[lox]

Team,

I'm having an issue with a persistent and very intense misfire. Story time...

• Car ran sweet, but had some faults in ISTA regarding O2 sensor line disconnections. Went to investigate and noticed the pre-cat sensor wires had chafed through. When I moved the wire to look closer, they split the rest of the way, so all 5 wires were split clean open and touching each other. The post-cat O2 sensor had a fault too, but the connector/wires/sensor looked fine.
• After unplugging both O2s and restarting, the car ran like a dog with multiple cylinder misfire and fuel injection disabled codes, which I figured was due to it not getting a signal from the O2 sensors and thus cannot calculate proper fuel trim.
• x2 new O2 sensors ordered, fitted, and adaptations reset. No more codes from the pre-cat, but still a line disconnection code from the post-cat (WTF, it's brand new). Misfire still very much prevalent, & long term fuel trims sat at over 70%.
• Continuity of the wiring between the O2 sensor all the way back to the DME pins confirmed, no opens or shorts to eachother/GND/B+.
• Coil and plugs swapped from cylinders with persistent misfires to rule out plugs and coils.
• Haven't removed the injectors to swap and test yet, however using my stethoscope I can hear them all firing properly (before they get shutoff), resistance is good across their pins, and using my oscilloscope shows a very normal looking waveform but with a very long pulse width modulation time - likely due to the strangely high fuel trim.
• I'm assuming the actual misfire is being caused by excessive fuel in the cylinders, nearly all plugs are soaked after a run.
• Additionally, I have a valvetronic servomotor fault that's been there since before the misfire. Replaced the servo motor as wiring between the servomotor and the DME pins was confirmed - no opens or shorts to eachother/GND/B+. After replacing, the fault was still present.
• MAF/MAP/IAT running values verified and all are as expected.
• Haven't checked for air leaks yet. However worth noting that despite my PCV/CCV valve seeming fine, there is quite a big vacuum when trying to open the oil cap. Unsure if it's normal on this engine.
• i-step software update performed on the DME to rule out software, no change.

These faults and my findings are slowly leading me to diagnose a faulty DME (possibly shorted or damaged somehow by the O2 sensor wiring fault?), as it's complaining about x2 faulty elements (O2 sensor and valvetronic) that don't appear to be faulty and neither do their wiring back to the pins on the DME itself - not to mention the oddly high fuel trims.

Any help or advice appreciated.

Thanks,

Lox

[rtl32]

Finally a case that isn't just another random asking why their newly stage 2 car has terrible timing on the original plugs.

Oil cap vaccum is normal. Mine sucks it back out of my hand with confirmed PCV function and spotless smoke and boost tests.

The car is literally telling you it's not getting the proper data from the oxygen sensor/s. Don't look for extraneous issues when you began having issues attributed to that direct issue. The disconnect code and screwed up trims are because of this. Running without proper O2 sensor signals is like trying to drive around a track blind by feeling where the curbs are through your steering wheel. The likelihood of another massive failure happening at the same time is low to none unless it is because you ran into the curb because you were blind. You would likely have other codes if it was an organic DME fault. However the valvetronic code does sound like something - especially since it didn't get fixed with the new motor. But those are different systems.

Please post the codes and a log. Including oxygen sensor voltages on the log.

Does your exhaust smell like straight gas?


From simple to less likely:

Did you get the right sensors? Are they installed in the DP in the proper orientation front/back? Are you sure you didn't mix up the blue and black connector and connect them backwards on accident?

Sounds like your plugs have had a hard life lately. Maybe they got totally soaked out, could explain the misfires. Put some new ones in and see if that changes everything - swapping them around won't do much when you're having problems across the board.

Do you have confirmation that the valvetronic mechanics are actually able to move and are functioning? The car commanding changes but it mechanically not happening could explain somewhat.

What are you comparing the pulse width to, in order to get to the conclusion that it's long? The DME is so in tune with those things if that was an issue (and not a symptom of another issue) it would throw a code.

[lox]

Originally Posted by rtl32

Finally a case that isn't just another random asking why their newly stage 2 car has terrible timing on the original plugs.

