One of my downstream O2 sensors is running rich. If the cat is bad, would the O2 sensor run rich or lean?

 

It would read as Rich.
When a catalytic converter fails, it cannot effectively clean the exhaust gases. 
The downstream O2 sensor then receives exhaust with a high amount of remaining oxygen and interprets this as an excessively rich air-fuel mixture in the exhaust. 


 

It could also be a failing O2 sensor. What is the upstream sensor showing?

 

I was able to get a quick scan this morning at warm up. Here is what I found. Tell me what I am looking at.

 

STFT's are close (reading lean but "within spec") while LTFT's have a huge discrepancy. I don't remember the exact range of op temp cats, but I believe it's higher than what you shared?

Is your car fully up to temperature here? How does your live data look from other points associated with air/fuel mixture?

 

The original data set was 5 minutes after start up in the morning. I added data set below in parenthesis after driving 30 minutes. While the bk1 cat is 20 degrees less, after driving 30 minutes both seem to be in spec. One thing that stuck out to me was the LTFT Bk 1 number not changing. I checked the data 4 times today and while the other fuel trim numbers changed, the LTFT bk1 number never changed. Perhaps 12.5% is a threshold Mercedes has set to activate the CEL as a permanent message and will not change without a repair or clearing the code. I have not tried clearing the codes and seeing if this returns, but I may do that this weekend.

 

That is interesting the B1 value never changed. B1 total fuel trim is 15.1, definitely worthy of notice while B2 is looking normal. At idle does the STFT bounce back & forth between lean & rich? Does upstream & downstream consistently tell the same story? Hopefully you do just have a lazy B1 downstream sensor.

I feel like I may have read similar about that 12.5% stuck value as a default, but perhaps for a dead/dying sensor rather than CEL trigger? I could be off on that. I had an extreme rich condition not long ago (sorted via new secondary air shut off valves) & my CEL didn't come on until I reached -25% LTFT (on both banks.)

How does all this shared data operate under load vs idle? If you sat in P & then brought RPM's to about 2500, hold for a bit, then let off it would be interesting to see how the B1 LTFT & B1S2 voltage behave.

 

I've had my LTFT over 70, and goes past 12 all the time. So I'd say it's broke and you need a new pair of widebands. They're also substantially cheaper than a Cat, which is probably fine.
If they're like my widebands, you will not like changing them. I bought replacements ~3 years ago, and twice I've been under there taking a stab at swapping them. The first time I said F that! The second time I had the exhaust lowered and figured it would be doable now. Nope... Reason is the plug is mounted waaaay up high for some reason, so I can touch it, but can't grip it to unplug it. Also, I can either see the plug or touch it, but not both. Very annoying...
So come the day I need to replace them, I think I'll just cut the wires and solder the new sensors in.

Roth; on mine I've had STFT go beyond -25, but it does peg out at 25.0. Once at 25, odds are it won't fault out or anything, but sometimes it does. I'm not certain why but I think it's how lean it is while at 25, or possibly how lean x time?
I've never had a CEL or code from it. What it does is stop working and that one STFT will read 0 until I restart it. Fyi for my 2016.

 

I had both O2 sensors in bank 1 replaced, but still have a LTFT Bk 1 of 19.5%. I went by another Mercedes repair place today to talk with them and they said not to worry about it as long as there is no CEL and the car is running fine. I am not sure how I feel about that, but am unsure how to proceed.

 

for my 2013 E63, I found it very hard to get to those plugs just like you. I definitely had to do everything blind and fully under the car. When doing the driver side I had to be under the car get towards the rear using my right hand. And for the passenger side, I had to be under the car using my left hand. Once you get your hand to the plug, you twist it out of the bracket.. this was definitely something where I had to shove my hand in there as hard as I could, but only could move my fingers. Do you wanna know what’s even more fun? Putting the sensor connectors back into the bracket…. Similar to what you saw, lowering the exhaust doesn’t really effect access to the bracket that holds the sensors.. if you have a Boroscope, you can send it up there so you can see the shape of the bracket if you have not done that. Unclipping the connector from itself wasn’t too bad as long as you can get your finger on the tab.

In my case, one of my downstream sensors was reading as if the cat was bad, but as mentioned up above, it was the upstream sensor, giving a bad signal to the fueling. I replaced all four.

 

Davery,

Those trims would bug the crap outta me and I would not be able to let it go.

