"Tired" O2 Sensors?

[miamibob]

Have a 1995 GM with over 186K miles on it. Got it with over 185K so not sure if the O2 sensors are original. No codes but is it possible they weaken to the point of negatively affecting MPG without throwing a code? Have the socket to change them but have never done so. Do they tend to be REAL tight after all these years? Any tricks to remove them and replacing them?? Any additional info would be most appreciated....thanks....MB

[86VickyLX]

Have a 1995 GM with over 186K miles on it. Got it with over 185K so not sure if the O2 sensors are original. No codes but is it possible they weaken to the point of negatively affecting MPG without throwing a code? Have the socket to change them but have never done so. Do they tend to be REAL tight after all these years? Any tricks to remove them and replacing them?? Any additional info would be most appreciated....thanks....MB

In the order of the questions you asked...
Yes, sometimes, heat and pb blaster.
Clean the threads before you install the new ones, and use the little packet that comes with them to keep the threads from getting stuck later. Also, when you're done, make sure that the wires don't contact the exhaust, or anywhere where they would get pinched.
I think you may have 4 O2 sensors (OBD-II). 2 upsteam of Catalytic converters, and 2 after the cats. The first two I think are on the manifolds, and if not, the down pipes.

[miamibob]

Great info, as usual! Any REAL benefit to changing them without bad codes? Thanks....MB

[86VickyLX]

Better gas mileage. I'd suggest a full tune up as well. Plugs, wires, air filter, fuel filter, oil change. The whole nine yards (if you haven't already)

[miamibob]

All the rest have already been done but sound ideas. Any thoughts about the sensors??

[91waggin]

On my girlfriend's 98 Escort, which admittedly is not identical, the o2 sensors would throw codes occasionally. We would reset them and they wouldn't show back up for awhile. When we finally got around to replacing both the pre- and post- cat o2 sensors, she got an average 6 mpg bump.

[gadget73]

I wouldn't bother switching the two rear ones. Those are there to make sure the catalytic converter is working. They are not used for fuel mixture trimming. The front-most sensors, one per side, is used for that.

[BerniniCaCO3]

Redundant, but I've a friend with a 1999 ford ranger. Wondering why his fuel economy was so bad (we told him it's because it's a pickup!), then finally got bad enough he got an intermittent check engine light.
We pulled the code, changed the O2 sensor, light went out and he got maybe 2mpg back.
It was NOT one of the front fuel-trim sensors, it was halfway down the exhaust, but I guess it mattered a little.

By the way --and someone do correct me if I have this wrong, I strive for autoomniscience but only have 10 months of intensive studying -- but I believe a check engine code is triggered simply when the O2 sensor's signal falls out of the normal parameters that the computer expects. All this really means is that the voltage at the computer is not what it wants. It could be a bad O2 sensor, but it could be corrosion in the harness eating voltage (if sensor after brand new sensor continues to trigger codes!), and it could be, shocking, that the sensor was working correctly all along and doing its job, and that some other problem has led the engine to run way rich or way lean, like a vacuum leak or stuck fuel injector. (You'd probably know a fuel injector stuck open though, by the rich smell of gas and dark smoke from the exhaust at all times, even after the engine's warmed up).
We haven't gotten into this in class, but I think obdii systems with the right diagnostic software, let you access real time signals? That would let you know if the signal from an O2 sensor is playing at the borders of normal but not officially into check-engine territory yet. Not worth the trouble for you, just fwiw.

[BerniniCaCO3]

I ran into an older indian man at a pepboys who was running around checking his battery and alternator wondering if it would help him...
his story was that he had an O2 sensor code, the dealer (antwerpen dodge I think) replaced all of his O2 sensors. No dice. Then they told him it was the computer, and replaced that. Still no fix.
So he's out of money and out of patience, and is wondering if battery voltage could trigger an O2 code! Poor guy.
I'm just guessing that especially with the new O2 sensors, it must surely be some other root cause, like a vac leak, that would lead good O2 sensors out of acceptable range.

[86VickyLX]

I ran into an older indian man at a pepboys who was running around checking his battery and alternator wondering if it would help him...
his story was that he had an O2 sensor code, the dealer (antwerpen dodge I think) replaced all of his O2 sensors. No dice. Then they told him it was the computer, and replaced that. Still no fix.
So he's out of money and out of patience, and is wondering if battery voltage could trigger an O2 code! Poor guy.
I'm just guessing that especially with the new O2 sensors, it must surely be some other root cause, like a vac leak, that would lead good O2 sensors out of acceptable range.

Lower than normal battery voltage can fuck with how an EFI system runs. If the charging system is not up to snuff, or if it's overcharging, weird shit can happen, and it can damage the computer and other various things. Sometimes if you're lucky, it'll blow fuses. Due to the increase in charging system voltage, obviously the amperage is way up as well.

