Erratic lambda

[lezsi]

I've got a strange wbo2 issue on a car recently.

It's a BMW M44 turbo engine, and was running fine for a year on VEMS (1.1.26)
Recently the engine was disassembled and the whole thing rebuilded mechanically (not touching VEMS at all), and since then, lambda values are erratic especially at low load and idle. The engine runs fine, doesn't misses or so.


It seems like fluctuating around the real value.
datalog: http://quasar.dynaweb.hu/~lezsi/bmw/v3.3_n001250-2010.09.19-11.39.09.vemslog

I've checked that my wideband config settings look normal.

- I've gone through firmware versions: 1.1.27, 1.1.78, 1.1.81
- Replaced the VEMS unit to another good-working one
- Replaced the wideband sensor to another good-working one
- measured that all wbo2 wiring has good contact and has no short-circuit to each other

Still, no results yet.
 

[z0tya]

Maybe watch battery voltage. There are 0.3-0.5 V glitches...

[[email protected]]

I find it strange that someone would change firmware versions because the one variable that has changed is the engine.
A sticking valve, mis-timed or different cam, bad valve clearance might cause a problem.
A bad or badly connected alternator, poor engine ground, failing injector, intermittent spark, failing sparkplug could also cause a problem.
Fuel pressure fluctuation, loose vacuum hose, air leak would possibly cause an issue.

[lezsi]

Maybe watch battery voltage. There are 0.3-0.5 V glitches...

Thanks for your look, this is a possible problem on this car, but as I look into the logs made years before (on this car), it was always doing this voltage fluctuation, and there was no problem to the lambda at that time.

[lezsi]

I find it strange that someone would change firmware versions because the one variable that has changed is the engine.
A sticking valve, mis-timed or different cam, bad valve clearance might cause a problem.
A bad or badly connected alternator, poor engine ground, failing injector, intermittent spark, failing sparkplug could also cause a problem.
Fuel pressure fluctuation, loose vacuum hose, air leak would possibly cause an issue.

Thanks Rob, the firmware change is not related to this problem, but winter starting problems on E85 using older fw.

Now I can't say that there are no mechanical issues here, but otherwise the car is running fine, stable idle, strong vacuum, good power, no serious missfires. The only problem is a bit high fuel consumption which is 18-20L of E85 in city. But as long as I cannot really tune it to nice lean lambda values, this might be normal.

I looked at this car's lambda for quite a few hours by now, and usually at idle the lambda looked like a sine-wave, averaging the (probably) real lambda value, with the period time of ~1s. This is not really visible on the log I posted.

[mattias]

This sounds almost like the problem I'm having with an Audi (AAN), using 1.1.81 firmware. It runs fine but has too high HC values at idle as tested in a 5-gas analyzer. Despite that, it gets great fuel economy (better than stock) and puts a smile on the owners face. Still this has to be solved to pass annual inspection. Like Rob said, there could be many reasons in your case and this one.

[lezsi]

This problem is now resolved by a replacement lambda probe.

I know it's lame, especially because we started the problem solving with replacing the sensor with a known-good one (a few years old, a bit sluggish but good working one).

No matter, it is operating fine with an almost-new sensor, hope it keeps for a while...

[mattias]

Had to mention that the Audi I was having problems with had a dead catalytic converter, the metallic "web" was disintegrated at the back of the unit.  This happened before the installation of VEMS PnP, as the owner tried to get help from a local "chip tuner" which made the car run very rich and just bad - ultimately leading to me helping the poor guy out.

[AVP]

it would be nice to have an option in the future that VEMS would be able to let you know if a problem is apparent with any of the sensors, similar to what OEM ECUs do with bringing the car into safe mode when something doesnt look right.

Say the wideband is reading off or the voltage is out of bounds, it could bring in a notice where the problem is coming from, so that we have an idea of where the problem is.

I had a torn vaccum hose and basically a boost leak which i couldnt detect other than having 200rpm later boost, and the car was still boosting as high as it should and mixtures were perfect.