Oil cap vaccum is normal. Mine sucks it back out of my hand with confirmed PCV function and spotless smoke and boost tests.

The car is literally telling you it's not getting the proper data from the oxygen sensor/s. Don't look for extraneous issues when you began having issues attributed to that direct issue. The disconnect code and screwed up trims are because of this. Running without proper O2 sensor signals is like trying to drive around a track blind by feeling where the curbs are through your steering wheel. The likelihood of another massive failure happening at the same time is low to none unless it is because you ran into the curb because you were blind. You would likely have other codes if it was an organic DME fault. However the valvetronic code does sound like something - especially since it didn't get fixed with the new motor. But those are different systems.

Please post the codes and a log. Including oxygen sensor voltages on the log.

Does your exhaust smell like straight gas?


From simple to less likely:

Did you get the right sensors? Are they installed in the DP in the proper orientation front/back? Are you sure you didn't mix up the blue and black connector and connect them backwards on accident?

Sounds like your plugs have had a hard life lately. Maybe they got totally soaked out, could explain the misfires. Put some new ones in and see if that changes everything - swapping them around won't do much when you're having problems across the board.

Do you have confirmation that the valvetronic mechanics are actually able to move and are functioning? The car commanding changes but it mechanically not happening could explain somewhat.

What are you comparing the pulse width to, in order to get to the conclusion that it's long? The DME is so in tune with those things if that was an issue (and not a symptom of another issue) it would throw a code.

Hey rtl32,

Thanks for the reply man, really appreciate the guidance. To respond to a few of your points/questions...

• O2 sensor disconnect code is now gone after reseating the connector for the post-cat sensor (it does seem to be intermittent, but given it's the post-cat which is only used for emissions and not fuelling, not too worried for now).
• They're deffo installed the right way around and connected to the right connector, 5 wire wideband is pre-cat, 4 wire narrowband is post-cat.
• Exhaust smells fairly rich until misfire detection starts shutting off injection to multiple cylinders (as expected).
• During running I read out a load of pertinent data from the DME in ISTA to see if anything looked out of place (temps, pressures, etc), but I couldn't spot much that concerned me... see what you think?
  
  IAT: 16oc (same as ambient temp outside)
  Ambient Pressure: 1014 hPa (atmospheric pressure)
  MAF: 40.5 kg/h (seems fairly standard)
  MAP: 531 hPA (vacuum)
  Charge Pressure: 1025 hPa (post turbo, pre manifold, slight pressure as expected)
  Valvetronic Angle Actual: 179o (no idea, lol)
  Valvetronic Setpoint: 183o (no idea, double lol)
  Rail Pressure 14.36 MPa (looks good)
  O2 Sensor Pre-cat Voltage: 1.64v (I think this is saying we are rich, from memory these wideband sensors should report between 2.2 and 2.8 if at lamda 1. Less than 2.2 says rich, more than 2.8 says lean, but I could be wrong)
  O2 Sensor Pre-cat Air Ratio Value: 1.27 (no idea, triple lol)
  
  
• Point taken on the plugs, they have had a hard time and don't look or smell great. I'll get some new ones in but I do fear they'll foul again pretty quickly!
• I'm comparing PWM duration on the injectors and defining as long only by limited experience with Bosch piezo injectors. Not suspecting the injectors themselves are at fault, but more that the longer PWM duration is a symptom of something else and could be over fuelling the engine for idle conditions and soaking the cylinders and plugs, thus causing misfire.
• Valvetronic motor and eccentric shaft are both moving freely as far as I could feel (they needed to be for me to install the new motor), however the fault is still weird given it's still not happy even after a new servo motor and confirmed wiring continuity back to the DME - this is what makes me wonder if it could be a DME issue. I confirmed that the valvetronic supply from the IVM is supplying 12v to the valvetronic system too.

I'm leaning towards either a partially fried DME, or that the valvetronic fault is somehow choking the engine (and this may also be caused by DME issues given every other component in the valvetronic system bar the DME has been ruled out).

As ever, any assistance or suggestions are greatly appreciated!

