When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
You just confirmed this as a narrow band O2 sensor. "The OEM are .7 to .8 but in the extreme 0 to .93" . If voltage range is only 0-1 v then those are narrow band O2 sensors. Stop screwing around and get 2 WB O2 sensors, and install pre cat, one on each bank. then log them and correlate to boost, AFR, timing , rpm and load
The changes you are making are and have caused the limp mode and transmission issues. Put the stock OEM tune back on.
I guess I didn't word that correctly so I modded the pix
This is the sensor in front of the cat: https://www.boschautoparts.com/p/pre...sensors-17355-
The program sees WB, the ECU software can't work without WB and the oem part # is WB. So hopefully the wide band thing is put to rest.
Won't be using the oem tune. The limp modes are apparently torque calculation issues, not something actually bad. My transmission issues are not that it isn't working as designed, it's that as designed is what I or anyone who has driven a non-computer controlled trans would consider unacceptable. I recall ages ago a stick shift that would not allow you to shift from 1st to 2nd, you had to go from 1st to 4th. I think it was a Vette in the 90's and it was of course an emissions thing. Obviously people had a serious issue with that. Here I don't see the difference, it's a tranny that doesn't do what you want or even what a normal person would consider acceptable. I'm more **** about how a trans works so it's extremely frustrating that I can't change this one.
As for the limp mode, which is not really limp but a total shutdown, it's when some math in the ecu doesn't jive and is called a "torque calculation error". When you push certain things too far but don't adjust some other data to match, it doesn't jive and shuts down. The problem is HP Tuners program only sees some stuff in the ecu so apparently I can't fix that using HP Tuners, all I can do is try to work around it. But it's all like Hellen Keller trying to assemble a huge jigsaw puzzle because I can't see what is happening, why it triggers a shutdown and I have no idea what my changes actually do. So I make a change, then test it which takes 1-3 days. There are so many things to change and so many ways to change each one there are literally trillions of combinations. You also have no ideal how one change might effect another so even if you have one dialed in it may be ruined by one or more changes elsewhere. So while I've been doing it almost daily for a little over three months, I only have a small fraction of it figured out, and the rest is really just a guessing game.
A normal car wouldn't be an issue. I could tune a normal car to be where I want pretty easily but this isn't a normal car. I could request 50% more power and off I go, no problem. The problem is the more you ask the more complicated it gets. For me, I'm starting out so low that 50% more doesn't even get me back to the stock level so I'm asking for >100% more power and that's a problem.
I have accepted that I will never know what the actual problem with the car is but based on misc clues I believe it's a calculation error in the software. Either the math is bad or some data is bad, I can't tell. End result is the ecu sees a higher power output that is actually happening. So when it's making ~250hp or so it thinks it's making 400 and simply stops making more. Since I can't see the data that feeds into the calc, or the math itself, I'm screwed. All I can do is adjust what things I can adjust to try and compensate for the issue.
Ok, this all seems very weird. I haven't been able to read this very long article, but how 'bout this (in case it hasn't been tried yet - if so, please ignore)....let's try going back to the beginning. Go back to the absolute initial setting. Change one thing at a time from there until you see failure. Accommodate the failure from there.
Kanata: Thanks, but not that simple. Problem exists with zero mods so nothing to change and see.
jb123mb: Define multiple times. Can you be more specific, like at certain rpm or load, like putting around, freeway or just full throttle? Or how about what you're doing leading up to it, like is it more likely to do it when hot out or when you're harder on it, freeway or slower? Any clues to narrow down when...
Heat seems like a logical reason, and it's programmed to do that in several ways. Water temp, oil temp or air intake temp will trigger said programming. Maybe your intercooler pump died and isn't cooling the intake charge? The pumps last ~5 years, so people tell me. I suppose it could be cat temps too, if they were restrictive from age and/or deposits. Another, but I doubt it, could be the turbo solenoid if it's acting up when hot.
Then there's the turbo wastegates which see a lot of opening/closing, far more than a normal turbo and apparently it wears out the pivot points. That should an issue no matter what but the ECU does control the wastegates differently when cold vs hot, and again when too hot.
