Plugs and a "bank too lean" code adventure.

[ishoottrap]

Living on the Edge

A couple weeks ago we went to Canada and took my wife’s 2012 Edge.  I drove and let’s just say I’m a bit more spirited of a driver than she is.  First car I passed, it missed pretty severely.  Runs fine until very near full throttle then begins to miss.  I figured it might be the plugs, I've had that happen before when they get wide enough and with 121k miles plus that was the general wisdom of the net…..  Anyway, I changed those with little fanfare and while I had the intake bits apart, I thought I might as well clean the MAF sensor which had caused me grief on our ’97 Expedition at around 130k.  That one is pictured below labeled “97 sensor” and has the important parts exposed and easy to clean with a squirt of electronic parts cleaner.  I could see a layer of fine particles on the upstream side of both parts before cleaning and afterwards, clean as a whistle.  This one, however is pictured below as “12 sensor” with the important bits buried somewhere in a passageway.  I gave it a quick blow out, couldn’t really see anything and was afraid to use the air compressor so just used my breath.

 

Now, changing the plugs went swimmingly, got everything back together and went for a few miles up I90 towards Snoqualmie pass with quite a bit of full throttle driving and it ran like a top.  All was right with the world until the following day when my wife drove it.  That night, she told me it made a zzzzzzz sound and the little engine light came on.  I pulled it into the garage, popped the hood and discovered I had stuck one of the 5 vacuum hoses on its fitting but not finished pushing it on so it was only engaged about ¼” out of 2” and the clamp was nowhere near the fitting.  Ha!  I put it back in place and checked the others which were fine, cleared the code (both cylinder banks too lean), stuck out my chest and told my wife how brilliant I was despite the fact that I had screwed it up the previous day.

 

Things were fine for a day or two and then I got the same report, zzzzzzz sound and the engine light came on.  Hmmm, I took it for a spin and occasionally, under very moderate acceleration I could get it to exhibit pre-ignition knock and sure enough, after a bit of this the check engine light would come on.  Crap, has to be a vacuum leak I must have missed something.  I tore it all back down to the point of taking the intake manifold plenum off again.  Gasket was perfect, couldn’t find anything out of sorts, put it all back together and even went so far as to look up the torque on the manifold bolts rather than just winging it.  No change. 

 

I should probably mention that I had been using my handy dandy $20 OBDII blue tooth adapter and $1 analysis application to look at various parameters.  Chief among them were “Air Fuel Ratio Commanded” and “Air Fuel Ratio Measured”  These bounce around between approximately 14.4 and 15.0 and were relatively close to one another for most of the test drive.  Nothing really stuck out so I poked around a bit on the net searching for the code that the OBDII displays mapping to “bank too lean”.  One posting mentioned trim values.  The injection system control is based primarily on the MAF and then “tweaked” if you will by the O2 sensors.  There are 4 readings of interest, 2 for each bank, long term trim and short term trim.  I found them in the sea of variables and displayed them on the way to work today.  See Fig 1, at idle they were as shown (long term trims on the top, short term trims next going down then air fuel ratio commanded and measured and finally, MAF and manifold vacuum.

The long term trims appear to max out at 29.7, positive is more fuel, negative is less.  This display shows that at idle, the AFR is basically maintained by boosting the long term trim to the max and then adding a bit of short term trim.  Idle MAF is a pretty constant 2.1 grams/second.

 

In my experience, running anything constantly at the rail is never a good sign.  Here’s an example of steady speed  where the trim is slightly off max (see Fig 2).  When I got home, I experimented a little with the various vacuum line connections.  I took them off and plugged the hole or clamped them shut with absolutely no effect on the idle readings shown.  I finally took the MAF sensor out and using my air compressor nozzle from a foot or so ways, I shot a little air into all the openings on the sensor.  After re-installing, the difference was instantly notable (see Fig 3):

The MAF had jumped more than 50% and the short term trims were much lower.  We went to dinner and on the way back, cruising at constant speed I saw (see Fig 4):

Both short and long term trims were nearly zero and no hint of pre-ignition knock ever.   While I should probably give it a day or two before proclaiming victory, I’m thinking that $21 was probably one of my better investments.

Edited December 6, 2019 by ishoottrap
Add tags

[omar302]

Welcome to the Forums. A very long first post ?. Hope it remains fixed. I don't remember a similar issue being reported before due to the MAF Sensor or cleaning it. There is always a first time (or my memory is is not good).

[ishoottrap]

Yeah, I was convinced it wasn't the MAF but as the saying goes, when you have eliminated the impossible, whatever remains, however improbable, must be the truth.  I must have dislodged some crud which ended up in a bad place with my first attempt at "cleaning" the sensor and then a little stronger blast dislodged it.  I have a new sensor on the way, I'll probably go ahead and install it but the current sensor seems to be behaving.