I have a 2016 E550 and boost psi is way low which is causing about 50% power loss. No explanation as to why. The ECU is commanding the wastegates to maintain <8psi, even down to 4psi. No CEL, no applicable codes, nothing seems to be wrong anywhere, yet here I am.
So I'm looking for an alternate way to control it, like using boost psi to manually (or electronically) control the oem control valve (part # A0081535428) which uses vacuum to hold the wastegates closed. Yes I replaced that valve just to verify for sure it was not the problem. I'd assume/hope people have done a mod like this to bump oem boost without buying a tune? I have ideas to make a mechanical or electronic mod, but it's rather involved so I'm basically fishing for optional ideas and any input from those that have done it, or hopefully an aftermarket one somewhere ready to go. Or better yet; someone who knows what my problem is...
I do not want to replace or mod the oem vacuum wastegate diaphragms, too much work imo but maybe if it's the only option. Plus messing with those would be all that much more work to revert back if I finally fix my issue.
Optionally I'm considering tricking the ECU by lowering voltage from the boost psi sensor but I worry the ecu will see abnormal pressure the rest of the time and freak out. So if someone has tried that, please let me know what happened.

I've been living with this issue for over a year fyi. Bought the car ~Mar '20 and put an OE tune in it a few days later. It was awesome for about two months, then it wasn't. I've been stuck with low power ever since.
If I remove the tune I simply lose more power. So without the tune my power is basically ~50% of stock, with a tune I'm ~50% of tuned power. Xentry shows nothing that I can see. Autel, nothing. Everything I can think of checks out. I'm guessing it's the ECU, which from what I've read is highly unlikely but highly unlikely things happen to me regularly. I'd replace the ecu just to see, but I can't find a used one and I probably don't want to know what a new one costs. One guy online "thinks" ~$5k but who knows. Lotta $ considering a new ecu is zero guarantee it'll fix it.
I posted about the power loss issue about a year ago, which some may recall, but the only thing new I learned since then is the ECU is commanding the wastegates to this lower psi.
My old post: https://mbworld.org/forums/mercedes-...igure-out.html

Thanks for reading my sad story. Please reply with the magic fix

 

I am no expert but this looks like a waste gate with changeable springs made for the 550 https://www.ecstuning.com/Mercedes_B...ger/Wastegate/
Could it be adjusted with vacuum and a ecu or just springs?
It would be interesting doing this way vs a tune. What is the stock boost suppose to be? using android car scanner it appears my stock 550 is only producing 8 PSI.

 

Sounds like you may have an issue too, maybe even the same issue. It's supposed to be 13psi, then with my tune it's some unknown amt higher. I don't know because it didn't last long enough for me to check. I assume 16-18 but just a guess.
Those in the link you posted are external wastegates, totally different deal.
One problem I have is if I tinker with the control valve or misc other ways, the ECU will simply command even harder to control boost, and it'll win. It's kinda tricky to fool it which may be why the mfg's are doing it this way now? So this is why I was asking what others have come up with because surely I'm not the first to try.
I was thinking last night and I figure bypassing the ecu altogether is probably the best plan. Using another oem boost sensor fed into a 555 timer, which I'd use to make a pulse signal (like oem) that actuates the oem wastegate solenoid. My guess is oem uses a pulsed duty cycle because otherwise the solenoid valve would likely stick? At the very least I think it would be sketchy and give different boost levels each time. It would be cheap to make, then a little trial and error to get it dialed in. Just have to drill a hole in the intake which I don't really wanna do, and find the time to make this toy.
Thanks for the reply

 

Well... you could try to re-engineer the entire thing and hope that you can outrun the ECU. Or, you could take it to someone who can troubleshoot the actual problem. You've got either a mechanical issue like a leak or a bad turbo, or an electrical one like a bad sensor or valve, or even a bad throttle body. It could be incorrectly sensing manifold pressure, or intake air temperature, or whatever else... lots of data goes into a decision to control boost. Maybe even coolant temp. And doesn't that engine have those electrically actuated tuning flaps in the intake? You don't know yet where or what the problem is.