Fyi my trims were ~10% when I got my car at 28k miles. I noticed others with the same engine were also seeing 5-10% trims at all times. The required 10% Ethanol in my state is part of it, but mostly it's simply an error. Maybe the gas in Germany is different or something, who knows.
You can actually adjust it in the ECU, because I do, but it's technically illegal. I adj mine because I run ~50% Ethanol. I'm sure it would run fine without adjusting it, but getting the trims to average 0 is basically an OCD thing. Plus if O2's were to fault out, or who knows what kind of error to cause trims to fail, it will now continue to deliver the correct fuel. If it was relying on 25% trims, which I suppose is about where it would be, and that trim went away and I floored it, I'd imagine my eng life span would be measured in seconds. So there's that...
Tinkering with it how I ended up getting crazy trims like 70, because I first adjusted in the wrong direction :o

Yours, and especially just one side, still makes me think something is physically wrong. If not the wideband sensor, then I guess I'd go with dirty-ish injectors? Maybe even a vacuum leak or the PCV has given up the ghost, but I'd assume a leak or PCV fail would effect both sides more or less equally, but the PCV is on that side. If it were me I'd add a bunch of injector cleaner to see if it changes, and if your PCV is original I'd replace it anyway. When they go they carbon up the valves much worse than normal.
While I've never paid much attention to the high pressure fuel pumps, or the plumbing, maybe one feeds one side and one the other? If so, and one spit out some debris, it would only clog injectors on that side. Just a thought.

I would also check trims as you drive and floor it just to see what happens. I saw you said you checked four times, but I don't suppose your tool can record live data for playback? Because they move all over the place all the time. The exception is the LTFT's freeze when floored. I forget if it's load or rpm, or maybe both, just that they do.
Fyi I record my trims 100% of the time for the past 3+ years, from before I start it to after I shutdown, so I have a very good feel for them. I only record the widebands because I could care less about the O2's in back, which are just emission sensors.
And I would most certainly check and clear codes. It's possible it's hung up because of that, but just a guess. I'd likely also do a system reset by disconnecting the batt for a while. How long is needed I don't know, and while I think 5min is likely enough, overnight can't hurt to be sure. Do yourself a favor and be sure the windows are all rolled up, doors closed, then pull the batt cable. On mine, if I don't do that, it screws up the window settings and it's usually a battle to reset them.

Have you checked your ECU plug for oil? In case you haven't heard, it's a thing with our cars and can cause all kinds of mystery ghost issues.


Baltistyle, thanks for feeling my pain, and the tips. Perhaps I will try again next time I'm under there changing the oil. I do recall the first time thinking there was no way, but I tried that second time anyway. I recall again thinking; yup no F-ing way. But it won't hurt to try again, with your tips My guess, knowing myself as I do, is they won't get replaced until one actually fails.

 

@Chevota

The fuel rails are joined by a crossover pipe.. A fuel trim issue on one side could be a dirty or sticking injector...or could also be a spark issue if it's a rich condition.

When my HPFP(s) went bad, I had severe rail pressure drops when the car got into boost.

P.S. I datalog very often and it's helped greatly to diagnose problems right when they happen.

 

I just had the plugs replaced a couple of months ago. I will try some injector cleaner, although I generally don't think they really do much.

 

Here is my question to those that know more than I do. My LTFT percent appears high, which would indicate a lean issue on bank 1. It is my understanding the LTFT is an average number over time, which explains why it doesn't change as often as the STFT. The STFT Bk 1 percent is within spec everytime I take a reading. . Does the LTFT use the STFT data over time to calculate the LTFT average percent? If so, that would mean at some point the STFT is reading a high percent in order the create an LTFT of a high percent? I just don't see how a normal STFT percent could create an abnormal LTFT percent.

 

Thanks Siegmann I'll look for that crossover next time I'm poking around.
I wanted to pull my intake at one point, for a look-c at my valve carbon and maybe add a Meth injector or two, but the straw that broke the camels back to not do it was the Inj lines. Not just a pita, but I kept thinking little bits of crap would get in there and partially clog an injector(s). Like the little slivers that tend to break free from the threads of the fittings, or random dirt that may be in them.
The old saying applies; if it ain't broke... Which I often ignore and sometimes regret.
I've had rail psi drops too, but my fault from adding too much Ethanol.

I've never seen or had to deal with a sticking Inj, but I wonder if maybe some lube would help? I used Lucas Inj Cleaner for a while, which is mostly lube. If I had to guess it's just Marvel Mystery Oil with cleaner added? I think I used two gallons of it, like 6oz per tank, but stopped using it because it lowers octane more than I like.
So, knowing a little about Inj design, wouldn't they tend to stick or hang open rather than closed? I'd have to assume it's not fully stuck in either position or he'd get misfire code, and obviously feel it, so I'd have to assume it's simply not opening enough, or fast enough. Is that a thing that happens? If so, then maybe that Lucas stuff would be ideal?