Another thing to look at is fuses. We had a 1996 Ford Ranger Extended bed (I think) at scotts. It came in for a bunch of stuff, including replacing the O2 sensors. There were 3. 2 upstream, and one downstream (single exhaust). Ivan had replaced all 3 sensors with brandy new boschs. We let it run for a while and the check engine light came on. We scanned it for codes. O2 sensor heater circuit failure. We weren't entirely sure what the problem was, until I noticed the tach was not working. Scott had a hunch of a dead fuse. Sure enough, when I checked the fuses there was one with no power on one side (key on powered fuse) Turned out to be the fuse for the O2 sensor heater wiring and the tach power (not signal). I replaced the fuse (15amp) and bingo, no more light.


In regards to how the O2 sensors are checked via the computer.
I am talking about narrow O2 sensors, not the A/F or widebands O2s, though I figure it's about the same, just a broader voltage scale...
Anywho
The computer watches the upstream O2 sensors this is used for fuel trim adjustments. As the computer does some minor tweaks, the sensor will switch from lean to rich, and vice versa. The O2 sensors work on a scale of about .250v (lean) and .850v (rich)middle being .450 being stoiciometric (14.7:1 A/F ratio). The O2 sensors now, and some since the mid '80s like on our SEFI cars, are heated, which help them start to read faster (so they switch properly). Non heated ones work, but it takes them longer to get warmed up for the computer to start working in Closed loop operation (where the computer looks at O2 sensor values). In addition to heaters, some computers have timers to see how long from start up till O2 sensor switching takes. If it takes longer for the sensors to switch, it will throw a code for failure to warm up. Sometimes this is normal, if the car is in exceptionally cold weather, and so on. If it's more frequent in warmer weather, then I'd expect a resistance/corrosion problem in the heater circuit, or the heater itself is bad.
They also use timers to see if there is a problem with the engine. What I mean by that is, if the computer sees that the O2 sensor hasn't switched from rich to lean in a programmed amount of time, it will trigger a check engine light immediately, a rich code for either bank 1 or 2 (OBD-I only) Same idea for lean.

Though the OBD-II cars have a little more adjustment, and adaptive memory to deal with extended rich or lean conditions. If the readings finally make it out of the computer's adaptive limits, then and only then will it trigger a check engine light for this...


Now, the only reason we need O2 sensors is so the catalytic converters get fucked up. Since no engine can run at stoic all the time (acceleration and decel would definitely damage the cats. The above applies to Ford OBD-I and OBD-II cars unless otherwise noted.

Below here applies only to all makes with OBD-II...
Once you understand the above, now introduce the catalyst monitor. This is basically a program in the computer that looks at the O2 sensors after the cat for activity. Generally, the readings on these sensors they will look like the upstream sensors when the cat is not warmed up. And when the cat is warmed up they will be stable and won't fluctuate much. This means that the Cat is doing something. Which is what the computer wants to see.

Now with active monitoring, the computer will purposely change the a/f ratio so it sees a change in activity. The upstream O2 sensor values should change whether or not if changed lean or rich. It will then expect to see a reaction after the cat to see if it's properly working. If no change or a reaction that took longer than expected, then it will store a pending trouble code. It will only throw a check engine light if it sees 3 consecutive trips with this outcome. And the light will stay on for 3 consecutive drive cycles where is sees no problems (in regards to this). The 3rd trip, the light will go out. The code will stay in computer memory for 80 cycles before completely erased from memory (unless someone goes in there with a scan tool).

Codes normally seen there are P0430 and P0420 which are catalyst efficiency codes.

[miamibob]

All VERY useful info but not sure I should bother to change them without getting any bad codes. What do we all think about that??? MB

[86VickyLX]

All VERY useful info but not sure I should bother to change them without getting any bad codes. What do we all think about that??? MB

Replace them, they do get contaminated after a while, and take a while to respond. They're not that expensive.

[BerniniCaCO3]

yeah, thanks, 86Vicky!

The designed lifetime was for 100,000 miles.
At 180k and 15 years, yeah, I'd put them onto my laundry list of regular maintenance items, codes or no.

Friend's ranger was '99 and had 80,000 miles, needed one.

They can also get "poisoned" by silicone (no, I don't know the chemistry involved), or just sooty. Sometimes even removing and cleaning works.

Yup, you'll need to be able to exert some force on 'em.
On my particular engine setup, they were in the tops of the exhaust manifolds, and I had a bear of a time getting to them. The driver side I cut the wires on the old one (tossing it, after all!) and used an offset box wrench (15/16"? They're all some standard size), the only thing that could fit there, and yanked real hard!
On the passenger side some other accessories were in the way going in from the top (in retrospect maybe I should have just unbolted them), and going from the bottom with a flex-head ratchet, the right combination of extensions maybe including a wobble extension, and then, because I could not get the force needed with the ratchet, a giant screwdriver/ prybar to at least break it free.

Liquid wrench is your friend.

Yours might well be easier. Some of them are REALLY easy to get to, right there staring you in the face once you're under the car, easier than any spark plug even.
This is probably the last time you'll do them, but no harm in using antifreeze. If you still have the car in 5 years and one goes bad, you'll thank yourself.

[BerniniCaCO3]

umm, *antiseize.

[86VickyLX]

Antifreeze contaminates them too. So does burning oil. Silicone pretty much does the same thing. The fumes will coat the sensor and cause it to give bogus readings. Use sensor safe rtv on exhaust or intake stuff.