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.
Yeah I plugged right into the turbo inlet where the black line is. Pumped to around 11 on vacuum and also on pressure. Both held with no issues. Did the same with all the hoses so I'm thinking it's probably the check valve issue.
Now, the vacuum pump has two suction port.
1 big one for brake booster and 1 small one for turbo solenoid. Each has its own check valve
The brake booster has a test port, you try reading vacuum there and see if there is vacuum gauge needle shaky dancing or not ?
If you have shaky dancing needle at brake booster test port, it can mean that the big connector check valve maybe also kinda stuck, or your vacuum pump itself has a problem.
Now, the vacuum pump has two suction port.
1 big one for brake booster and 1 small one for turbo solenoid. Each has its own check valve
The brake booster has a test port, you try reading vacuum there and see if there is vacuum gauge needle shaky dancing or not ?
If you have shaky dancing needle at brake booster test port, it can mean that the big connector check valve maybe also kinda stuck, or your vacuum pump itself has a problem.
I tested at the other check valve. The smaller one on the left in the pic you posted. Have not tried at the brake booster....however I did have someone depress the brake pedal while I was checking vacuum and the vacuum actuator. The rods did nothing different and the shaking didn't change. Seems to me the brake booster is doing its own thing fine, but the other is having issues.
Then again, it very well could be the hose that goes from the solenoid to the vacuum pump, but since I can't manage to get it off in order to test it, I'll just have to assume the check valve.
I have not actually. I posted all of my findings. It's either the hose going to the pump...or the pump itself. However....since I performed all of these tests? The check engine light has disappeared and full power has returned. I have had no issues at all since. It's kinda strange actually. Idk if maybe something was just loose. But I'm kinda paranoid about it and still thinking of replacing that lil check valve since it's so cheap. Maybe even all the hoses. I'm not sure. But I haven't had one issue since then. I also haven't checked to see if the rod is still shaking tho
UPDATE: Got the check valve installed and also replaced the vacuum line that goes to it since it broke. Noticed quite a bit of oil inside the vacuum pump where the check valve goes and also on the check valve itself. Replaced both and wastegate rod has stopped shaking back n forth. They both pull and hold on startup and power has returned. I wont say this is the final update because anything can happen once its cooler out but so far so good. Ill report back if anything changes. Also, the oil in the vacuum pump concerns me, im sure itll need to be replaced soon.
As an update: my issue has been solved. There does seem to be a little hiccup on cold starts while idling for maybe 10 sec or so. But driving is normal and power is normal(tuned). The check valve solved my issue and was cheap to replace. Less than $30
i have a '14 GL450 with M278. got P0299 a couple of days ago. i just checked my wastegate rods on both sides and they are not moving. i am planning to check the check valve anyway because it is simple to check.
my next step is to check the hoses.
VW2MB, that will fit but it's chinese junk. The valve inside is simply lame. It may work on a stock eng, maybe, but it won't on mine.
I found a better orange one on aliexpress but still not as good as oem.
I had to buy a whole new pump to get a good valve.
i have a '14 GL450 with M278. got P0299 a couple of days ago. i just checked my wastegate rods on both sides and they are not moving. i am planning to check the check valve anyway because it is simple to check.
my next step is to check the hoses.
thanks
Yeah that's the video I found and the part I replaced it with. Works perfectly fine on my end
VW2MB, that will fit but it's chinese junk. The valve inside is simply lame. It may work on a stock eng, maybe, but it won't on mine.
I found a better orange one on aliexpress but still not as good as oem.
I had to buy a whole new pump to get a good valve.
It depends on the boost you want, but at some point it'll let you down. Have you checked it? Check at higher rpm too, because the chinese one is heavier and can't open or close as fast. There is also the cold starting issue where the gates move and put unnessassary wear on the hinge, which is the main failure point of these turbos. My guess is lower vac would allow even more movement.
i got the vacuum pump check valve today so i went and replaced it. i did it because even though my waste gate rod is not shaking visually the check valve replacement is the easiest and cheapest part i can throw at.
well, i wasn't too lucky. however, i noticed that the old check valve i pulled out was oilly. is it normal for the vacuum pump check valve to be oilly? does that mean that i have a oil leak in the pump?
i need to decide whether i should go ahead and replace the vacuum pressure solenoid or go and get my mechanic to do a vacuum line leak test. my CEL is lit immediately after the ignition so i feel like it is the solenoid sending signal to trigger the CEL.
If the pump has oil in it there is either a leak at the engine connection or oil is being sucked into the vacuum system via a vacuum line. I would start hunting for cracked, damaged or loose vacuum lines.
It's normal for it to be oily as the pump is bathed in oil and the valve is millimeters away from it, flinging oil all over it. The check valve, if working correctly, "should" keep oil out of the vacuum line because air flows from the turbo solenoid to the pump. Or maybe I should say very little oil should get in the line, and certainly never reach the turbo solenoid. If it reaches the solenoid then it tells me the check valve has failed.
That old oem check valve may be near death, maybe not, but I wouldn't throw it away if it's still intact. As long it's intact it's better than a cheezemo one from china.
There was mention of code P0299 for Turbo #1. I discovered this is the only code it will generate because as far as the ecu is concerned there is only one turbo. Say one turbo seized up, either one, you would simply get a low boost code for Turbo 1 because you cannot make boost on one turbo and the ecu can't tell which it is. All it knows is boost is too low. Also, it won't trip unless boost is 2.9psi or less for 5 seconds, and I'd imagine most people won't see that unless boost is extremely low to zero? It took mine over a year to finally trip that code because it wouldn't drop to that level until ~5krpm, so long before it timed out it would shift.
In my case there was nothing physically wrong with anything, it was the ecu commanding the low boost. I'm not sure but I may be the only one with this problem. Or there could be many of us but I'm the only who figured it out? I dunno, I just know I'm the only one I know of. There are many others with low boost, but usually they never report back with a fix, or anything at all. If they do, it was something else or they never figured it out and got rid of the car etc.
Ecu, or bad solenoid, or vacuum pump, check valve, diaphragm, vac line etc; all you'll get from the car is a Turbo 1 low boost code, if you get that. It's all there is and it's almost zero help. I'd ignore it while you try to find the cause. It doesn't hurt to have the code, unless you're having it smog checked. If so, clear the code a couple days prior and don't trip it again.
The cars are not nearly as smart as people think, and for me, all the codes are all but useless. I'd imagine there are many M278 engines out there with low boost, but unless they give enough gas for long enough to trigger the code, they will never know. Or it could be like mine where it's on the edge of triggering the code, but never does. Then when it does trip, it tells you nothing but the obvious, like a Weather Rock. When I had my low boost issue I even had the dealer look at it and the tech didn't find anything wrong and didn't trigger the code. After I complained, the service manager tested it and again found nothing wrong and also did not trigger it. I had to take the manager for a ride and trip the code. What was more weird, to me, is how could they drive it with such a massive power loss and not notice. So a car that is obviously hozed could be fully checked out by the dealer, twice, and pass. In fact, of the three times I've swallowed my pride and taken a car to the MB dealer, not once could they find the problem. Actually four times; a braking issue on the wifes car that the dealer screwed up and then couldn't figure out, but three different engine issues on three different cars with 0% success rate and each time the dealer gave up. Nice, make a car so complex that not even the oem certified mechanics and all that equipment and data can't troubleshoot. Where I work, if something fails like that, they'll do whatever it takes to find out why so it won't happen again. MB; sorry, don't know, now go away.
I don't know if 299 code will clear on its own over time or not, but if the CEL is still on, and bugs you, then just clear it and carry on. Low boost is something you're on own to troubleshoot and fix, and as mentioned, I doubt the dealer can fix it either. So you need to check the mechanical parts, and the best first step imo is by measuring vacuum to the turbo solenoid. The less vacuum the less your boost, more or less. I should say; you need a minimum amount of vacuum to make the oem boost, and the greater the boost and rpm, the more vacuum you need. So to really test it you need to verify you have enough vacuum available. Then check that the wastegates are getting it at ~4-6k rpm under full throttle.
