Is my wideband sensor dying?

[manga_blue]

Every time I go to WOT then a second or so later my wideband reading jumps to AFR 22+. The car is still making good power and feels healthy, even at these extreme readings. There's nothing really to indicate anything wrong with the motor except a slight flatness above about 6000.

The usual suspects for this condition are either an ignition miss (which dumps unburnt fuel and unused oxygen down the pipe), a lean fuel miss (which dumps just unused oxygen down the pipe) or an exhaust leak which allows entry of extra oxygen upstream of the sensor. I think I've eliminated all of these. I'm pretty sure it's just a stupid reading error. Is this what a dying wideband looks like?

It's a Bosch LSU 4.2 running through an Innovate LC-1 and an SSI-4. The sensor sits downstream from the last Y in a 4-2-1 system and is correctly pointing up hill. I get the same results with and without the SSI-4. It's all purely there for logging purposes at the moment. It doesn't contribute directly to engine management. I just need it for tuning.

This trace is pretty typical. It's a pull through 2nd and 3rd at WOT. Black is RPM, red is TPS, purple is AFR.



This has been getting steadily worse for a year or two, i.e. the lag time between going WOT and 22 AFR is getting shorter and shorter. I've:
- cleaned and checked injectors and injector loom
- replaced the Toyota COP coilpacks: twice
- beefed up power supply to the COPs
- had the exhaust system pressure and smoke checked for leaks upstream from the sensor
- recalibrated the LC-1

[rascal]

Phil, If you hadn't added the last line I would have picked this as your problem.

Mine had been running fine for 18mths without any issue up till last meet when suddenly it started doing similar, reading 22.4 more often than not. (without any interruption to power levels)

I removed the sensor, and fresh air calibrated it, and put it back in and now its working perfectly again.

Can you borrow a known working sensor off someone to confirm it is/isnt your sensor?

[manga_blue]

Thanks Russell, I'll try re-re-calibrating first then. Maybe I got the last one wrong. I just spoke to Peter at wbo2: it's not a typical sensor failure but it's possible, as a controller failure is also possible.

BTW I don't remove the sensor to re-calibrate any more, it's too easy to twist and damage the wiring where I have it. I just remove #1 plug, turn the motor until #1 exhaust is open and stick a vacuum cleaner hose in the exhaust pipe for a while.

[beavis]

I've read you should re-calibrate after every race meet.

[zossy1]

I've read you should re-calibrate after every race meet.

That's probably overkill but LC1s are notorious for going out of calibration.

[gslender]

Also, even widebands will read incorrectly for too much fuel or too lean. Takes a lot, but can be both.

[manga_blue]

Re-calibrated but found damage to the airbox along the way. Can't try starting it for a day or two until all the adhesive sets.

[manga_blue]

No luck.

I did a full "first time" re-calibration: heater plus free air. The resulting readings are pretty much the same as before except that the lag between going to WOT and the sensor going to max lean has increased to around 2.5 secs, but then the lag between going off-throttle and the sensor returning to normal readings has got much longer as well. I also did a run at half throttle and the lag was over 4 secs.

Could it be that the LSU4.2 sensor is experiencing heat failure under the extra heat and flow? Maximum lean reading on a Bosch wideband means over 5V output, does't it? Seems weird that it would fail by putting out higher voltage rather than no voltage though ...

The wbo2 people say replace the sensor, then replace the controller if that doesn't work. Anyone got any other ideas before I replace the sensor?

[Trackphotos]

Anyone got any other ideas before I replace the sensor?

Borrow one from someone local to prove it first?

[zossy1]

I agree with starting at the sensor. Thats what I'd do.