The easiest way to check what it's doing is log some data. The app Torque Pro and cheapo OBD dongle will tell you a lotta stuff. <$20 total so pretty cheap for the info. This is the dongle I use amazon.com/dp/B01HXGX8V6 and it works as good as ones much more expensive I tried, like OBDLink. I also have this which is very powerful for checking codes and misc things that most OBD scanners can't amazon.com/dp/B07QD4GZW4 It's also handy for clearing codes that most cheap scanners can't.
I use HP Tuners to collect data now. It's $400 for their dongle and you can read all there is available to read, unlike Torque Pro, and unlike Torque you can read all of it at once.
Then for an extra $300 you can make changes to the car, which I'm doing to compensate for my issue.
This is why I know there are several temp related ways it kills power. The oem settings start lowering power before the car is even warmed up, then lower and lower as temps rise. So I programmed mine to not start lowering until 220F water temp, then power drops very quickly. It sits between 180 and 215F under normal driving. I also use it to prevent the turbo wastegates from cycling so much, so now they stay closed 99% of the time instead of constantly moving. That mod alone is worth the cost of the program because turbo replacement is big $.
That reply was the short version via my phone while at work. Here's a much more helpful one:
My friend bought a Bluedriver for $100 ~two years ago because it was all there was for apple. It didn't seem like it was good for anything more than the $10 dongle, and if I recall you had to use their app which means it's much worse. It's possible there are other cheaper ones out now but I'd imagine the same problem with apps?
So I'd go with the $10 Android dongle and a work around, like a cheap Android tablet, cheap burner phone or a Windows laptop. The latter you can use a program called Nox to run Android apps on it: bignox.com which is free and easy to use.
If all you have is a mac laptop then you'll see on that same website there's a mac version! Whodda thunk it? So one way or another you can use the cheapie dongle.
Plus you may want to get that Autel dongle I linked above and it's Android only so you'd need to do a work around anyway. The Autel is a must have imo, but I get into more trouble than most so I need it. Still, think it's a must have for anyone.
There's the Torque Pro I suggested for $5, but also Torque Lite for free: https://play.google.com/store/apps/d...owl.torquefree
I can't recall what the difference was, just that I needed Pro. So it's very possible the free one will do what you need.
If it won't allow you to download the app (it happens if the device you try to download it with is not on the approved list), let me know and I'll get it for you. Let me know if you have any issues with any part of all this and I'll do my best to help.
After reviewing my driving habits, my issue seems more prevalent at speed, when i accelerate in the city away from traffic lights, in a spirited fashion, i don't have any issues, but when i am on the Interstate doing 70 mph and floor it multiple times, it seems to loose power the more i go through this process.
i wonder if it could be related to an under body aero sensor issue, telling the ecu to back off, due to a fault code?
Madman: No luck on data from anyone. The only info that has been shared is 0-60 and 1/4 mile. Their 0-60 #'s are much better than mine (best of 5.0) and I can't test 1/4 but I know what the answer is.
jb: I don't know what that sensor is but there are lots of limiters and a lot of temp related ones. If you're not electronically limited by heat, heat will kill power anyway.
For example, if that intercooler pump died the intercooler will work normally at first, but then the air just get hotter as the stagnant water gets hotter. It'll take a lot longer for that water to cool than it takes to heat up so when you lean on it while it's hot it's like not having an intercooler at all.
You really need the dongle/app so you can see all your temps.
Year later update: Car runs great after extensive tinkering via HP Tuners. There's very little info on how to use HP on a Benz, and what info I could find is just basic stuff to bump power in a car that doesn't have my issues, Like the OE Tune changes I saw in there, which do little to compensate for my issue. So I had to push it substantially further to get it to work, and make lots of other changes not normally done, and I'm guessing at it the whole time. Plus running into all kinds of obstacles because it usually didn't like me doing the things I did.
