Bank 2 misfiring at low revs Volvo 1999 S80 2.9L

Symptoms

Engine vibrates at low revs, up to about 1500, due to misfire. Diagnostics have fault codes to say that fuel trim can't be maintained at part load, and that oxygen sensors can't keep engine within spec. The front oxygen sensors (new) are balanced and within spec. The rear sensors read slightly high (0.6-0.8 cf 0.45-0.55 normal) but also balanced.

From a cold start, all cylinders run OK. As the engine warms up, bank 2 (cylinders 4,5,6) start to misfire. This seems to correspond to a drift in injection interval which ends up 1 msec longer (3.1 msec) on bank 2 than on bank 1 (2.1 msec). Removing the ignition coil from #5 has no effect on the engine idling. Removing #4 has a small effect, only #6 seems to be contributing from bank 2.

Things investigated by my dealer:
Swapped spark plugs
Swapped ignition coils
Checked coil wiring
Checked variable valve timing solenoid & wiring
Checked compression - all slightly low, but level and well within 10% variation
Cleaned throttle unit. Updated ETM software
Installed upgrade to ECM software
Checked wiring to oxygen sensors, and that they were not cross-wired
Checked ignition timings
Checked fuel pressure
Leak test on inlet manifold
Checked e-vac (don't know what this is)
Checked air mass meter - reading slightly high at 1400 kg vs 1200 spec, raising revs brings it on spec
Checked knock sensors

At this point, still no certainty as to whether the problem is mechanical or elecronic. Nothing (except the injector timings) shows any difference between the two halves of the engine.

One possibility might be the inlet valves on bank 2. Perhaps if the oxygen sensor on that side has been faulty for some time, an overlean mixture has partially burned the inlet valve seats. Not enough to effect a compression test, but enough to cause problems at idle. The solution would be to remove the cylinder head and re-grind the valves.

It's not clear to me why an inlet valve leakage would lead to the observed phenomena. The only thing I can think of is that mixture leakage means loss of power, so the management system would try to compensate by giving the injectors a longer duty cycle.

My dealer says he could try a more sophisticated cylinder leakage test using compressed air to see if bank 2 differed appreciably from bank 1. I've also heard that there are remote camera sensors that allow you to investigate the innards of a cylinder through the spark plug hole.

Another possibility is that there could be something obscure wrong with the ECU. Unfortunately you can't just drop in a substitute - it's tailored to the car.

Does anybody have any other ideas?




History

Lambda sensor warning light was on back in May 06 - dealer at service said it was a front oxygen sensor. Since the car had 75K on the clock I replaced both, using US OEM imports at less than half the price quoted by Volvo. I expected this to fix the sensor warning, and hoped it would also fix a cold-starting problem (lots of cranking before the engine caught) which had been there for several months.

But the warning light remained on, and the car remained hard to start. Inspection with scope showed the oxygen sensors working normally. Two spark plugs were very degraded (one of them #5), but with 30K miles on them so perhaps not surprising; replaced but to no effect. The car was driving fine, and I was going off on holiday, so no further work was done at the time.

Some hundred miles later the Engine System Service Urgent light started to come on, but with no apparent loss of driveability. About two months later the vibration started, at which point I took it to the dealer for diagnosis - see above.

A moral for me is - take the warnings seriously while it still might be cheap to fix! The car has had the lambda sensor warning popping on for a while every now and then ever since I got it, so I'd started to treat it as an intermittent nuisance rather than a genuine alert.

 

I had a very helpful suggestion on the Brickboard forum to investigate the injectors, and in particular to swap the sets bank 1/bank 2 to see if the problem moved. It did, and the culprit seems to be a faulty injector #5.

 

Thanks for the update.

Injectors are not common on the S80's to go bad.