Now mind you, I'm all in favor of fun electronics projects to do things on cars, but in this case you're effectively trying to come up with a solution worthy of Rube Goldberg to avoid fixing a problem. The fact that you haven't yet found the problem doesn't mean it's not there, it's obviously there. It's time to dig deeper or pay someone who can, there's no shame in it.

 

I do feel some shame in having someone else fix it, but I also kinda doubt they can. The only time I ever took a car to be fixed for something I couldn't figure out, which was another Benz, the dealer couldn't either... I had to Goldberg that car too. I am considering contacting another mech, but the last guy flaked and the one before that said he would need to troubleshoot all day, and road tests would be two mechs.Whatever $$$/hr x 8 did not sound appealing to get what is most likely a wrong or no answer.
No moving parts in the manifold, sensors all work. Turbos work, but they can't if the ecu opens the wategates. I can see them on a camera I had setup under the hood, so they boost, sometimes up to ~17psi, but just for a split sec until the wastegates open. Then the wategates are moving back n forth the whole time to control it to the lower psi.
One thing I can't test/replace is the ECU itself. So if the ecu is glitched in some way I'd never know. If it's just getting a signal it doesn't like and pulling boost for "safety", then it would explain my issue perfectly, but what signal? What could it be seeing that won't show up on Xentry or anything else I've tried?
I sure wish I knew what all signals could cause it to do this? It would of course help to locate the issue, or trick it, or disable that stupid safety function, but how can I find this info? Does it even exist?
Thanks for the re

 

You have to have something pissing off the ECU to make it limit boost. Watch all your temp sensors when it happens, is the intercooler system working properly?

 

All temps are good. Even when it's cold out it does it, or hot, doesn't matter. The air intake temp is high imo, but it is also normal which is from the very long metal intakes being nuked by engine heat. Need to wrap those things...
I just went into Xentry again to poke around for a couple hours and there was just some CAN bus glitch a couple weeks ago, which happens now and then. It was implausible data from traction system, which I haven't seen in a long time. Usually it complains about the pass headrest or radio, and if there is more than one they always have the same mileage stamp. So whatever glitch is causing that I don't know, maybe pot holes or something, but clearing them changes nothing.
So contacted a local mech after I posted earlier. I'll just have to sit there in his shop all day bored out of my mind, then shell out a bunch of $. I just hope it's not for nothing but I'm very pessimistic on the subject.
Meanwhile I have the 555 timer thing planned out in my brain if/when they fail to fix it. Makes me wonder if I can get away with similar to the throttle so it actually opens when I say to, but that has feedback so I'm sure it'll freak out. I really really wish I could just get into the program and change everything annoying, and turn off all the safety features. The "safety" features have directly caused some very near misses that would've never happened without them. Funny term "near miss"...
Thanks for the re

 

If I were troubleshooting this, I'd want to get some good data logging on the car and take it for a drive. I'd want to watch intake air temperature, throttle position, % load, manifold pressure pre- and post-turbo, fuel trim, knock sensors if the engine has them, misfires, and probably pre- and post-cat o2 on both sides, just for good measure. And I don't know about the M278, but mine has a setting for telling the car it's running on low octane fuel, so I'd want to verify that is still set correctly.

If the ECU is sensing more manifold pressure than there really is -- it will reduce boost. Seeing high IAT -- reduce boost. Got cylinder misfiring -- reduce boost. O2 sensors telling you the mixture is too lean and you can't adjust it -- reduce boost. Preignition or detonation detected by the knock sensor or coil pack feedback -- reduce boost. Fuel pressure is too low -- reduce boost. You get the picture. There is a ton of input data that the ECU uses to determine what it wants to see for boost pressure, and that decision is made in concert with mixture, spark advance, valve timing, and on and on.

Again, I'm not an M278 guy, but I have done a little bit of reading up on the M275. If the two use a similar scheme to manage the wastegates, then yes, it's a PWM signal to a valve that controls pressure to the wastegate actuators. So let's say you successfully design a device that massages the PWM signal to run more boost pressure than commanded. As you correctly guessed, it's not an open lop system. The ECU will simply command less and less boost to compensate, until it finally realizes that it doesn't have the control of the wastegates that it needs -- welcome to limp home mode, would be my guess.

