Late to post but got a couple things done this weekend. My new OEM 02 sensors arrived, and I was able to get them both installed. The car still ran rough immediately after, but I'm assuming there is some amount of time for the PCM to start receiving the signal and then to make adjustments. I should have let it idle longer to verify, but got frustrated that it wasn't a more immediate change and moved on to other things I needed to do.
It will take it a little while to warm up enough to use the upstream sensors for feedback. A couple minutes would be a good guess.
I removed the drivers side PCV from the vacuum line as others recommended earlier. I still need to find a place to mount my catch can which is rather large, but I have everything I need for that once I figure it out. I also removed the intake tube from the throttle body and hit it with some carb cleaner to make sure the blades weren't sticking. It appeared to be entirely closed and while I did PB blast the idle stop screw, adjusting it didn't seem to make a visual difference, I would have assumed based on the 19.2 (%) reading from live data on my scanner that the blades would appear to be open at idle, but they didn't. After some adjustment the percentage dropped to 18.6 (%) which makes me wonder if the TPS is faulty, or needs adjustment. I know the earlier SN95 were adjusted by loosening the screws and moving it until it read proper voltage. If it's not adjustable then the sensor may just be bad?
This part here made me wonder. I know old EECIV foxes and EECV new edges, but the SN cars I'm not as well versed with. Generally "idle" is a unique operating mode for the engine. If it thinks the throttle is cracked, it won't operate in idle mode, which means no benefits like using spark to keep a steady idle. On the old foxes you had to adjust the TPS to get the right voltages. On the later EECV cars the EEC does an auto learn on startup to determine what closed is. I know even though mine registers as something like 200 counts with the throttle closed (EEC using an A-D converter to turn volts to discrete counts), it actually interprets that as "0% throttle- throttle closed." I would make sure your scan tool isn't doing it's own interpretation based on raw voltage or something of that nature. What matters is if the EEC thinks it's closed, and sets the idle flag.
If you have a datalogger I would check to see if the idle flag is set, what spark/fuel sources are, and when you go monkeying around with the throttle, keep in mind that's going to change the IAC behavior. If it logs IAC integrator you want that to be zero.
When re-tightening the intake elbow, it caused the lower connector to pop off after I went to start it... car sounded cammed and boost gauge was showing like 4-5 for vacuum. Got it reconnected well, and started it and it was very smooth. Being the scientist I am, I can not point to any specific cause because I didn't isolate each change like I should have lol.
Could have been something that was intermittent or just bordering on faulty, and you managed to dislodge it all the way. Sometimes you just get lucky.
I wasn't able to get the wideband running correctly, as I assumed it used a LSU 4.9 sensor, and found out it's old enough it uses the LSU 4.2 sensor. I couldn't find any cheap alternatives that looked trustworthy so I guess I'm out $80 for a Bosch unit. If anyone has used a cheaper alternative with good results please let me know.
There's really only a couple manufacturers of those sensors, basically just NTK and Bosch. You can find knock off ones on places like Alibaba, but I sure as hell wouldn't trust those. You definitely want a genuine part- just a matter of shopping around to see who has the best price.