O2 Sensor Problems

[mwsku83]

My wife drives a 1999 Pontiac Grand Am GT (3.4L V6) and, luckily, hadn't had any major problems with it since she purchased it in 2002.

Then things got ugly.

Back in the summer of 2006 it started drinking coolant. I tore the engine down and took the head in to get milled and replaced all the gaskets from the head gasket up. No more problems that year.

Then in the summer of last year my wife and I drove it to Wisconsin for a friend's wedding and noticed poor gas mileage during the trip. The car ran fine the whole way, but after the rehearsal and dinner I was driving back to my friend's house and the car starting surging and stalled while at a red light. I started it back up and it did the same thing...surging (or "searching" as some might say) at idle, sometimes RPMs dropping so low it would stall. I limped the rest of the way with some fancy footwork to keep the RPMs up at stoplights and decided to look into it the next morning. The next day my friend read the DTC causing the MIL/SES lamp to come on and it said something about an O2 sensor (can't remember what the code or description was now, just that it was O2 sensor related). I started it up and it ran and idled fine, so I decided to risk the 300 mile drive home before I did any more investigation. Only I never did any more investigation, because besides poor gas mileage, it has run fine ever since with no check engine light coming on.

Until this past week...

My wife was driving back from work and noticed it was doing the surging at idle again. I took it for a drive and noticed it ran ok as long I kept the RPMs up over 1000, but when it came back down to idle it would surge, but then the surging would eventually even out to a steady idle. My first guess was the EGR valve was sticking, so I took it off and sure enough, the pintle was sticking! Aha! I cleaned out the EGR valve with carb cleaner and made sure the pintle moved up and down smoothly without sticking and re-installed it. I then took the car for a 20 minute test drive around town and on the highway and it ran great. My wife drove it the following day and didn't have any problems.

Then this morning she came back from the store upset because it was surging again. The SES light was back on and a quick check of the EGR valve showed that it was working properly without sticking. I checked for vacuum leaks...none. I also put in a new air filter. When I did the head gasket in 2006, I replaced the PCV valve, spark plugs and wires, etc, so it shouldn't be any of those. I then checked fuel pressure at the fuel rail and measured 41psi at key on, 37psi at idle...both within spec. Not a fuel delivery problem. Then I cleaned out the MAF sensor and throttle body. That didn't help either. I also cleaned out the IAC valve, but it didn't appear that dirty or sticky anyway. At this point, the car was running so poorly at idle it would start and immediately stall. As a last resort before giving up, I decided to believe the code scan from last year and replace the upstream O2 sensor. After driving it for 30 miles, the light has gone off and I'm back to a smooth idle.

Was it really that simple? Did the ECU report a code that was right on the money, not one that was being caused by something else in the emissions system failing/operating poorly? Can a bad O2 sensor really cause an engine to run that poorly? Or will the car run fine for a while and start acting up again?

Any thoughts, previous experiences, technical expertise would be more than welcome.

 

[kd5byb]

I'd say it's possible.

Example: Co-worker has a '95 F150 that was setting intermittent O2 sensor codes. The check engine light would come on long enough for him to notice, but not for as long as it took me to remember to bring in my code reader. It was getting really awful gas mileage - like 10 MPG and that was with the V-6. It also seemed to run rough to me.

Well, one day the planets line up and it throws the code when I've got the code reader with me and it says that one of the O2 sensors is wacky. I then look at the fuel trims and the side with the bad sensor is around +20% long-term while the other bank seems normal at about -7% long term fuel trim. He replaces it and the gas mileage goes up significantly and seems to run smoother to me.

But here's the thing - when the computer senses that an O2 sensor is bad, it should revert to "limp home" mode. Or, better said, it knows to ignore that sensor and reverts to settings that it knows will waste gas, but will get you home. Limp home mode shouldn't require any fancy footwork. But, in the case of this F150, limp home mode was definitely rougher than normal.

Another example - back when I had my 1994 Ford Probe, I would also get intermittent check engine lights. It was also running like garbage, using too much gas, etc... Since I didn't have a code reader and I don't think that the 1994's were OBDII yet, I yanked the sensor and took it to my workbench to test it. (I seem to recall the Probe boards had lots of posts about O2 sensors going kaput on the '94 Probes) To test, I put a voltmeter on it and used a propane torch to heat up the element AND deprive it of oxygen. As you apply/remove the flame, the output voltage should change rather dramatically. Turns out it was "lazy" - yeah, it worked, but it was really slow to respond. Strangely enough, as I kept fooling with it, heating it with the torch, it seemed to "wake up" and begin switching like it should. Back into the car it went and I had no problems with it. Gas mileage went up and driveability returned to normal!

