O2 sensor codes and issues

[JohnW63]

I have a 95 SHO. Some years back the plug for the rear O2 sensor melted together. No way to unplug it. Long story why. I got another upper plug and some wires from a junk yard, but I never moved forward on doing much about it. The smog center guy said I'd better not leave that sensor in much longer because that bank is running too rich and it will foul up the catalytic converter. So, I got off my butt and wired the new pigtail in, at the end of the wires of the old O2 sensor. Just cut the old sensor off and used crimp connectors to hook them together. Before the swap, the error code would state things were to lean, which causes that side to try to run rich to compensate. When I wired it up, I got two wires crossed. I think it was a heater line to either the ground or signal line. Of course, I still got a check engine light. After going back on the lift I found the error and corrected it. I drove up and down the street and got no codes. Yay! But, on the way home, I got a light, then it went off, then back on. Eventually it stayed off. The light didn't go off, when it was wired wrong. It just took a couple blocks to finally turn on. I checked the codes and now have a 173, rather than 172 stating it's too rich now. So, the question is... Is the computer trying to figure this all out, or did I mess up my O2 sensor, when it was wired wrong ? Not used to having a check engine light go off, that wasn't due to oil pressure.

 

[JohnW63]

Should I post this in another section ?

 

[luigisho]

Couldn't hurt. Not a common occurrence so not sure how much experience is floating around among the collective. Maybe post in the v6 mailing list and the facebook group to get more eyeballs too.

 

[rubydist]

So here is most likely what is going on. The pcm has a table for how it adjusts the fuel trim for each bank. Think of a spreadsheet with engine speed on one axis and engine load on the other axis. Each cell in the table has a value for how the fuel trim is adjusted. Allowed values are from -20% to +20% in each cell. The way it sets the value in each cell is that the engine needs to be operated in that cell (this much load at this much engine rpm) long enough for the feedback loop to say "we need to adjust the fuel trim for this cell" and tweak it.

Since you have been running without a oxy sensor on that bank for so long, the fuel trims for each cell are messed up. Now that you have an operating oxy sensor again, it will take a while (as much as several weeks) for the engine to spend enough time in each cell to adjust that cell back to a proper fuel trim. When the pcm recognizes that the a/f is way off, it will set that code and turn on the cel for a period of time (I think 10 seconds is the minimum time, iirc).

So, you have two choices. 1 is to just drive the car for a month and the code and cel will likely stop. 2 is to disconnect the battery overnight so the pcm is reset to the default values and then drive it a while to let it get dialed in. Both have pros and cons, so there is no "best" answer on which to do.

 

[JohnW63]

I was wondering if there was a "relearn" period. Since the Check Engine light goes out a few times, I wasn't sure what it meant. I was about to buy another OS sensor, thinking I had messed it up, by having it mis-wired for a bit. If pulling the battery for a bit will help clear it, I may do that, as I need to get it smogged, and they won't pass it, if the check engine light comes on.

 

[zak]

If you sent heater power down the signal wire its likely the sensor is fried (in which case it will signal constantly lean). You can check by sticking pins through the wires and checking to see it it is still switching above/below 0.4 volts or so cyclically with the engine held at say 1300 rpm. Cover the pin holes afterwards with dielectric grease. Never use crimp connectors for this, any voltage drop even a minor one will alter the signaling. Needs to be soldered and shrink wrapped.
Is this an MTX or ATX car?

 

[JohnW63]

Zak,

It's a ATX. The code is sends it that it is too rich on that bank. I was going to try using a pair of sharp pins to test for voltage, but the wires are pretty stiff and dry. I would have preferred to solder the job, but I would need three hands. Two to hold the wires and and one to hold the soldering iron. Considering where the wires are, even with the "extension" that was made by the existing old sensor pigtail, I don't think I could get three hands in there. Perhaps I may be able to pretwist the wires, but that can be tricky, if you can't get both your hands in the space and still see what you're doing. I was thinking that if the sensor was fried, it would come and and stay on, pretty quickly. It takes a few miles for the pattern of coming on and off a few times to kick in. It's always a long period of on, then some off and ons.

 

[rubydist]

fyi, it takes a few miles for the pcm to go into closed loop mode - when the engine is first started it runs in open loop during the warm-up phase, during which time it ignores the oxy sensors. That is why the cel will not come on for a while with this type of code.