Jump to content

Photo

Leaning out over time during races

- - - - - Air fuel ratio lean troubleshooting

  • Please log in to reply
10 replies to this topic

#1
HeavyRightFooT

HeavyRightFooT

    Member

  • Members
  • PipPipPip
  • 11 posts
  • Location:Bay Area
  • Region:California
  • Car Year:1999
  • Car Number:77

Hi all,

 

Continuing to have issues with the car leaning out over time when on track.

Initially had made a thread about the car leaning out over time at idle, but unfortunately it appears to continue to do so on track as well.

 

Would appreciate any help or tips as to how to how to continue to troubleshoot this !

 

AFR leans out over time and could not replicate on a shop dyno (dynojet). 

ECU, MAF, primary 02 sensor, wideband gauge and sensor, alternator, throttle body with sensors, coolant temp, air intake temp, egr valve, fuel pump, fuel filter, have been swapped in an out for known working or new pieces without effect. Smoked tested with no leaks.

Wideband is logged and close to dyno tailpipe reading so assuming it reads correctly.

Fuel pressure is logged and stable throughout the race.

Voltage doesn't appear to drop. (14v+ during entire race.)

 

After its last stop on the dyno at the track we set it up rich to make sure it wouldn't get too lean at the end of the race.

 

Not sure if related as this started only recently, but car also sometimes hangs at idle (1500rpm, then at some points drops back down to 850) 

 

Attached a screen grab of the AFR log vs RPM during my last race this past weekend: green is second lap, red is mid race, and blue is last lap, all in middle of the main straight.

 

Coolant temp was around 190 all race.

 

Any idea what to try next ? Voltage drop across wiring somewhere tricking the ecu into leaning the car ?

 

Thanks!

 

Olivier

Attached Files



#2
bmarshall1

bmarshall1

    Member

  • Members
  • PipPipPip
  • 298 posts
  • Location:United States
  • Region:SW Florida
  • Car Year:1999
  • Car Number:23

From what I understand a 'hang idle' means your air/fuel screw needs adjusted, some call it the idle screw, it's on the passenger side of the Throttle Body, maybe behind a cap.  I think* the proper direction is to back it out until it no longer hangs.  If it drops then comes back up you went to far.  The computer will always try to keep it around 850 +- (at least this is what I Have been told and it works for me).

 

I can't be any help on the A/F mixture.


  • HeavyRightFooT likes this

#3
speedengineer

speedengineer

    Member

  • Members
  • PipPipPip
  • 390 posts
  • Location:Michigan
  • Car Year:1999
  • Car Number:84
What year is your car?

- You know for sure that the car used to not change AFR with time on track? I'm not sure what is typical. Difficult to tell from the screenshot, but perhaps 1 AFR richer by the end?
- Assuming NB, Have you plugged in an obd2 scan gauge to confirm the the ECM reported IAT and coolant temps are appropriate? I'd especially try to capture these values when your idle hang occurs.
- Ensure you don't have any exhaust leaks, could affect your wideband reading. Not sure why it would only leak later on in a session, though.
- What does the wideband read at idle when the ecm is controlling afr to stoich via the switching sensor? Anything other than 14.6 to 14.8 might indicate a wideband accuracy problem. How about when you come off the track with the car hot?
  • HeavyRightFooT likes this

Jason Kohler 

#84 SM

www.youtube.com/user/speedengineering

 

NASA Champs Winner - NASA Champs Winner We have a Winnah! - Won their 1st race... Congratulations! Bona fide - A bonafide Spec Miata driver

#4
HeavyRightFooT

HeavyRightFooT

    Member

  • Members
  • PipPipPip
  • 11 posts
  • Location:Bay Area
  • Region:California
  • Car Year:1999
  • Car Number:77

What year is your car?