Davery; I wonder too, but I just want to see if causes your trim to change for the better. Even if just a little it would be a clue, and if so I'd add more to see if it can fix it altogether.
Maybe it works great which is why I've never had a clogged injector, ever. Or it could be like my lucky charm that keeps Bigfoot away, and since I've never seen one it must work
Didn't want to do the plugs yourself? I did mine and it wasn't as bad as I thought it would be. I'm just too tight to pay for any mech work. Plus it's quicker if I do it and I know it's done it right. Plus sometimes I like to experiment and do things differently. E.g. I modified my spark plugs, and used different, and more, lube on the boots.

The STFT is very fast response to lean/rich conditions. You can watch it bounce around constantly as it tries to adjust. LTFT are not really very long, just longer than ST. So while ST can sometimes be a blur as it moves to fast, LT moves more slowly, like in seconds. So if STFT hovers ~10 and stays there, it may take a few seconds for it to decide that should be LT, and it will then shift that to LT and ST will avg more zero-ish and LT will be 10 higher. I forgot to mention that LT also doesn't move much, or can freeze, at idle too. So ST works at all times, assuming the O2 sensors are working. LT is lazy and at idle or at too much rpm/load it will turn off and simply lock at whatever value it was at that moment.
So in your case your trims could be 12.5 (assuming yours isn't broke and stuck there) while driving, but when you drop to idle LT could freeze, but if you need an extra 5% at idle ST fills that need.
I suppose another way to look at is a course and fine adjustment? I'm not sure why ST can't simply do it all, but I assume there must be some reason LT exists?
If your tool can't record live data, you can get a phone app like Torque Pro and an OBD dongle. Combined cost is ~$15. Then you can record ST & LT for an entire drive and put it in an Excel sheet chart to see how it acts compared to rpm and load.

 

My scan tool can read live data, but I have not found a way to save data. I may connect it, drive around and just monitor the STFT and LTFT Bk 1 numbers. If my hypothesis is correct, then the STFT number should increase at some point.

 

ST is likely always moving, up/down, most all the time, but the ideal goal is 0, with the help of LT. If your LT is working, it will do that, on avg. Or I should say try to.
Say I'm driving and suddenly I'm 20% lean for whatever reason. STFT will go to 20 in a heartbeat to compensate, then in a few seconds that 20 will transfer to LT and ST will end up reading 0 while LT now reads 20. If the 20% lean condition continues then ST stays more or less 0 from then on, and LT will stay at 20. If suddenly it goes back to normal then ST will instantly go to -20 (because LT is still 20 and slow to respond), then LT will drop from 20 to 0 as ST goes from -20 to 0.

If the fuel maps were correct and all sensors were super fast then there shouldn't be trim adjustments at all, but that isn't reality. There are many fuel map variables and physical sensor data that go into the ECU's decision on how much fuel to deliver. Maps are not 100% accurate, and sensors are neither accurate or fast. Take the air temp sensor for example, which exists primarily to estimate the air density, which is a factor in fuel requirements. Well, that sensor is VERY slow to respond to temp, which may change quickly, especially in an M278/M157. Thanks to the lagging data the ECU gives the wrong amt of fuel. The wideband O2's are quick and say; hey, 5% too lean foolio! The ECU says oops, my bad, here 5% ST fuel adjustment to make up for it. If the temp reading corrects quick enough so the fuel command is correct, then ST trim will zero out and that's that. If the temp reading takes too long then LTFT says; hey, ST, I see you've been 5% for a while, so I'll take that off your hands. Then ST = 0 and LT = 5. But now the temp reading will catch up to reality and ST will pull fuel to compensate because LT is too slow, then moments later LT will go back to 0 as will ST.

It kinda reminds me of work, and my manager (ECU) is trying to make everything work smoothly, but all those people make it difficult because they're not doing what they're supposed to, or doing it too slowly etc. Plus work loads change every day/week/month so the poor guy is constantly adjusting who does what and when. Then someone get a paper cut on their widdle fingy and has to go home, or whatever happens, and things happen daily, and he has rearrange again. Or something breaks, a part doesn't arrive in time and so on, and it never ends, like trims

I don't know what tool you have but you could rig something to have your phone or something record the screen? That way you could spend more time pushing the car, like more gas, flooring it, etc to get a better feel for what it's doing. I did similar once with a camera to watch my turbo wastegates to see if they were working right, or my ECU was screwing up. It was the ECU.

 

Here's my trims for an minute of extreme varying conditions. I should help and hopefully not make it more confusing. Lotta stuff going on there but you'll figure out
I was unable to post the 4MB .mp4, so here's a link to download it. I think the free service is only good for five downloads so let Davery get it first.

https://tinyurl.com/E550TrimsforDavery

Here's an explanation of what each item is to help: