The AUSubaru.com.au Forums

Hi all, this is my first post, hoping someone here can shed some light on this strange fault with my 98 outback. I've only recently purchased the car, it has 203k km on the clock.

The fault is that after accelerating to 100km/h, then sticking at 100 and cruising, the Check Engine light comes on, and stores an Oxygen sensor fault. The sensor is a very new Denso unit, it was changed just before I bought the car. The light will stay on until you back off the throttle for a bit and then it seems ok, until you stick the boot in again and then cruise.

I have stuck my mates wideband oxygen meter on it, and what appears to happen is that after the mixture goes rich on booting it, it "gets stuck" and doesn't go back into closed loop 14.7 af ratio, instead it hovers around 12, and the computer seems to stay open loop and slowly work back to around 14.7af, but in open loop. Watching the voltage output of the ecu oxygen sensor, it sits around .3v and slowly comes up to around .5v. To get it working properly again you need to take your foot off the gas for a bit, let it go lean, then it seems ok.

So it appears that the oxy sensor is working, but something goes amiss after a period of full throttle.

Any suggestions would be much appreciated, it's a strange one. I'm going to log the voltages coming from the TPS and the MAF and see what they are doing.

Thanks.

Adam

I would probally look at the M.A.F sensor as they can cause funny things to happen to mixtures. Are you using a wet(oiled eg k&N) or dry(paper) aircleaner. Subys maf sensers dont like wet aircleaners as it leaves residue on the heating element in the maf senser. Try cleaning with carby cleaner and replace element with paper one.

I'm pretty sure it's a dry filter, but I will definitely check the MAF sensor. Thanks.

Yeah i would agree above - I had that problem in the past due driving in the tunnels all the time where there is diesel fumes are and leaves diesel oil on my filter element and was on the MAF sensor - used metho to clean it (Poured metho on it - didnt touch it with anything and i wouldnt)..
Cheers
AP

Well we just took it for a spin with the logging pc connected, reading the TPS, MAF, ECU narrowband O2 and the external wideband O2. The only thing that looked like it was throwing bad values was the ECU narrowband sensor, just as the fault code 32 from the ecu is indicating. When you floor it, the narrowband sensor seems to die for a while, it just flatlines, and stops mimicking the wideband. Then when you close the throttle for a bit, it seems to come good again. It's almost like the sensor is being contaminated after it sees the richer mixture on full throttle. We think it also went lazy once after seeing lean mixtures on a period of closed throttle too, but this didn't seem to happen much.

I guess I'll stick a new oxy sensor in it and see how it goes. The engine had just had new head gaskets, so I'm wondering if the sensor has been poisoned by an adhesive or sealer that he used when he rebuilt it.

Does anyone know of a cheap replacement sensor I can use in this? The genuine one is $380, a bosch one that a local mechanic assures me works fine is $150. I guess thats not too bad.

Thanks for the suggestions. I'm pretty sure the MAF sensor is ok after that test, it's values looked fine during the fault condition.

Problem seems to be fixed. The bosch sensor is working fine so far. Hopefully it will stay that way.

Hope this thread is useful one day for someone else with this problem.

Yeah I have the same problem,I think iwill bite the bullet and get one too.

The sensor I bought came from a petroject parts dealer. I think they just put the right connector on a bosch sensor. Unfortunately mine was wired incorrectly and I had to swap the pins around in the connector. Let me know if you want the petroject part number.