- You know for sure that the car used to not change AFR with time on track? I'm not sure what is typical. Difficult to tell from the screenshot, but perhaps 1 AFR richer by the end?
- Assuming NB, Have you plugged in an obd2 scan gauge to confirm the the ECM reported IAT and coolant temps are appropriate? I'd especially try to capture these values when your idle hang occurs.
- Ensure you don't have any exhaust leaks, could affect your wideband reading. Not sure why it would only leak later on in a session, though.
- What does the wideband read at idle when the ecm is controlling afr to stoich via the switching sensor? Anything other than 14.6 to 14.8 might indicate a wideband accuracy problem. How about when you come off the track with the car hot?

Thanks for the note !

 

-The car is a 99, built it over the last year. Only has 2 race weekends on it. New engine went in over the winter. Has done this every time it went to the track as far as I can tell. Before that no idea, it was a running street car I used to commute and drive around and it was not throwing any codes.

 

-Yes about 1.5 point leaner by the end of the race. The values for the position of the marker are in the columns on the left.

Car was set on a dino and was ending way lean during first race weekend. Re-set pressure after a session on the mobile dyno at the track during the last weekend, now starts the race way rich.

 

-Plugged my ODB2 scan and both IAT and Coolant temp look appropriate. Wasn't able to reproduce the idle hang, but it did have a code stored p0507 this time around. So replaced the air control valve with a cleaned spare and checked connection. Reset the idle. Code cleared now.

--> Does the ECU need a certain number if incidents over time to throw this type of code ? as in could the issue have started to cause a problem but the code only appear after a while ?

 

-As far as I can tell there are no exhaust leaks. But I'll go ahead and recheck and re-torque the manifold and down the line as well.

Both the primary o2 sensor and the wideband are in the down pipe, the primary in the stock location in the elbow right after the end of the exhaust manifold and the wideband 6-8" further back. Wouldn't an exhaust leak before that influence the primary as well and make the ECU go richer ? Although I imagine the Fuel pressure regulator effectively caps how rich the system is allowed to be.

 

- At idle the Wideband usually start around stoich for a while and then would eventually drift to leaner. I was told not to worry about idle readings so I did so, but since it continues during races over time I want to make sure I understand what is happening.

 

-Attached is a graph of the logged wideband with the car hot, left to cool down idling on the Dyno after 15min of pulls loaded trying to heat soak it to reproduce the issues.

Coolant temp (not on the graph) at the start is about 195 and goes down and stabilizes around160 mid way.

You can see the overall drift up (leaner) over time. Start around 14.7 and ends around 16. (for reference horizontal dashed line is 13, top one 20)

I condensed the graph to make it easier to see the drift over time and pay less attention to the extreme peaks.

Idle was between 850 and 900rpm.

 

http://photos.app.go...T5PKaLQF6cf23T6

 

I'll start it cold and let it completely heat up for a while see if it happens again now that the IAC has been replaced.

 

Thanks

 

Olivier



#5
HeavyRightFooT

HeavyRightFooT

    Member

  • Members
  • PipPipPip
  • 11 posts
  • Location:Bay Area
  • Region:California
  • Car Year:1999
  • Car Number:77

Tried to add an image but for some reason it didnt work and showed a broken icon in the preview...  is there a trick to using that function ?

 

For now just added a link.



#6
speedengineer

speedengineer

    Member

  • Members
  • PipPipPip
  • 390 posts
  • Location:Michigan
  • Car Year:1999
  • Car Number:84

Do you have a way of logging the stock primary O2 sensor (switching sensor).  Theoretically, at idle it should be toggling back and forth ("switching") between 0.1 and 0.9 volts.  This would indicate the ECM is successfully controlling fueling to achieve 14.7 at idle.  Or, at least what it thinks is 14.7. 

 

If the Primary O2 was switching for your entire 15 minute idle period (indicating stoich) while the Wideband O2 continuously leaning out, then I would suspect an issue with either the primary O2 sensor or the wideband O2 sensor.  I tend to suspect the wideband sensor to be more likely the issue, since it also leans out on track when the ECM should theoretically be fueling in open loop mode.  

 

If the Primary O2 was not doing any switching at idle, then it is possible that the wideband was accurate and the car did lean out during the idle period.  Then you'll have to figure out why the car doesn't run in closed loop at idle.  This may or may not be related to the leaning out that you see on track.  