At this point, I'd try replacing the suspect sensor.

later,
b

 

[NYSteve]

83,
O2 sensor issues can cause all sorts of issues. The purpose of the O2(anyone feel free to jump in if I am off on any of this) is to communicate to the ECU how rich or lean the air/fuel ratio is by the content of oxygen in the exhaust. This ratio is never absolutely perfect, and will be minutely rich or lean. The O2 sensor, will tell the ECU to adjust accordingly. As an O2 sensor gets less sensitive, so does the precision of its ability to monitor and adjust the fuel/air ratio so your car can run horribly because the air/fuel ratio is off. Also, as you can imagine this confuses the ECU, as other sensors are telling it information that doesn't match, ie air in the engine, timing, fuel being pumped at injectors, etc

My suggestions:
First get the car scanned(autozone does this free). Even if the engine light isn't on, it can be stored as a ghost code.
2nd, sounds like it could be a bad Throttle Position Sensor(TPS). Has this been checked?

Hope this helps
Steve

 

[mwsku83]

Steve,

I'm looking at a bluetooth scan tool online, otherwise I might go pick one up myself. If I decide not to get one I'll take it to AZ and have them scan it. I didn't want to take it anywhere earlier today the way it was running, but all seems to be well again now so I'll probably go tomorrow.

I have not checked out the TPS sensor, although it is something I had considered. If the car starts acting up again I'll look into it.

I am by no means a mechanic, but I'd like to think I have a basic understanding of what the O2 sensor does. I just have a hard time believing it alone would be responsible for the car running as poorly as it was, but I guess it's possible. I figured the ECU would be smart enough to compensate for a bad sensor (limp home mode as suggested).

I just want to get an feel of what else it could be in case the problem comes back, as I'm running out of ideas myself. I'll keep the TPS in mind...anything else?

Thanks for all the input.

 

[brianl703]

The O2 sensor is how the computer fine-tunes the air/fuel ratio. I shouldn't say fine-tune--the O2 sensor can effect quite gross changes in the air/fuel ratio, up to 25% before an error code is set.

 

[JTK]

Does the exhaust smell rich when it's acting up? IMO, if an O2 sensor is the cause of such a noticeable concern, it would throw specific codes of it's own; O2 sensor malfucntion, low or no activity, etc. EGR shouldn't be an issue at idle, nor should it effect MPGs. Does this engine have a MAP sensor as well? Main thing is, keep scanning for codes.

Joel

 

[eljefino]

I love the TPS idea, it burnt me on a saturn.

The insidiousness of this sort of thing is it won't throw a code.

Saturn, and I will stretch to imagine all of GM, recalibrates the TPS's voltage to zero every time the car is started. So it could be .33v one day and .35v the next but the computer will be cool.

If the thing jumps from one to the other though as you are driving the computer will think you have your foot on the gas slightly. The idle speed program says not to attempt setting the idle speed even if the throttle is open a teeny tiny amount. Long story short your idle is all over the place.

There are cut and dried codes for lazy o2 sensors, P0133 IIRC. Qualification is throttle between x and y, z seconds go by and fewer than q cross counts of voltage have gone by. In a spot of engineering intelligence this trouble code basically says do not pass go, change this part out.

 

[T-Keith]

A coworker had a 03 Silverado with one bad O2 sensor. It was pouring black soot out the tailpipe. Yes a bad O2 sensor can cause a car to run poorly.

Bottom line: if there is something wrong and you know what it is, Fix it.

 

[NYSteve]

El Jef-
Back when I had my Caprice/Impala, heard a lot of TPS issues that sounded exactly like this. We must have been in the same GM Intro to Fuel Injection Malfunctions class taught by Professor Trial-by-Fire, haha
Steve

 

[WindwoodTrader]

Has anyone had any experience with a sensor going bad apparently as a result of using an additive such as MMO?
My Land Cruiser with 120K and getting 15/17 just got a "Check Engine" and I haven't gotten it in yet. I've been using MMO in the gas for a couple of years now.

 

[mechtech2]

O2 sensors can go bad is an improper sealant is used.
Like an old style RTV that has silicone. It gets in the PCV, then the manifold, then cylinders, then to the O2 sensor.
MMo is not going to foul an O2 sensor.
Sure, a ridiculous amount may, but so will a ridiculous amount of plain gasoline, which is supposed to be in there.

 

[Vikas]

You got a check engine light, so naturally you should just *assume* that you need new oxygen sensors. Since it is Land Cruiser, you will need four of them. After you replace them and when the check engine light comes back on, please get the code. When the code says P0420 (or P043), please replace the catalytic converter for the left (or right) bank.

Hope this helps

- Vikas

 

[EricF]

Heated o2 sensors are slated to last 90k minimum. Anything over 90k is simply luck of the draw.
MMO shouldn't affect the o2 sensor at all.

It simply could be: just that time.

 

[Vikas]

The point was without even knowing the code, it is foolhardy to change the O2 sensors. Get the code first.

- Vikas

P.S. I am talking about the guy with Land Cruiser 120K miles