My 0-60 time is down to 4.5sec, which is apparently about equal to stock? I think that # may be AWD and I'm 2WD. I can't get traction to do any better, even with a set of wider soft racing tires, so not much I can do. After 60 it pulls good so 60-121 is 7.5sec which is much better than stock. So I can only guess at 1/4 mile but I rekon 0-121 in 12 is worth at least a 12.0 slip.
I have a Dragy but it's Bluetooth, and like all Bluetooth devices it won't work for me, so despite many attempts I have just one 0-60 run. Waste of $150.
Still all stock except the K&N filters. Only new parts are plugs n coils. Note; changing the oem spark plugs (pita) with 60K on them did zero, nadda, nothing. No power, no mpg, not better idle or anything whatsoever.
In the end it does answer the question about being a physical problem or not, and is clearly the ECU as I suspected. The mystery problem is still there, and still zero clue what it is, I just overrode it. I may be the only person on the planet with this issue, but if anyone else ever does, this was my fix.
I have managed to get some readings from other cars with my motor but they are not the same. One thing that is very noticeable is fuel delivery. According the data, I dump in a lot more fuel. About 100% more than stock and over 50% more than tuned cars. I can see dumping in a little more, but this is weird. It's not actually rich or anything so I don't know what to make of it.
Another is the turbo boost control, which on all others the psi is very steady, like rock steady. So steady I don't even know how it's physically possible considering the system. My psi is all over the map. It seems like boost climbs too fast, then it tries to pull it too aggressively so now it's too low, then overcompensates to bring it back up, and repeat. It's worse at lower rpm which I thought was killing my 0-60, but in reality 4.5sec is when everything goes well, when it doesn't I'm back to 5.0 or worse. It's usually wheel spin because it's very difficult to control, but when I do, it's 4.5.
Note; I'm using the HP Tuner software to tell me my time, and I know the data is delayed so I may be quicker? I just figured it's also delayed from the start so the 0-60 run should still be more or less accurate? Or perhaps there's more data as rpm/power increases but the same bandwidth so the it slows as I go? Too bad I don't have a Dragy.
Next is new O2 sensors, which I've had for a year but after crawling under it for a look-c I dread it more than the plugs. Especially when I know it won't change anything.
Also bought a Meth injection setup a year ago. I have no excuse for not installing it other than I spent any Meth time on tinkering with the tune. Not sure what that will do for me but only one way to find out.
Anyone found a solution? My CLS550 is doing the same thing, boost up to 8 psi exactly then it falls off. I do have a code for a P2540 low pressure sensor, not sure exactly where that is but could definitely be causing the issue and causing the ECU to pull power, planning on swapping the fuel pump and sender unit. I agree with OP of the thread, it is almost definitely an ECU issue and not vacuum or a boost leak. Have not gotten around to much testing but I have fixed the common leak in the vacuum diaphragm and that didn’t fix it. Xentry doesn’t report anything except for the P2540.
Nope...
I've never had P2540, even when fuel psi drops. You can also check to see if the eng is making what it believes is full power. Not sure if Xentry can, never tried since I have a much better program for that. When my ecu is oem it says it's making full power, and boost is ~7.5 at peak, then peters out to ~2 at 6k. If it were pulling power then it would show that in the #'s, but it isn't. Since it thinks everything is fine the only fix was to tell it to make more power.
Even with that OE Tune I had it was slower than stock, so that tells you how far off target it is.
Not sure if a tuner would make such a tune, so imo you gotta do it yourself. HP is ~700. Dongle $400 + $300 for access to one Benz. Even if you don't tune, the $400 dongle is sweet. You can watch everything from before start to shutdown, every 1/100th second of it.
If you have a physical issue, my #1 suspect is the vacuum pump check valve. It's easy to pull it for a look-c. One screw, not to hard to reach. Just be wary pulling it out because sometimes the valve falls apart when removed. Check if the diaphragm is in one piece :o A new one can be had on ebay for like $12, and may or may not make full power because the design sucks. I ended up buying a new pump with a good valve, so basically $100something for a tenny piece of silicone.