Luke

[aerobod]

Is the replacement pre-cat Lambda sensor an LSU 4.9 or LSU ADV? The ISTA provides output mapping for both, but they have a different relationship between Lambda value and voltage.

If the voltage is measured relative to the sensor output as opposed to ECU interpretation (often the ECU will have it's own 0 to 5V range that is normalised to the Lambda Sensor range), then Lambda 1.00 is 1.5V for the Bosch sensors, whether 4.9 or ADV. If it uses a normalised ECU range for voltage, then 2.5V is usually mapoped to Lambda 1.00. This would indicate you are running lean if it is the sensor output voltage.

Here are the Bosch datasheets for each sensor:
https://www.bosch-motorsport.com/con...or_LSU_4.9.pdf
https://www.bosch-motorsport.com/con..._pre_Turbo.pdf

[lox]

Originally Posted by aerobod

Is the replacement pre-cat Lambda sensor an LSU 4.9 or LSU ADV? The ISTA provides output mapping for both, but they have a different relationship between Lambda value and voltage.

If the voltage is measured relative to the sensor output as opposed to ECU interpretation (often the ECU will have it's own 0 to 5V range that is normalised to the Lambda Sensor range), then Lambda 1.00 is 1.5V for the Bosch sensors, whether 4.9 or ADV. If it uses a normalised ECU range for voltage, then 2.5V is usually mapoped to Lambda 1.00. This would indicate you are running lean if it is the sensor output voltage.

Here are the Bosch datasheets for each sensor:
https://www.bosch-motorsport.com/con...or_LSU_4.9.pdf
https://www.bosch-motorsport.com/con..._pre_Turbo.pdf

Thanks for the reply and links to those useful documents, aerobod. The pre-cat is an LSU ADV, and judging on the chart in that document and my recorded voltages in ISTA (1.65v) I'm running around lambda 1.1 which is a lean AFR of 16, not rich as I'd guessed.

Oddly, ISTA reports the air fuel value from the pre-cat as lambda 1.3, which doesn't add up, but anyways. The DME also reports a mixture adaptation value of 1.0 (no adaptation required in either direction) surely meaning it thinks its stoich? But if we are truly lean then shouldn't the adaptation value be 0.something?

rtl32 / aerobod - any idea what's going on here fellas?

Thanks in advance for your help...

[aerobod]

Next I would re-check the plugs carefully for any mismatch in colour, then look towards a possible faulty injector, bad plug or faulty coil in any cylinder where the plug doesn’t match the rest, as the Lambda values aren’t necessarily as consistent as expected. I know you have already done some of this, but have you eliminated the ignition by having the misfires stay with a given cylinder after moving the plugs and coils?

From a logging perspective I would see if any individual cylinder values (such as timing) are also different on any one cylinder.

It is also worth borrowing or buying a cheap fibre optic camera if you don’t have one, to have a look around inside of the cylinders and look for any signs of pre-ignition on the piston crowns. There are quite a few on Amazon for around £50: https://www.amazon.co.uk/Inspection-...ode=5136303031

[lox]

Thanks for your help guys. Oddly after blow torching and cleaning the plugs, letting the cylinders dry out, and putting everything back it seems to now be on 6 cylinders, although the rough running figures in ISTA do show a bit of unevenness, which I can physically feel too - mainly on cylinder 1 and 4.

The coils on cylinder 1 and 4 are showing a lower amp draw using the amp clamp and oscilloscope (despite resistance of the coil and plug looking good, and ground continuity for the coil and plug via the block also looking good) so I've gone ahead and ordered 6 new plugs and coils. I'll drop an update in when the new plugs and coils are installed and tested. Not sure why the fault was previously not moving cylinders when swapping plugs and coils over, which is normally the go to method for troubleshooting.

I've removed injector 4 (possible bad), and injector 3 (seems good) to have them bench tested at a facility to help rule out faulty injector.

Valvetronic issue appears to be unrelated, looking back through ISTA history logs it was there before the misfire. Raised a separate thread for that: https://www.2addicts.com/forums/show...2#post28748692