i'll try to keep this brief but I'm baffled.
Been getting a lean bank one code on a 97 3800
I was able to find a small leak with carb spray while running between the lower intake and the plastic plentum. Knowing of the egr tube issue on these and at 155k miles I decided to replace the upper plentum.

Still getting the lean code. looked like crazy for other sources of un-metered air. seems like none. Got to thinking, just maybe, o2 related. I found (from previous owner) a generic, i think, 4 wire o2 sensor that was spliced into the factory o2 sensor disconnect plug. the two heating wires and the black wire and the grey wire were not spliced together to match up at the plug. I mean that the grey was hooked to a white and so on. Mind you I was not getting any o2 sensor codes. I cut the 4 wires, added connectors, and plugged them all in so they lined up, blac to black, grey to grey, etc.

Now I get a No activity code at the o2 sensor!! The lean code went away.
Mix em back up and lean code is back and o2 code goes away.

Installed another o2 sensor thinking maybe that one was goofy. Lined the wires up according to color match and No o2 sensor activity code is back.

It does run a little funny, lopes a little, or chugs when the o2 sensor is wired up according to correct color but Not when the wires are mixed up.

Poltergeist? Gremelins? Nutty mechanic? Any ideas? If I wasn't already bald this car would have me there.

First, get teh correct O2 sensor, go to www.rockauto.com and look up what you need, do not get a Bosch O2 sensor. Plug it in and see if that works. When dealing with cut and splice things, we have no way of know if it'* right or wrong. If you have access to a live data scanner you can scan the O2 sensor, it should go up and down. When unplugged it shoudl read about 450mv

It isn't a problem putting the right wores together because both the wiring harness end and the O2 sensor end are color coded.

Can a bad o2 sensor trip a lean code? Why would having the o2 sensor wires mis-wired not trip an o2 sensor code and run good, and wireing it right go back to the lean code and run bad?

Do you know what brand the sensor is?

Until you scan the O2 sensor for live data it'* really hard to tell if it'* working or not.

Don't know the brand of either o2 sensor. I'm so tired of piling money into this car that I pulled one from the junk yard today. I know ill get curcified for that but in my experience o2 sensors don't go bad very often but can trip a code for other issues. Also wired wrong gives no code? Wired wrong runs good? Something funny here. Hoped someone may have dealt the same issue.

O2 sensors can get slow before they fail. If the sensor is clogged with carbon it will give a lean code. If you have a leak, that can cause a lean code, so look for vacuum leaks and exhaust leaks between the heads and the O2 sensor.

Thanks for putting up with me.
I got out the stethoscope and found an exhaust leak (that I could hear but ignored for now because if location). It is after bank one and bank two manifolds come together and context to the down pipe to the converter. The o2 sensor is upstream of this gasket leak. It'* only a couple of inches but upstream none the less so it shouldn't bother the sensor.

Really bugs me that wired right it runs bad and trips code o2 (no activity). Wired wrong, no o2 code but lean code and runs good.

Also, unless I'm totally missing something, there is only one o2 sensor above the cat. How does it know lean bank one when the sensor is after both exhaust manifolds join together?

That'* the only sensor before the cat so it'* called bank 1

maybe its wired wrong at the computer and not the sensor so when you wire it to what you think is right at the sensor it says its wrong?