12 replies to this topic

[rik]

Hello. Got the engine running yesterday (its been sitting for 5 years!) and have been doing some checks today, getting it up to temperature etc..

For the first few minutes the AFR on the wideband is good, a little lean 14.8/9, but after a few minutes it goes full rich, 11 or so

Have made a log of it here https://ufile.io/wvj9u and the map is here https://ufile.io/z3yz0 and also screenshot below.

I have made the conclusion that the primary lambda is faulty since it drops below 0.5v suddenly (for no reason, I didn't touch anything), indicating lean, but both the secondary and wideband then indicate rich. My WB sensor is linked in to the EGR pin, and to get the actual afr the formula is AFR=(V*2) + 9 ... so lower voltage means rich

Please let me know any thoughts before I go and order a new primary lambda

 

[Exmantaa]

Without studying your log file, the ecu controles the mixture from primary O2 sensor. If that one goes lean, the ecu will enrichen the mixture.

So that is exactly what you are seeing; a faulty primary O2 sensor and the ecu just reacting to the (incorrect) lean signal; making it pig rich...

 

Edit, now seeing the log file and pic =>

Yup, at ~480s your primary O2 sensor somehow stops working.

Edited by Exmantaa, 06 July 2018 - 08:22 PM.

[jimn]

If that bottom pic is wide band vs primary lambda it looks like it’s just gone open loop at idle for whatever reason. Might be dodgy primary lambda might be something else.

Are you running a tmap load reading? What’s that logging?

From the log it looks like the car is quite happily running closed loop then shits itself and goes closed loop and rich.
Do you have any fault codes?

I’m not 100% but perhaps what the primary is telling you is correct. Wide bands and linear lambdas run different voltages and they both change at the same time. If they both correlate to the same afr (which it looks could be true) the ecu has decided to go closed loop safe mode because of something🤷🏼‍♂️. Air leak or similar?

Can you log fuel trims?

[jimn]

Sorry. Just read that back and it makes no sense..
The second 2 closed loop statements should read open loop.
..it looks like the ecu seems to want to fcuk the lambda off and go to look up/limp settings for safety.

(I can’t read your logs on my phone, just looking at the screen shots you’ve posted)

I should really have a look at the log on my computer to make myself more helpful.

Edited by jimn, 06 July 2018 - 11:09 PM.

[rik]

top graph is wb, goes from ok to rich

bottom graph is primary lambda (black), goes from ok to lean and secondary lambda (red) goes from lean (i guess it makes sense since its post cat) to very rich

 

map sensor is reading 47-50 kpa.. seems a little high for idle? would expect around 20, so could be an intake air leak. If it still happens with new lambda sensor i'll look there

 

no error codes

 

 

[Exmantaa]

Nothing to do with map sensor values. It's in closed loop mode and that is dictated by the primary O2 sensor...

Check primary wiring/connector, if not then replace. (with a decent one)

[siztenboots]

looks like air leak, cracked exhaust maybe

[vocky]

NGK Lambda Sensor OZA531-GM6 https://www.sparkplu...nsor-oza531-gm6

 

exactly the same part as OEM

[rik]

Found an old (but new) universal one in the Attic, fitted and seems to have solved it! Vocky, I've never tried an NGK one, do you rate them?

[vocky]

Ngk supplied the Oem z22se sensors


 

[oakmere]

I use Ngk NTK sensor with OBD Tuner. As Vocky says same as original.

[siztenboots]

lambda heater circuit failure?

[Arno]

Found an old (but new) universal one in the Attic, fitted and seems to have solved it! Vocky, I've never tried an NGK one, do you rate them?

 

From the graph you can already see that the narrowband lambda is wonky or something is up as it's hitting 0.0V which should never happen if it's running normally.

 

Narrowbands should really cycle between 0.2 min and 0.8V max with some ECU's tightening control to switch points around 0.4 and 0.6V.

 

Yours looks lke it was already offset downward and dropping to 0V output voltage.

 

Indeed, as concluded the ECU will at some point decide that the primary lambda is bad and switch to open-loop (and go rich) for safety.

 

An exhaust leak upstream of the primary lambda can also cause issues as the signal also gets 'shifted' because of extra oxygen that gets drawn in. Although ususally not down to 0V

 

Bye, Arno.