I've probably tried everything in my effort to fix this bs, so if you have questions, feel free to ask.
Nope...
I've never had P2540, even when fuel psi drops. You can also check to see if the eng is making what it believes is full power. Not sure if Xentry can, never tried since I have a much better program for that. When my ecu is oem it says it's making full power, and boost is ~7.5 at peak, then peters out to ~2 at 6k. If it were pulling power then
Thanks For the reply. I have changed the vacuum nipple on the pump and mine was shattered. My car scanner app calculated the engine to be making 450hp on a pull but it certainly didn’t feel like it, as the torque was only said to be 240 which seemed about right. My AFR’s sit around 13.5 on a pull, with the MAP sensor reading about 23 PSI peak. In my m113k car when it had an issue with the knock sensor it pulled boost completely, which is why I’m thinking this could be a sensor issue and is not a physical issue. Xentry found absolutely nothing except the p2540. I’m still going to check it for any leaks but I’m pretty sure I won’t find any. What exactly did you do in HPtuners to fix the issue? My buddy has the dongle already so that should save me $$$ and just need to spend the $300 for access to the car right? Thanks for the reply also, I really appreciate it.
What are you using to measure that 450HP, or did you mean torque? Torque will read that, but I've never seen HP unless it's tuned. Well, except for boost spikes because the ecu/hardware cannot control boost quick enough so sometimes you can a get a lot more boost for a brief moment. With my issue and low boost mine could go to like 17. It helped because it tells me my turbos are working and I don't have a leak etc. I think the best condition for that was locked in 2nd gear ~25mph?, then stab it. It usually does not work because the throttle opens too slow, but you'll know it when it works. My tires would break free for maybe 1/4 second before it kills boost.
It will pull power if it sees ping, and mine pings a lot on 91octane. It pulls spark of course, but also boost. When I correct the timing so it won't ping, or add octane, I get a bump in boost and power.
I judge how things are working by looking the Turbo solenoid duty cycle. Not sure what stock is supposed to look like because I never checked when it was stock AND running right, but my guess it would be closer to 50%? If it's at 90% (your max), then you know it's trying to make boost, but cannot. So a bad turbo, bad diaphragm, solenoid, check valve, leaking vac line, oil in the vac lines etc can all cause that. Or it may be trying but venting your boost out a blown gasket, or far more likely, a hose between the turbo and intercooler popped off, which has never happened to me but apparently it's common? Fyi a #*%!$ Rat chewed a vac boost line a few weeks ago and killed boost. So annoying, and illustrates that so many things can go wrong.
So I would check the vac line between the check valve and solenoid for oil. Since your valve was fubar it no doubt let oil in little by little, possibly adding up to a lot. It would mostly get sucked out with the new valve, but the question is did any get past the solenoid and into the turbo lines and diaphragms? I think it's unlikely enough can get in to cause a problem, but I've read it happens. I pull the line between the vac pump and solenoid, and blow it out. Then to check if the check valve is working, reconnect it at the pump and suck and blow on it (insert jokes here). Even the valve was new it could've fallen apart during install. Or a chinese one may kinda work, but not actually seal well.
Then pull the two turbo lines at the solenoid and suck on them and watch the wastegate rods pull closed. There is a spring the diaphragm has to overcome, but you can do it if you're good at sucking :p Then you should see the rod(s) goes xx far and stop. Each side should move exactly the same, and by the same amt with the same vac applied. So you can do both together and check that they open/close the same and same distance. I used a 1/4" tube from my RO water system as an extension for sucking.
Then there's the solenoid itself, which I consider a crude device and if I want to lose power I can simply install the "new" one I bought. So you never know if it's really good unless it can make at least the min vacuum it's supposed to, and flow well, which you have to have on the car and drive it to see. My new/crappy one works well enough to make oem boost, but it cannot make what I want. It also takes longer to make boost because there is less vacuum at all times, so turbo lag is greater.
Hello Chevota,
Pulled the vacuum lines going to both wastegates and checked them and they’re both holding vacuum fine, wastegates actuate fully and then spring shut again. The line between the solenoid and the vacuum pump had no oil in it, however I did have a crack in one of the rubber V shaped connector hoses, I glued it shut with some rubber cement and it sealed properly but I still didn’t hit more than 8 psi. My butt dyno says the car is pretty quick so maybe I’m just being paranoid and my car scanner pro app is calculating the boost wrong, I will borrow a friends Dragy and see what the 60-130 and 0-60 are. However the MAP sensor readings do align with the calculated boost readings.
My car scanner pro app calculated the instant HP and torque. You are right, the torque is very low, only around 250, and the HP seems right around 400 when the car is making 8psi. I’ll include an image of the graph produced by the car scanner app below. This is one of the highest boosts the car made at 8.5, and as you can see the AFR’s were perfectly normal indicating that the car is pulling the boost and not a boost leak as it would run rich right? The strange thing is that I don’t get full boost at all, my CL55 when it had a knock sensor wire chewed through was good for 1 pull to full boost then it would cut after that and I would have to reset it in Xentry to get the boost back for 1 pull, but the CLS has never once hit full boost.
Can't really compare the CL to the CLS.
The calculated boost is MAP minus ambient, so it's good. The power reading not so much. It should be doing the math based on the torque the ecu spits out, and rpm. You can't make 400HP with <250ftlbs. So I wonder if it's using the ecu torque, or is it using your phone to guesstimate? Either way it's fubar. Will that app show more stuff, like speed, throttle position, turbo duty cycle, spark timing and spark retard, maybe Injector Pulse Width? Because what it shows isnt much use.
Will be interesting to see the Dragy #'s. I'd prefer an 1/8th or 1/4 pass, but wth, do 0-130 and get it all
I don't understand the boost screenshot because I can't see rpm or speed. It says it took 10 sec to reach peak boost? All I can think of is you locked in a high gear on the freeway? When I do that I usually do it in 3rd starting ~2500, to get a better response from it.
My first app to read car data was Torque Pro, which I think it could record 20 things? Then it could output to .cvs to put into an excel chart. So all the stuff I mentioned, if you can, and post the .cvs.
Can't really compare the CL to the CLS.
The calculated boost is MAP minus ambient, so it's good. The power reading not so much. It should be doing the math based on the torque the ecu spits out, and rpm. You can't make 400HP with <250ftlbs. So I wonder if it's using the ecu torque, or is it using your phone to guesstimate? Either way it's fubar. Will that app show more stuff, like speed, throttle position, turbo duty cycle, spark timing and spark retard, maybe Injector Pulse Width? Because what it shows isnt much use.
Will be interesting to see the Dragy #'s. I'd prefer an 1/8th or 1/4 pass, but wth, do 0-130 and get it all
I don't understand the boost screenshot because I can't see rpm or speed. It says it took 10 sec to reach peak boost? All I can think of is you locked in a high gear on the freeway? When I do that I usually do it in 3rd starting ~2500, to get a better response from it.
My first app to read car data was Torque Pro, which I think it could record 20 things? Then it could output to .cvs to put into an excel chart. So all the stuff I mentioned, if you can, and post the .cvs.
Sadly can’t log those, I do think my friends RTD HP tuner dongle can log those so I’ll try and borrow it from him. The app is pretty limited but it’s just basic OBD so really can only get what the car gives it. My butt dyno says the car is a lot slower than my C400 was which was a 10.4 second 60-130 and a 11.9 second 1/4 mile car. The photos weren’t really useful, just showing some peak numbers and not a full pull.
The power numbers are calculated using the fuel consumption, I don’t exactly know how the app is doing it. But it is showing 450ft lbs of torque and 250hp, not the other way around. One thing I noticed is that the throttle position and relative throttle position never go above 80%, even when throttle is fully pushed past kickdown. My AFR’s are perfect at 14.7 and the timing advance is around 12° and doesn’t drop during a pull.