 

 

Side note, not sure which wideband you own.  I have an AEM, and it's 5v output signal was quite noisy which made accurate datalogging difficult.  I put a simple RC (resistor and capacitor) filter on the output prior to my datalogger - it really cleaned up the signal.  See the attached image.  

 

Attached Files


  • HeavyRightFooT likes this

Jason Kohler 

#84 SM

www.youtube.com/user/speedengineering

 

NASA Champs Winner - NASA Champs Winner We have a Winnah! - Won their 1st race... Congratulations! Bona fide - A bonafide Spec Miata driver

#7
Steve Scheifler

Steve Scheifler

    Veteran Member

  • Members
  • PipPipPipPipPip
  • 2,816 posts

...
Side note, not sure which wideband you own. I have an AEM, and it's 5v output signal was quite noisy which made accurate datalogging difficult. I put a simple RC (resistor and capacitor) filter on the output prior to my datalogger - it really cleaned up the signal. See the attached image.


Great tip! I haven’t noticed noise issues in our data but we never used AEM widebands. A lot of people do and it’s probably not just AEM so definitely something to watch for.
  • HeavyRightFooT likes this
Instigator - Made a topic or post that inspired other Broken record - You are starting to sound like a broken record.

#8
HeavyRightFooT

HeavyRightFooT

    Member

  • Members
  • PipPipPip
  • 11 posts
  • Location:Bay Area
  • Region:California
  • Car Year:1999
  • Car Number:77

Do you have a way of logging the stock primary O2 sensor (switching sensor).  Theoretically, at idle it should be toggling back and forth ("switching") between 0.1 and 0.9 volts.  This would indicate the ECM is successfully controlling fueling to achieve 14.7 at idle.  Or, at least what it thinks is 14.7. 

 

If the Primary O2 was switching for your entire 15 minute idle period (indicating stoich) while the Wideband O2 continuously leaning out, then I would suspect an issue with either the primary O2 sensor or the wideband O2 sensor.  I tend to suspect the wideband sensor to be more likely the issue, since it also leans out on track when the ECM should theoretically be fueling in open loop mode.  

 

If the Primary O2 was not doing any switching at idle, then it is possible that the wideband was accurate and the car did lean out during the idle period.  Then you'll have to figure out why the car doesn't run in closed loop at idle.  This may or may not be related to the leaning out that you see on track.  

 

 

Side note, not sure which wideband you own.  I have an AEM, and it's 5v output signal was quite noisy which made accurate datalogging difficult.  I put a simple RC (resistor and capacitor) filter on the output prior to my datalogger - it really cleaned up the signal.  See the attached image.  

 

Unfortunately not equipped to log the signal of the primary o2, but did observe it via my obd2 scanner.

 

Started the car cold-ish and left it idle and warm up for about 40 min checking from time to time.

Car Idle dropped down to 850 a lot faster this time around. (Also its not that cold here so not as much to warm up).

Primary 02 was switching between 0.1 to 0.9 as expected. IAT and coolant temps looked accurate. ECU was trimming fuel accordingly and AFR as displayed on the wideband stayed hovering around 14.7 +-0.3  all the time after starting around 12... not appearing to drift up at all once the car was warmed up.

 

Usually it would start to drift up a bit at least in the mid 15 to 16 by then...  We'll see what it looks like next time I'm at the track end of the month.

 

Thanks for the tip for the RC filter to smooth out the signal ! Great one  :thumbsup: . I do use an AEM wideband and the signal is indeed quite noisy. 

 

Do you just solder the resistor in line and the capacitor to that and off to a ground and cover with heat shrink ? Any need for structural material to help it stay together ?

 

Thanks

 

Olivier



#9
speedengineer

speedengineer

    Member

  • Members
  • PipPipPip
  • 390 posts
  • Location:Michigan
  • Car Year:1999
  • Car Number:84

 

 

Thanks for the tip for the RC filter to smooth out the signal ! Great one  :thumbsup: . I do use an AEM wideband and the signal is indeed quite noisy. 

