Downstream O2 voltage changing, other wonky readings

Joined
Jul 13, 2020
Messages
569
My car seems confused. Whenever it's running poorly or differently, I find something with my OBD reader that isn't right. I recently installed an OEM MAF sensor after semi-frequent issues with the old one. But, just like the old one, it worked great for only a few days. Fuel trims were right at 0, then gradually started going negative. Again. Yesterday LTFT was at -8.9%, so I unhooked the battery, cleaned and replaced the MAF sensor and it went back to normal. Today, my LTFT started trending negative again and I noticed my MAF grams per second reading was inconsistent and unsually high (around 5.6gps at idle.) I also noticed my downstream O2 sensor voltage was fluctuating. All sensors were "ready" by the way. The O2 would drop as low as 0.08 and get as high as 0.7. It wasn't fluctuating as fast as the upstream, but it would not stay the same for more than a couple seconds. I really don't know what's going on. Can any of you make sense of this? I'd really appreciate it.
 
Did you notice the voltage on the downstream go down to close to zero when you were coasting? Then up close to .7 while cruising and fluctuating around that voltage?

If so? completely normal operation with a operational catalytic converter.

As for grams per second at idle, what were the conditions you were seeing 5.6?

A rule of thumb on an un modified engine at hot idle, neutral/park, no head lights, accessories off, no AC, no cooling fans, you should see close to your engine displacement for grams per second. So the vibe has a 2.4? I'd expect to see in the 2.4-3 range with the conditions I just described.
 
Did you notice the voltage on the downstream go down to close to zero when you were coasting? Then up close to .7 while cruising and fluctuating around that voltage?

If so? completely normal operation with a operational catalytic converter.

As for grams per second at idle, what were the conditions you were seeing 5.6?

A rule of thumb on an un modified engine at hot idle, neutral/park, no head lights, accessories off, no AC, no cooling fans, you should see close to your engine displacement for grams per second. So the vibe has a 2.4? I'd expect to see in the 2.4-3 range with the conditions I just described.
I'm pretty sure I saw the downstream fluctuating at idle. The MAF reading was at idle with the AC on, so about 100 more rpm. I'll check again with this new info in mind. Thanks.
 
Leaked fuel injector or fuel tank cap?
Possibly. if I have a leaking injector it's not leaking all the time. I may need a new gas cap, but would that have any effect on fuel trim/O2 sensors?
 
A/C on is a lot more horsepower than just idling, so increase in air seems in line. Fuel trim within 10% is not a problem. There are manufacturing tolerances in everything thus there is a trim system. Long term stuck at 0.0 probably means you're running open loop.

The MAF should read zero with engine not running. The temperature sensors should read as expected.
 
A/C on is a lot more horsepower than just idling, so increase in air seems in line. Fuel trim within 10% is not a problem. There are manufacturing tolerances in everything thus there is a trim system. Long term stuck at 0.0 probably means you're running open loop.

The MAF should read zero with engine not running. The temperature sensors should read as expected.
MAF reads 0.46 gps engine off.
 
Interestingly (or not) 0.46 is about 8% of 5.6.
Is it reading high? It's still a bit warm under the hood. I may just be paranoid since I got a CEL for P0101 a few times this year.
 
I'm just saying that there's a +8% offset error in that part alone, thus a -8% trim occurs. That does seem a bit excessive off of zero. The specification for acceptable error might be in the service manual.

Cleaning the MAF is not a normal service procedure it is sort of a desperate measure to avoid buying a new one.

Is the car running badly or are we just overthinking things?
 
I'm just saying that there's a +8% offset error in that part alone, thus a -8% trim occurs. That does seem a bit excessive off of zero. The specification for acceptable error might be in the service manual.

Cleaning the MAF is not a normal service procedure it is sort of a desperate measure to avoid buying a new one.
Yes, I was thinking it was the disconnecting of the battery rather than cleaning the MAF that had any effect. It's strange because I've had fuel trims close to zero for long stretches of time. I got an OEM MAF sensor because I thought the Denso first time fit may had failed. This is strange.
 
Long-term fuel trim is at -7.8%, but what's more troubling is my downstream oxygen sensor voltage is fluctuating at idle from 1 to 7
 
Long-term fuel trim is at -7.8%, but what's more troubling is my downstream oxygen sensor voltage is fluctuating at idle from 1 to 7
You are talking about two different things here. Downstream O2 sensors have precisely zero affect on fueling. They are only there to make sure cat efficiency is up to snuff and unless you're triggering a CAT efficiency code, forget about it because it has no bearing on your perceived fueling issue.
 
The downstream O2 at idle won't give reliable readings. It should stay above 0.550V at the least but usually ~80% of the highest pre cat O2 reading, when cruising down the highway.
 
From what I recall, modern Subarus do use the downstream O2 sensor for some weighted use in fueling. At least I recall reading that while learning tuning on my previous 08 STI.
 
From what I recall, modern Subarus do use the downstream O2 sensor for some weighted use in fueling. At least I recall reading that while learning tuning on my previous 08 STI.
I've heard that too. Must suck when the take the cats out. You can use a good wide band o2 that has a narrow band output emulate the 2nd O2 voltage to make most ECUs happy by setting it up for a constant 0.650v or so.
 
You should be looking at your STFT, your LTFT won't be populated until many complete heat up/cool down cycles have been completed.
Are your upstream O2 sensors heated?
 
I'll try to be more clear.
I have had CEL's a couple times this year. They were all for system too rich, system too lean, and/or MAF sensor out of range.
My fuel trims would be close to zero at idle for weeks or months then suddenly go past +/- 10 and I'm trying to address this before the CEL comes on again.
Which brings me to my downstream O2 sensor. I know it doesn't affect AFR, but it fluctuates across the entire range at idle (warmed up, all sensors long-since ready and functional) which possibly indicates Cat inefficiency. My cat is also fairly new and was replaced about 8 months ago, so I wouldn't expect to replace it again any time soon.
I can't figure out why my fuel trims randomly go haywire and replacing/cleaning the MAF sensor helps, but only temporarily.
I will work my lunch shift today and report back with detailed information.
 
Put the other MAF back in and see if engine not running reading is closer to zero. I think one cause of P0101 is not reading close to zero before startup.

randomly go haywire

Wiggle the MAF plug and wires while watching the live reading, if MAF or IAT goes haywire there's a bad connection.

Downstream O2 is supposed to cycle, but slowly. The chemical reactions in a functioning cat slow down the changes in oxygen occurring upstream. If the downstream starts to exactly follow the upstream this means the cat has lost its chemical activity and the system will throw a P0420 code.
 
Back
Top