I am thinking that there may be a catalytic converter problem as I was under the car and they were unnaturally hot. I also noticed the LTFT for bank 1 is about 3% and the LTFT for bank 2 hovers around -2%, which is a bit strange. Not awful but still weird, which makes me think that potentially one of the cats is having an issue, although I would expect a p0420 or 430 for a catalytic converter issue. Might just punch them out and see what happens, I want to do it anyways if I get the car tuned. Need to figure out this boost problem first however. I’ll log on HP tuners with as many sensors as I can get and upload it here, hopefully you or I can spot a problem. Thank you for the help, I really appreciate it.
The RTD dongle will record all the data like mine. I don't understand those RTD things. Apparently the same price as mine but you can't make any changes yourself.
The fuel ratio doesn't richen up much at full power but should go to 13.5, maybe 13 but that's the oem limit if you have more boost.
Timing should be moving around so with that and the fuel not moving I'd blame that app and/or dongle.
My trims are all over the place, as are most I've seen. I wouldn't worry about it.
Do you want my HP Scanner settings? I geeked out on it, so might save you some time. It's attached. Just swap it for yours and you should instantly have my setup
The RTD dongle will record all the data like mine. I don't understand those RTD things. Apparently the same price as mine but you can't make any changes yourself.
The fuel ratio doesn't richen up much at full power but should go to 13.5, maybe 13 but that's the oem limit if you have more boost.
Timing should be moving around so with that and the fuel not moving I'd blame that app and/or dongle.
My trims are all over the place, as are most I've seen. I wouldn't worry about it.
Do you want my HP Scanner settings? I geeked out on it, so might save you some time. It's attached. Just swap it for yours and you should instantly have my setup
Hello Chevota, apologies for the lack of responses. After talking to a few guys on the M278 Facebook group, they told me that the correct boost pressure for the M278 in the CLS-Class, E-Class, and M-Class models is actually 8 psi, peaking at 12psi on some pulls. This M278 is detuned compared to the M278 in the S class, CL class, and SL class and such. From what I’ve gathered 8 psi is normal with 12 PSI reserved for those higher output 278’s which make way more torque than us. I didn’t find a single leak or issue with any of my turbos or lines and the car doesn’t throw a low boost code like I’d expect if it was actually low on boost. I punched the cats out and the car feels much stronger and healthier, and it now feels torquey like how I would expect this car to feel so I’m assuming I had a catalytic converter issue. Doesn’t mean I don’t have a boost issue, but from what other people said what my car does it perfectly normal, and the M550i my car was keeping up with serves to reinforce that notion. Maybe your car had the higher power output ECU in and something happened to cause it to revert to factory settings? No clue.
I've heard that about the boost, but not sure if I buy it. Plus, is that 8psi peak, or maybe what it's supposed to be across the torque range? I forget exactly but the claimed tq range where it basically flat lines at peak tq was huge, like 1600 to 5000 or something. That suggests to me that whatever peak boost is it likely stays at that psi until that highest rpm? Mine would peak ~7.5 (excluding spikes) for a brief moment, then drop fast. I don't remember what it was across the range but 6k was ~2psi.
To throw a low boost code it has to be very low on boost, and be low for 5 seconds straight. On mine it didn't give a code until I was showing the MB service manager that the car sucks. Then it threw a code again last month or so when a frikkin Rat chewed my vac line and I had zero boost, but again it had to be in boost mode for 5 sec so it didn't trip right away.
Good to hear about the cats. I often wonder how mine are doing, and wonder what would happen if hogged 'em out or removed them. Seems to be fine considering the power is better than most people with no cats. Plus I'm in California and they do not heed the 8th amendment. Getting caught hogging cats could ruin your life. Still very tempting tho... If I had a spare set of pipes n cats I'd totally try it.
I still don't know wth my issue is, but the power does vary a bit with time, apparently. I noticed with tunes that it varied a bit, but come smog time I had to use the oem tune and power was a little better than it was before. Still 7.5psi at peak, but at 6k I had 3.2psi. Not great but better than 2. And 0-60 was 4.9 instead of >5.
05-14-2022, 09:46 AM
