efi Erratic Wideband O2 Sensor Output, Ideas?

[AustrianVespaGuy]

Hi guys,

 

I've been noticing that my Wideband sensor has been reading VERY erratically lately.  I haven't noticed that the car runs any different though and at idle it will smooth out quite a bit (see below).  Any ideas what might be causing this?  I've thought of a couple possibilities but can't decide which might be more or less likely:

 

1.) Vacuum or exhaust leak - haven't found anything visually or listening through a piece of hose, but will check tightness of all of the manifold connections tonight just in case.  Not really sure it that could actually cause this significant of a fluctuation in the readings though.

 

2.) Bad injector - already tested all four to make sure they clicked open with 12V and shot some carb cleaner through them, but didn't seem to make a difference.  If one of the solenoids is somehow operating too 'slowly' though, I think it could cause this, but no idea if that is a common failure mode for injectors or not.

 

3.) Bad O2 sensor or bad O2 controller.  The good idle readings make me think these are working fine, but again maybe this is a weird way that these things can start going bad.  Will do another free air calibration of the sensor, but I'm doubtful that will do very much.

 

Any other thoughts suggestions?  Thanks guys!

 

-Carl

 

[AustrianVespaGuy]

Oh, and it certainly didn't used to behave like this; here's a datalog from about a month ago for comparison (AFR readings are the yellow line up top, red line below):

[jimk]

Looks like you were just sitting still and stabbing the throttle.  The graph is full of transients and your transient settings do not reflect a good enough tune to handle it.

[AustrianVespaGuy]

Hi Jim,

 

Ok you're right I didn't take a very good snapshot there, but here's a better one where you can see that the AFR is doing the same thing during a very UN-transient period.  Again, comparing to where I was a month ago in Datalog_2 above (and no changes to the tune since then), there's definitely something a bit wonky going on here.

 

[M3M3]

i would first check O2 sensor calibration and then check all the WBO2 wiring.  particularly the grounds.  I have found the output to be very sensitive to proper voltage and grounds.  if you are using separate data aquisition system and wbo2 system they must share same power source and ground.  how does the gauge reading compare to the data output values?

 

then check for air leaks in intake system and exhaust (upstream from o2sensor)

[stuff]

Are o2 sensors sensitive to ignition noise?

[AustrianVespaGuy]

Hi Marshall! Trust you guys made it back up north OK this week?  This wonky O2 signal is the only trouble I had driving home. I only have the crappy digits gauge that came with the kit, so since it refreshes so slowly compared to the log I can't tell much from it, but the numbers are also jumping around so I guess you can say they agree.  Will double check all of the 12V/grounds when I re-calibrate tonight.  If I have a leak it's not a very big one; do you think a small leak could cause this much disturbance?

 

Colin, yes I do think they ARE sensitive to ignition noise, but in this case I can't see how any noise would have changed from a month ago since I didn't change the wiring or anything like that. . . 

 

Will keep trying things and let you guys know what happens!

 

-Carl

[ndog]

my afr did that after my steering arm clipped the data wires.

[AustrianVespaGuy]

Decided to venture away from the O2 sensor for awhile and instead check on the injectors.  I can here all of them working fine, but then I pulled the plugs and #3 looks MUCH worse than the others; could this be an indication of an injector that's not closing all of the way?  I'm having trouble discerning if this is fuel or oil deposits though.  Sure hope it's not oil since I'm only ~3500 miles on the freshly rebuilt engine. . .

 

 

Plug #2                                                Plug #3

 

-Carl

Edited May 31, 2014 by AustrianVespaGuy

[AustrianVespaGuy]

my afr did that after my steering arm clipped the data wires.

Just saw this Ndog; were you able to solve this by rewiring or did you need to replace the controller?  Thanks!

-Carl

[jimk]

Those specs look like aluminum.  Are there specs on other plugs?  If so, the timing table needs some work and confirm your sync between the crank and what the ECU is saying.

[ray_]

What type of plugs are you running and how did you choose them? Even though I know you've redone engine management, I'd stick with copper plugs on the M10 unless you have more educated advice.

 

GL,

[AustrianVespaGuy]

Nope, it's only that number 3 plug that looked at all bad; 1,2, and 4 are all just fine (maybe a tad rich) look the same as the #2 pic above.  They were new iridium plugs that I put in with the rebuild, so for now I'm going to chock this one up to a little extra gunk from the rebuild/break-in.  Swapped out all four for new copper plugs last night (I agree Ray, not really sure how I let myself stray from copper previously, but all better now) and will keep a close eye on that #3 for awhile.  Have all of the O2 wiring still *undone* and will redo today just in case, and then next step I guess will be to replace the O2 sensor.

-Carl

[AustrianVespaGuy]

SOLVED!! Ok well not actually solved yet, but I'm 100% sure that I've found the problem!  Even though my logs show reasonably smooth MAP and TPS inputs, I must be having some huge noise spikes in the TPS signal, because my TPS acceleration enrichment keeps kicking in randomly!  At idle, the TP: AE light will come on sometimes, pulse width jumps from ~2.6 to ~3.6ms, and then the AFR bounces around like crazy!  MUST be the issue.  Will check TPS sensor wiring next.

 

-Carl

[AustrianVespaGuy]

Crap, sorry to resurrect this but still having issues.  I rechecked all wiring, made sure all sensor grounds come back to one common pin on the MS board, and installed a large capacitor across the supply voltage to both MS and the Innovate controller to make sure they don't see a lot of noise on the supply side.  All sensor signals look really good now except for the wideband O2.  Interestingly, it stays pretty stable at idle and while cruising, but seems to go all crazy at higher pulse widths.  So now I'm thinking it could be either injector(s) or maybe fuel pressure?  New datalog attached.

-Carl