drop and AFR's will slowly climb to 15.x, followed by engine speed increasing and AFRs normalizing. Starting from a stop the AFR graph looks like a sinus wave. Bottoms out, leans horribly, then returns back to stoic. The correction values show the ECU is trying to remedy that as well, pulling fuel during the rich spot and adding fuel during the lean spot.

 

AF learning values changed a bit (and they confuse me):

A: -8.10%

B: -2.10%

C: 1.70%

D: 0.00%

 

Im really really really hoping it's a basemap issue.

MTBwrench's Stage 3 5EAT #racewagon 266awhp/255awtq @17.5psi, Tuned By Graham of Boosted Performance

 

Everyone knows what I taste like.
Link to comment
Share on other sites

I can't tell from your posts if you have actually replaced the MAF sensor with a new one already or if you just cleaned the original sooty one and reinstalled it.

 

I see you have a GS intake. Does your current map have an off-the-shelf scaling for the intake your tuner had or was the MAF rescaled during the tuning process?

Obligatory '[URL="http://legacygt.com/forums/showthread.php/2008-gh8-238668.html?t=238668"]build thread[/URL]' Increased capacity to 2.7 liters, still turbo, but no longer need spark plugs.
Link to comment
Share on other sites

I can't tell from your posts if you have actually replaced the MAF sensor with a new one already or if you just cleaned the original sooty one and reinstalled it.

 

I see you have a GS intake. Does your current map have an off-the-shelf scaling for the intake your tuner had or was the MAF rescaled during the tuning process?

 

 

 

I have indeed replaced the MAF. I installed a paper filter on my GS intake as well. As for MAF scaling, it is scaled for my intake, however the tuner has not seen it in person yet to make any adjustments so it's a generic GS scaling. Goes on the dyno October 21st.

 

 

Sent from my iPhone using Tapatalk

MTBwrench's Stage 3 5EAT #racewagon 266awhp/255awtq @17.5psi, Tuned By Graham of Boosted Performance

 

Everyone knows what I taste like.
Link to comment
Share on other sites

That generic GS scaling could be part of the issue, not sure though. When I first installed my GS CAI and did the CL MAF scaling, it somewhat resembled the approx +12% over stock scaling that others on the forums have come up with, but it wasn't exactly that and it wasn't a flat % over stock. I used one of the OS MAF scaling tools across 2 or 3 iterations to help w/ dialing in the #'s.

 

Since the behavior seems to persist after (among other things you've done) installing the GS dryflow element, this is probably not contributing to the issue, but: did you take particular care to locate the filter element exactly 1" on the intake pipe as the GS instructions mention?

Link to comment
Share on other sites

That generic GS scaling could be part of the issue, not sure though. When I first installed my GS CAI and did the CL MAF scaling, it somewhat resembled the approx +12% over stock scaling that others on the forums have come up with, but it wasn't exactly that and it wasn't a flat % over stock. I used one of the OS MAF scaling tools across 2 or 3 iterations to help w/ dialing in the #'s.

 

Since the behavior seems to persist after (among other things you've done) installing the GS dryflow element, this is probably not contributing to the issue, but: did you take particular care to locate the filter element exactly 1" on the intake pipe as the GS instructions mention?

 

Funny you mention the scaling issue, as I was perhaps thinking the same thing. Though I might throw more money at it in the form of a new front O2 just as a last ditch effort... I've been noticing lately that it's slowly becoming more and more off from my wideband in the last week.(reads lean when wideand doesn't/reads okay when wideband reads lean) Perhaps the sensor has a high latency now and that's my issue. :iam:

 

And for the GS filter: Funny you mention that... I definitely did not not. I'll defend myself by saying that I put it back on exactly as much as I did before, and that was just fine before, but I know that's a bad defense.

MTBwrench's Stage 3 5EAT #racewagon 266awhp/255awtq @17.5psi, Tuned By Graham of Boosted Performance

 

Everyone knows what I taste like.
Link to comment
Share on other sites

Funny you mention the scaling issue, as I was perhaps thinking the same thing. Though I might throw more money at it in the form of a new front O2 just as a last ditch effort... I've been noticing lately that it's slowly becoming more and more off from my wideband in the last week.(reads lean when wideand doesn't/reads okay when wideband reads lean) Perhaps the sensor has a high latency now and that's my issue. :iam:

 

And for the GS filter: Funny you mention that... I definitely did not not. I'll defend myself by saying that I put it back on exactly as much as I did before, and that was just fine before, but I know that's a bad defense.