The fact that the wastegates move at all tells you it's most likely not an issue with the PWM wastegate control output from the ECU. If absolutely nothing else is wring, no sensor is reading incorrectly, then it very well could be the boost pressure regulating valve itself leaking or obstructed. They're not terribly expensive, really none of the sensors are. But you won't know without some datalogging and a spirited drive, with some detailed analysis afterward. And if your mechanic doesn't understand this, it's going to take them a very long time to figure things out.

 

There is actually far more things the 278 can be upset about than on the 275, I agree though if you watch the correct data you will see what is happening.

I'm sure with proper diag you can fix it, going to a hacked up boost control solution, isn't a solution.

Saying "everything is good" doesn't mean much, what are your IATs like pre and post intercooler?

 

HP Tuners supports the 278 now, they will let you tune it yourself. I don't know how many of the stock limiters you can defeat with the software, but probably more than enough to blow it up.

 

Safety things I'd like to disable are not actual safety items, they're stupid safety items like; the car throwing itself in park while moving. Tire psi warning is set way too high. Dash lights are way too bright. Manual shifting is not shifting when asked, nor is does it stay in manual. It's not upshifting after flooring it and letting up. Brakes sometimes engage much harder than I ask, including when I don't ask at all. ABS is increasing stopping dist. Stability control often kicks in when it's the last thing I need. Eco mode needs to be off unless specifically requested. All these lights coming on when I open the door, turn the car off, or turn alarm on or off, they all need to stay off. Having to hit the brakes to start it or shift out of P. The automatic three flash turn sig needs to go. Plus others I forget at the moment.

I can certainly chart other things over time so maybe I'll do that. It's just a pita since bluetooth usually doesn't connect, and when it does I'm usually close to my destination.
The knock sensor trips only for a brief moment when I gas it, but I can only see it using Xentry. I don't suppose there's a way to add that to Torque/Charts? I tried looking for the PID # but no luck, and it's look very involved to find it on your own and it appears nobody has.

 

Temps are good meaning they are normal. So I put the few charts I have together here.
Water temps is normally ~190 if I recall correctly, but I rarely check it because again I called it "good"

 

All these tests here were done ~1130am, weekday, because it's the best time traffic wise. In this case it was 30 July. One is scaled 0-6200 so you can see RPM, then 0-290 for Power, then 0-90 for the other stuff.
Last chart is a bunch of throttle blips put together on one very lucky day.
I can do tests at lower speeds off the freeway, and if I do it right boost will flash much higher for a split sec, but I can't hold the gas down long enough for a good read. Generally 2 sec is about it and I have to hit the brakes. 3am is different but cops are everywhere. Since the data lags so bad, in those 2 secs you'll see boost but something stupid like 50HP because it missed most of it. I think all data collection here 1/10th sec but data comes through when it feels like it.
Misfires: per Xentry, zero ever apparently, until my battery went low a month or so ago and caused massive issues including tons of misfires. Which is another gripe because the batt should not cause that any issues. It happens to others too, so obviously it's programmed to create problems. My guess is so to you take it in for repairs in exchange for your cash.
I charted O2 sensors last year, and while I forget what they were exactly, they were fine and I blew that off like the temps. Same with knock sensors, which activate for a brief moment when first floored, but apparently that's normal. I assume that is how the system tests your octane, because how else could it know?
Mpg is like ~21, which is good considering the traffic. On the freeway it's ~25, 30 if I drive the speed limit (per the untrustworthy reading the car provides). The 21 is based on actual miles per tank fills.

 

Last one is a bunch of throttle blips

together

 

Need to see ITA2, the temp after the intercooler. Need to see it on full throttle runs.

This needs to go to someone who knows how to diagnose it.

All the complaints you have about the stock car, sound like you need to go back to something in the 1980s at the very latest. You will not find modern cars that don't have safety features, and some of the drivability things you complain about are normal in almost any car with computer control, like the holding gears after full throttle (the car assumes you are in a hurry, so it holds the gear in case you want to keep accelerating at max). If it immediately shifted up the moment you let off, then you stepped right back down and it had to downshift and get ready again, you'd be more upset about that.

Proper diagnostics starts with good validation of the fault vs normal operation, you need good data to make good judgements on what is your defect and how to remedy it, don't just throw parts at it or guess. Get a good log of all the info, Xentry shows most live data with reasonable scan speed if you have access to it. If not, I highly suggest taking the car to a knowledgeable tech before you get much further into it. Will be the cheap route in the long run.

