O2 sensor failure modes?
 

Hi folks,

Marvin decided to take a funny turn and run like a bag of nails, one of the codes I'm getting if for the O2 sensor, a code 21, which on this model year appears to mean that the O2 voltage value is stuck in the middle somewhere.

I'm thinking "what the hell?, don't they just fail lean???" so this seems highly bizarre. There's other codes for constant rich or constant lean. I realise of course that this might not be the O2 sensor per se but some other fault, particularly as I'm getting a code for the speed/distance sensor at the same time.

Some sources say for 90 up the code means O2 sensor "stuck at value" and no mention of whether it's middling high or low. Either way, seems very odd I'm not getting a code for constant lean or constant rich as well if that's the case.

Really don't have much of a clue what's going on yet, I'm suspecting an issue in the speed/distance sensor circuit that shares a signal ground with the injector timing, TPS etc, that is taking out the SDS and buggering up the fuelling.

Any clues greatly appreciated.

Road Warrior

Years ago I had a similar problem on the BWH. If I remember correctly it was throwing a TEMP, MAP, BARO, and TPS codes all at the same time, and they were all reading low. It took the longest time to figure it out. Turns that my temp sensor was shorted to ground internally, and that was throwing everything else on the 5v reference off. After the temp sensor was replaced I had no further problems. You may want to check and see if all those sensors share some bit of wire in common, and make sure that is intact and not shorted to ground. Then check all the sensors with a VOM and make sure none of them are shorted internally.

-Jay

Thanks that's one of the possibilities I was figuring. I've got one ground loop that goes to a bunch of sensors, the O2 doesn't appear to be on it though unless it's internally connected in the ECU to a common signal ground rail. I don't know if that signal ground should float off chassis ground or not though.

Annoying thing is, that the wiring harness is so faded it's hard to tell colors, and with it being 20 years old, I'm worried about causing more issues digging into it.

You may want to do what I did on the Buick.... I had a suspicion that it was the TEMP sensor (Can't remember why now), so I unplugged it, and all the other codes disappeared. When that happened I replaced it. Maybe try unplugging sensors 1 at a time?

-Jay

Well I unplugged the coolant temp sensor to force it into open loop and it ran a tad smoother but not much, made it a bit more driveable, less stumble. a little more power.

What does your exhaust smell like when the engine is being revved up and it's stumbling?

Well being without a helper, can only go sniff it at idle, smells normalish, can't smell coolant or oil, (Unless I get it hot enough it seeping past the valveguides after 20 mins of hot idling, then it gets a bit blue but that's kinda "normal") of course it smells faintly of burning ATF since I put a couple of ounces in the tank with a bunch of IPA to see if it improved it if it was a clogged injector.

No bubbles in the coolant, and not losing any significant amount, meaning it might be a couple of ml down every time I run it for a while, but not the cupful you'd expect if it was squirting enough in a cyl to cause problems. It's been needing about a pint every 3 months/7000 miles or so, so there's some seepage somewhere but doesn't seem enough to be able to find or fix, or cause problems.

I wonder if the ATF fuel additive could be messing with the O2 sensor, coating it with residue that prevents it from reading.

I was getting the code before I put it in.

See if you can find in your service manual where to check the O2 output. On my 88 Toyota you could check two contacts on the service port and it would give you a 0-5v reading that told you what the sensor was doing.