 

I suspect the potential for MAF scaling being a contributing factor was what fahr_side was getting at earlier, but barring hardware issues hopefully the upcoming protune will square that away.

 

Regarding the air filter positioning, I don't know exactly how critical that 1" onto the tube measurement is but I figure if GS highlighted it in the install instructions it's probably worth adhering to pretty strictly. I measured/marked the 1" around the circumference of the tube to make sure I had the filter on 'straight' and exactly 1" from the tube end.

 

Ironically I've just discovered that my Perrin turbo inlet has a leak at the BPV fitting so now I'm going to have to either replace it or try to seal it :/

Link to comment
Share on other sites

Similarly, if tuned using the Cobb platform - I have a working instance of AccessTuner Race (which I use on my unlocked calibration file), have the same car (05 OBXT) and could take a look at some things, but am inclined to leave that to your current tuner. Plus, if they locked/protected the tune I wouldn't be able to see anything within it anyway.
Link to comment
Share on other sites

That generic GS scaling could be part of the issue, not sure though. When I first installed my GS CAI and did the CL MAF scaling, it somewhat resembled the approx +12% over stock scaling that others on the forums have come up with, but it wasn't exactly that and it wasn't a flat % over stock. I used one of the OS MAF scaling tools across 2 or 3 iterations to help w/ dialing in the #'s.

Okay, bearing in mind I did the final scaling on DW750 injectors, my average scaling is 16% higher than stock. The closed-loop area is about 20% higher and the rest varies between 5% and 20% higher. OTOH the open-loop area was adjusted slightly lower for the 750s than the stock injectors I used to do initial scaling. I'll post this scaling for you if you want to try it. I spent a lot of time working on it but you use it at your own risk and all that.

Obligatory '[URL="http://legacygt.com/forums/showthread.php/2008-gh8-238668.html?t=238668"]build thread[/URL]' Increased capacity to 2.7 liters, still turbo, but no longer need spark plugs.
Link to comment
Share on other sites

Are you tuned with open source, or with Cobb?

 

If open source, I'd be happy to look at your tune. I don't have Cobb's software, though.

 

I'm currently on a locked COBB protune. :(

 

Similarly, if tuned using the Cobb platform - I have a working instance of AccessTuner Race (which I use on my unlocked calibration file), have the same car (05 OBXT) and could take a look at some things, but am inclined to leave that to your current tuner. Plus, if they locked/protected the tune I wouldn't be able to see anything within it anyway.

 

See above :( I have a working copy of ATR as well... but I'm far too afraid to change everything necessary to make it work with my setup.

 

Okay, bearing in mind I did the final scaling on DW750 injectors, my average scaling is 16% higher than stock. The closed-loop area is about 20% higher and the rest varies between 5% and 20% higher. OTOH the open-loop area was adjusted slightly lower for the 750s than the stock injectors I used to do initial scaling. I'll post this scaling for you if you want to try it. I spent a lot of time working on it but you use it at your own risk and all that.

 

Thank you for the offer! But, like above, I can;t really utilize any of this since my tune is locked... What I will say though is that under boost it usually runs 0.1-0.2 richer than commanded fueling. Also, I've noticed it pulling small amounts of fuelvia live correction even while the front o2 is clearly reading lean. And I'm totally confused as to why my AFR jumps all over the place. On the highway I can watch my wideband oscillate between 14.1-14.9AFR, and its not cyclical like an injector is bad or something. :confused:

 

Beginning to think this is an ECU issue, but still paying its tune. Doubting it is tune as I don't understand how the ECU's logic would ever arrive at some of these conclusions.

 

I didn't read the whole thread so my apologies if this has been suggested/ checked already but what if you have a TGV malfunctioning.

 

Those are long gone :)

MTBwrench's Stage 3 5EAT #racewagon 266awhp/255awtq @17.5psi, Tuned By Graham of Boosted Performance

 

Everyone knows what I taste like.
Link to comment
Share on other sites

New Discovery!!!

 

Comm. Fuel final steadily rises(with very light load) as high as 15.8AFR while cruising at highway speeds, however actual AFR's remain mostly safe. :iam:

 

I think the "LGT members help MTBWrench troubleshoot his wagon" session is over, I need to take this to a shop for a once over to ensure it's not mechanical.