 

So, what you really want is not a Mercedes at all, or you want one made before, say, 1965 or so.

 

That temp is after the intercooler.
I know why the car holds gears, I just think it's exceptionally stupid to think that scenario is anything other than rare, so why would they program it in? If someone wants to hold gear, you do that manually. Oh, wait, that doesn't work on this car either. The car is incapable of predicting anything so why force stuff like that on us? It can't even open the throttle when I ask so I have to floor it 1-2sec before I need it. Basically everything in it that is electronic is worse that old school mechanical. If they feel the need to make whatever stupid feature for the masses, then give us the option to disable it. What is so hard about that?
As mentioned, the two places with Xentry that I contacted blew me off. I did find one other guy quite some distance away, so maybe some day when time allows and I feel like wasting a day and a bunch of $ all at once.

An old car has other issues that make it even less desirable, and more $ to make desirable. Looks being the biggest factor, but there really isn't a fix for that with the exception of a very few cars from that era. This is the car I want and it took me years to find it, I just don't like stupid things being forced on me. Just because these things may help most people on avg doesn't mean it's best for all people in all locations. Like adding 300lbs of US safety crap I'll never need. If you wore 15lbs of body armor you'd be safer too, but are you willing to do that? Of course not, that would be stupid, so why would I want it in my car? Or why would I want my cars electronics to cause a crash? It reminds me of that stupid self driving car by Tesla, which apparently was created by the same people that did my car. Did Elon honestly think it would not kill people? Isn't he supposed to be smart?
Lets take stupid safety bs a step further; say next year cars are to be speed limited via gps so the car cannot exceed the speed limit for your exact location no matter what, and will never allow you to accelerate any quicker than say 0-60 in 10 sec. Would you be ok with that?
So apparently I just have a lower threshold for BS than you guys.

 

Spend your money on an hour of dynamometer time for your car and then just make run after run logging data. Then peruse that data and analyze at your leisure without having to worry about being on the highway.

 

That sir is an excellent idea. Maybe with any luck I can find a performance shop that does Benz who may have experience with this particular issue. After all, I'd bet part of my problem is the vast majority of owners and mechanics would never even know this problem existed because there are zero clues except power loss. It still has the power of a "regular" car so I bet it would even pass a test drive at the dealer if I were to trade it in. So there may be a lot of them with this issue but maybe only hotrodders like me are aware of it. I would imagine hotrodders are 99% of dyno shop customers?

Meanwhile I'm trying to chart O2 and misc other things you guys mentioned but was unable to get bluetooth to connect today. Maybe Vediamo or DTS Monico can chart data. One of which may even hold the key to my issue, if I knew how to use them.

Thanks for the re's

 

Yep, I was often the only one who noticed customer cars with dead IC pumps, because my tuned 600 ran like a raped ape, and the ones with **** intercoolers felt like a dog by comparison. Once I straightened them out, every customer was exceedingly happy.

Did it for a guy at Rover who bought a trade in G63, he took it and had it tuned to "700hp". I drove it, felt it pulling boost/timing. Changed the plugs because he had an insulator coming loose, and I properly filled and bled his IC system, didn't say anything, just gave it back. He came back saying it felt like a whole different car, the plugs must have picked up 200hp lol. I had to explain why it was so down on power before, and it's a shame someone was willing to flash a tune on a truck with a non-working IC system.

 

So frustrating.... The Dyno shop nearest me, one city over, in business for >40yrs, said they can't help. After I explained what it was doing, he said without any codes they wouldn't have a clue what it is. He also added that he's never heard of a car doing that unless there were codes to go with it. He passed me to another guy, who apparently does custom dyno tuning, who said the same; No codes, good luck. I may have mentioned earlier that I've only had one code that was accurate, any car ever. That's excluding stupid stuff like a light bulb out, which for some reason it sees perfectly. Otherwise it's all false codes, or no codes. Makes me wonder why I've always had this issue because it seems everyone else gets codes that tell them everything they need to know.