 

Do you just solder the resistor in line and the capacitor to that and off to a ground and cover with heat shrink ? Any need for structural material to help it stay together ?

 

Thanks

 

Olivier

 

Fortunately, my setup didn't require me to install the resistor and capacitor inline with the wire.  You could definitely do exactly as you described though.  Might be a bit messy.  I would definitely add some structural material/plastic/cardboard to keep that section of the wire from bending, and tape it up real well. 

 

Alternatively, perhaps it would be worthwhile to pick up a bit of protoboard?  Cut a small strip of it off and solder the components and wires into it, then tape/heat shrink the whole thing.  


  • Jim Drago likes this

Jason Kohler 

#84 SM

www.youtube.com/user/speedengineering

 

NASA Champs Winner - NASA Champs Winner We have a Winnah! - Won their 1st race... Congratulations! Bona fide - A bonafide Spec Miata driver

#10
Tom Hampton

Tom Hampton

    Egregious Member

  • SMembers
  • PipPipPipPipPip
  • 2,060 posts
  • Location:Mckinney, tx
  • Region:South west
  • Car Year:1992
  • Car Number:41

Great tip! I haven’t noticed noise issues in our data but we never used AEM widebands. A lot of people do and it’s probably not just AEM so definitely something to watch for.


Doubt it's the aem. I've run an AEM for 7+ years. But, I use shielded twisted pair cable for all my analog data inputs. Shield grounded at the data logger end, only.

Single ended sensors with chassis ground are a recipe for noise coupling.

-tch
Build: www.tomhampton.info

video: vimeo.com/tomhampton

Support: X-Factor Racing

 

I didn't lose, I just got outspent!

Beta-Tester - Assisted us with beta testing the website. Donor - Made PayPal donation Bona fide - A bonafide Spec Miata driver

#11
HeavyRightFooT

HeavyRightFooT

    Member

  • Members
  • PipPipPip
  • 11 posts
  • Location:Bay Area
  • Region:California
  • Car Year:1999
  • Car Number:77

Hi All !

 

Resurfacing this thread as I could use a little perspective on this issue and how it is evolving:

 

Since the last threads I replaced most of the remaining possible interfering bits including intake air temp, o2 sensors, coolant temp, ecu again, the EGR valve, evap purge solenoids, VICS solenoid, smoke tested the thing again with no changes or success. Fuel pressure is logged out of the regulator and is steady all the time.

 

During the last event ran the car with the hose coming from the EVAP system pinched to completely isolate it from the engine and no difference. 

 

More time on the dyno, shop, track, or troubleshooting the issue, has not surfaced any possible sources outside of the engine itself yet.

 

I have to start the car with close to 60psi of fuel pressure to be able to avoid leaning out to a dangerous level by the end of the races.

 

In addition to the leaning out issues I am now seeing the oil pressure drop down to 10-15 psi in heavy braking zones... glurps. Always kept the thing topped up. Not feeling great about how this whole thing is going... 

 

Because of this, and after talking with the builder again, I performed a leak down test yesterday.

 

Warmed up the car to operating temp and saw 9 to 10% leak across multiple cylinders, into the crank case. (All coming out of oil filler cap and vent by the oil filler cap).

To double check I retested it and saw only 3% across...

 

Thinking it was kind of weird that it would not seal better hot than cold I warmed up the car again for 30 min and tested it again fully hot and found again about 10% leak to the crank case on 3 different cylinders...

 

Not sure I understand how the rings would seal properly cold and mildly warm but not hot... was expecting the inverse. But my leaning out issue does only happen when fully hot on track.

 

Can excessive blow-by cause the engine to go lean ?

Guessing it's being re-circulated through the PCV valve into the intake manifold, and trough the vent tube on the valve cover close to the oil filler cap to the intake tube coming from the air filter. So all unmetered air going back into the intake ?

 

Engine only has 4 race weekends on it.

Any perspective appreciated.

 

Thanks 

 

O







Also tagged with one or more of these keywords: Air fuel ratio, lean, troubleshooting

0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users