MTBwrench's Stage 3 5EAT #racewagon 266awhp/255awtq @17.5psi, Tuned By Graham of Boosted Performance

 

Everyone knows what I taste like.
Link to comment
Share on other sites

Weird. That parameter will show when the ECU is altering fueling to achieve a target AFR, but I can't imagine what would cause it to pull fuel like that... it would seem to indicate that extra fuel was going into the engine somehow, but how? Leaky injector? But how would that cause the leaning that was happening earlier? Basically, I got nothing.
Link to comment
Share on other sites

Weird. That parameter will show when the ECU is altering fueling to achieve a target AFR, but I can't imagine what would cause it to pull fuel like that... it would seem to indicate that extra fuel was going into the engine somehow, but how? Leaky injector? But how would that cause the leaning that was happening earlier? Basically, I got nothing.

 

I'm really fairly sure this is a mix of injector scaling, not very well matched injectors, and MAF scaling issues at this point. Whatever is going on, the ECU is telling the system to do it!

 

- I can see injector pulse width(latency added) changing rapidly by a few ms upon tip in, and it correlates exactly with the "rich-lean(stall)-return to stoich" issue I'm having with a large TPS delta.

-When I'm idling ~775rpm the ECU is adjusting the A/F learning B value... which I'm pretty sure isn't right

-A/F learning A is pulling almost 10%

-At highway speeds and light throttle, I can watch the A/F learning C range max out at 15% with short term correction at 10-20%.

-My AFR's are VERY unstable... with cruise control on at 70mph they will range evrywhere from 14.1-14.7. I feel like theyre either badly matched or the per cylinder injector trim is wonky

-I've encountered many situations where commanded AFR witll be 16+ yet correction is still adding fuel and learning is doing the same.

 

I'm rather tempted to put on my stock intake, adjust a COBB OTS tune for my injector setup, and see what happens.

MTBwrench's Stage 3 5EAT #racewagon 266awhp/255awtq @17.5psi, Tuned By Graham of Boosted Performance

 

Everyone knows what I taste like.
Link to comment
Share on other sites

I'm really fairly sure this is a mix of injector scaling, not very well matched injectors, and MAF scaling issues at this point. Whatever is going on, the ECU is telling the system to do it!

 

- I can see injector pulse width(latency added) changing rapidly by a few ms upon tip in, and it correlates exactly with the "rich-lean(stall)-return to stoich" issue I'm having with a large TPS delta.

-When I'm idling ~775rpm the ECU is adjusting the A/F learning B value... which I'm pretty sure isn't right

-A/F learning A is pulling almost 10%

-At highway speeds and light throttle, I can watch the A/F learning C range max out at 15% with short term correction at 10-20%.

-My AFR's are VERY unstable... with cruise control on at 70mph they will range evrywhere from 14.1-14.7. I feel like theyre either badly matched or the per cylinder injector trim is wonky

-I've encountered many situations where commanded AFR witll be 16+ yet correction is still adding fuel and learning is doing the same.

 

I'm rather tempted to put on my stock intake, adjust a COBB OTS tune for my injector setup, and see what happens.

 

That's not necessarily wrong, but it means your MAF scaling and/or your AF learning ranges aren't set up as Subaru intended. If your AF learning ranges are stock, if when you're idling you're above 5.6g/s but below 10g/s MAF, then you'll be in the B range and it'll be adjusting that. Whether or not that MAF figure is accurate or not depends entirely on your intake and MAF scaling, but I'd say it's entirely undesired to have idle MAF ranges fall under your B learning range (meaning something in the tune needs to change).

Link to comment
Share on other sites

 

I'm rather tempted to put on my stock intake, adjust a COBB OTS tune for my injector setup, and see what happens.

 

This might be worth it. If your injector tuning isn't dialed in, you'll be chasing your tail with every change in maf scaling and fueling issues.

Link to comment
Share on other sites

If you have Cobb's tuner software, a picture of the MAF scaling curve might be interesting. It should be a fairly smooth curve, but if it is wavy that might explain going rich and lean as the airflow varies and the ECU uses different parts of the curve.

 

And yes I'm totally grasping at straws here...

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.




×
×
  • Create New...

Important Information

Terms of Use