Rough running, AFR reading lean but engine running rich

I have a 2005 Prius that is running rough. Eventually it will trigger a hybrid system code when the ICE fails to start. Once it also triggered an air fuel sensor code (bank 1, sensor 1) that wasn't the heater-related code. I didn't record them then they happened.

Thinking the AFR sensor was bad I replaced it (the existing one was about a year old). No change in behavior. Digging deeper, I found the ECU registering the O2S1 voltage between 4.5V and 5V when the ICE is running. Apparently this should be around 3.3V and 4.5V is a very lean condition.

I went to replace the AFR again and the sensor I pulled from car was covered in black soot. This is after only about 30 minutes of running. I pulled the spark plugs and they were also black with soot as well. My understanding is that this indicates a rich condition (The air filter is new). So we have an AFR that indicates the engine is excessively lean, while reality indicates a rich condition. I checked the voltages at the sensor and at the ECU and they register 3V and 3.3V on each of the signal wires, so wiring doesn't seem to be an issue.

Thinking the only thing left is the ECU, I made a run to the junkyard today and replaced it. Tested and it still has the same issue. The ECU measures the AFR sensor voltage between 4.5V to 5V, and the ICE runs like crap. If I drive the car long enough, it runs worse and worse to the point where when the engine stops there is a horrible shaking and clattering and I'm afraid the engine will be damaged.

Today I did catch on the scanner where the AFR sensor read around 3.3V for a short period of time about a minute after I started the car (picture me doing fist bumps and celebrating my victory). Then the voltage jumped up and the engine started to run like crap.

With the old ECU the LTFT was about 20% and the STFT was around the same. When the AFR sensor read 3.3V, the STFT was 0%, but jumped up past 20% when the AFR reading rose to 4.5V. I haven't checked the LTFT with the replacement ECU. If I disconnect the AFR sensor the engine runs much better because it relies on the downstream O2 sensor.

I'm at a loss as to what to try next, and I'm starting to wonder if I'm looking in the wrong spot. Anyone have ideas about what would cause the AFR reading to apparently be the inverse of what it should be?

 

Update after some more testing: I noticed the O2 sensor was reading 0.0V. Removed and tested and reinstalled in the car and now the computer reads 0.9 volts and backprobing the connector reads 0.6V. The AFR sensor at one point drifted down to 3.5V, but when the ICE stopped and restarted it was back to the 4.5V-5V range.

 

The vehicle has 285000 miles on it. I have noticed misfires during cold startup since I purchased the vehicle about 3 years ago, but the rough running when warm started a few weeks ago. Now with COVID I only drive it once or twice per week.

I'm not familiar with techstream. I have been using bluetooth code readers for information, I have a Bosch reader and recently purchased a Veepeak reader to help troubleshoot this issue.

 

I replaced with an aftermarket sensor, Ultra Power 250-54006. The part number matches one from Walker, but those may be generic part numbers.

Are you suggesting I try the Denso sensor instead?

 

I suspected a vacuum leak as well, but digging into the problem, it turned out to not be the case. When I inspected the spark plugs and the new AFR sensor they were covered in black soot, indicating the engine is running rich, not lean. I checked by introducing a vacuum leak and the engine ran smoother, engine rpm increased, and the downstream (o2) sensor switched from rich to lean. The upstream (AFR) sensor remained lean throughout.

 

I pulled a Toyota labelled AFR sensor from a wrecked car at the junkyard today. Well, actually pulled the exhaust manifold with the sensor still in it because none of my o2 sensor sockets would fit inside the cylindrical heat shield. I used it to replace the aftermarket sensor I had installed. The car works as expected now. Success!

If there was a moral to this story, it's probably "buy sh***y parts, win stupid prizes". That was three budget sensors that all exhibited the same symptoms. For those reading this thread in the future, you probably want to shell out for a Denso AFR sensor or Toyota sensor.

I took it out for a short test drive and the car seems eager to accelerate. It might have been low on power for a while and I never noticed it. I'm going to take it out for a full drive cycle next, and test under cold start conditions in the morning.

I probably wouldn't have been able to fix it without everybody's suggestions. Thank you!

I'm attaching a screenshot of the engine running normally.


Edit: I forgot to mention that I had cleaned the MAF sensor a few days ago, and also swapped out with a junkyard part. There was no change in readings when I did this.

Edit 2: I put a bottle of Techron in the tank. I previously ran fuel system cleaner through this car , but I hadn't heard of polyetheramine (PEA).