As for an O2 chart; Torque/Realtime Charts can only see two sensors, not sure why it can't see any of the others but I also wonder what two it's looking at? Better than nothing but still have to get bluetooth to connect to see those. This past week was a fail, despite getting it to connect twice. I guess I shouldn't be surprised but both those times, once I got to where I'm going and I checked the data, I see it stopped seconds after starting. So I only have a few seconds of idle data to show for quite a bit of time trying.
Autel, which can see the sensors, is even more difficult to get working. Then you can only see two sensors at a time? Maybe it was three, I forget, but its issue is what it's willing to display at a time, and since it won't chart them it's of little use. Xentry is similar, can't see all the sensors at once and no recording/charts. When I did O2 last year I simply held/watched the Autel screen as I drove/floored it, which is a little sketchy. How do you people chart this stuff????
So I'll have to figure out a way to pull this off, plus do the fuel psi etc. I figured I'd just use a screen recorder on my PC, a poor mans chart, which I just discovered sucks >60% of my CPU while running so hopefully Xentry will still work right. May have to use my dash cam instead. Funny how high tech everything is, yet still so painfully difficult and often a fail.

Now I hear the EPA is going to crack down on ECU tunes, so my already slim chance of finding a replacement ECU will no doubt get 1000x worse as tons of people fight for a spare. An ECU was (imo) my last shot. At the very least it would eliminate the ECU as the cause. It would also open the door to my fantasy of reprogramming the stupid thing to do what I want. A huge stretch for sure, which is why I said fantasy, but who knows.

ItalianJoe; I see now why you had all the intercooler questions. I wish it was that simple... I had high hopes the dyno people would be like you, but I guess they're just automatons; "No codes. Does not compute" (smoke starts rising from ear)

 

If you were in south FL I'd say bring it to me to straighten out, but without seeing good data on what it's doing, hard to say why the ME is pulling power out. I'm sure it's doing it on purpose, so replacing it and re-tuning won't change it if you don't fix the parameter that it's seeing that is causing the reduced boost level. You could record in Xentry, I don't remember how exactly, but a graph function even across the screen would be helpful. Are you 100% sure your intercooler system is working correctly? Pump, coolers, full of fluid, no air? It's not the best stock, adding the heat load of more boost makes it even worse, and the tune may pull it back to lowered boost if it's too hot for safety, I know I would set up things like that in tunes I did for people.

 

I hope you're right about it doing it on purpose, which means it's fixable. What I do not understand is it doing something so drastic but no mention of it via Xentry etc. It seems to record everything else, even hitting the brakes hard is recorded, but it cuts power in half and not a word.
The intercooler pump motor is new because last year I thought maybe it might be hurtin despite the car only having 29k on it, but no difference. It now has 42k. Tank is full, water flows, temps match what others with the same motor are seeing. It also doesn't matter how cold/hot it is outside, or if the motor is cold, the power never changes.
I actually got bluetooth to work for 11 minutes today, but unfortunately all it recorded was fuel psi. Lower psi is idle, the rest is everything off idle to full throttle and rpm and everything in between. The internet says it varies from 1740 to 2900. I checked psi last year too but I can't find that old data. Not sure what to make of it being <2900. I'm guessing this is of no use but Torque/Charts apparently won't do fuel trim.
I did find an old Autel report that says long term fuel trim is -2.3 on banks 1 & 2. Not sure how long long term is, but fyi. I assume that means its shortening the injection duration by 2.3% compared to the fuel map. If so then it would also be safe to assume clogged injectors or low fuel psi would go into the + side. With the low mileage I don't believe inj, cats or anything else is hosed. Electronics I believe...
Meanwhile I'll work on Xentry screen recording of fuel trim and O2. Thanks for input btw. I'm super annoyed at all this but appreciate the input/help. And yes I'd bring it to you if you weren't 3k miles away.

Spent hours Sunday reading up on Vediamo, but all I really got from it is a headache. Lots of acronyms that I don't know wtf they mean, and instructions leave out important details so I guess it was written for people that already know how to use it. It also seems there is nothing relating to I want to do, or maybe I just have the wrong reading material. It did mention some things I want to change, like how the throttle input moves the actual throttle blade, but you only have a few facotry preset programs to choose from, nothing about making your own. Imo it should be simple; 0-100% pedal movement = 0-100% blade movement, and no lagging or other weird crap.