One easy test you can do today is to bypass the solenoid and plug the vac lines together with a tube, or simply hold them together. At that moment, with the eng running, the two wastegates should close very quickly and with force. It also bypasses the ecu bs that causes confusing things like shaky rods and the wastegates randomly opening/closing.
If you're familiar with vacuum, put your finger over the tube to see if it's high vacuum or not. Don't drive it with the solenoid bypassed (yet) because it can overboost very easily, but idle is perfectly safe. If vacuum is weak, then it's the pump, check valve, or a leak somewhere.
To test further you need a vacuum gauge, and a T fitting, because you need to measure the vacuum between the turbos and solenoid to see how much vac they see. The solenoid is always bleeding it off so vac at the turbo side will always be less, but is vac good enough? You also want to see what the vac is doing while you drive. So you need to drive it and floor it while the vac gauge is on the turbo side, and note the #'s. The #'s at idle and putting around are basically useless, it's under load that counts and more load the more important it is.
My vacuum from the pump is something like 26"?, and the most at the turbo side 22"? I'm going by memory from 2 years ago, but pretty sure that was it. So 22" is the real # I have to close the wastegates, plain and simple.
I then bought a brand new solenoid to compare, since I found zero info online about what normal vac is. With the brand new solenoid the vacuum was never better than 15", and that I recall clearly, with a big wtf? The solenoid is a crude device, and imo getting a good or bad one, or somewhere in the middle is probably a roll of the dice. 15" might make enough boost to make oem power, maybe, I never tried, but it certainly wasn't cutting it for me. Plus, why would anyone want one that is barely working? It would always be on your mind if the power seemed a bit low. So check the oem solenoid, and keep it if it checks out, or get another one and again check it.
When first checking vacuum while driving I ran a tube between the T and the gauge, which I had on the dash. I simply used a little tube, like a vacuum line, connecting the two. You can run it into the open window to make it easy, but I routed it so I could have the window closed because I was watching it for weeks. I wanted to see it under all conditions to get a good feel for what it was doing. I also had a Torque Pro app to watch boost and solenoid duty cycle (DC). DC is how the ecu controls boost. This way you can see if the ecu is causing your low boost, or if the ecu is trying to make boost, but physically can't. The ecu has no clue what the vacuum is, btw, so you have to check that yourself.
Deciphering the DC takes some understanding, especially since the DC moves all over the place all the time, but generally a higher DC (max of 90) means the ecu is trying to make boost, and lower (min of 10) means it's trying to lower boost. But boost can't exist at idle so it may read 90 or 40 or whatever, and this is one reason it's confusing if just looking at your wastegate rods. It really only makes sense when it's at say 2k rpm or more and under some load. Basically the gas pedal position and rpm will determine DC, more or less. The throttle position and rpm gives a torque value for the ecu to shoot for, lets say 300. The ecu then tries to make that torque. If it's making less then it opens the throttle and/or increases DC to get there, but like a PID loop there are factors. Plus there are safeties that upset it and lowers DC or closes the throttle for what seems like no reason, and imo no reason is accurate for most of it. The point is it's hard to understand why the DC is what it is at any moment until you watch it for a long time and see a pattern. And the ecu isn't watching boost btw, it's watching calculated torque. It calculates that at xx rpm, xx spark timing, xx manifold psi, xx fuel delivered, I must be making xx torque. If that is less than the example 300 being requested, it opens the throttle more, and/or increases DC. There are some limiters to prevent too much boost, but from what I see they are not used at all under normal conditions because the torque control programming keeps it in the safe range at all times. Since your rpm and gas pedal position are changing, and the target torque changes with it, and the ecu and engine are both slow to respond, it means the ecu is forever adjusting the throttle and DC to hit a moving target that it is too slow and stupid to catch in the first place. Even at steady cruise it's moving all the time. Say you want 40ftlbs to cruise on the freeway, but it will drift because it simply isn't able to maintain that like a non-computer controlled car. So it drifts to 41 and pulls back. Damn, now its 39, give it more gas. Crap 41 again, and this repeats forever. The O2 sensors are the same as well, and they screw up the torque too. An endless PID loop of failure. I recall a statement about carburetors; "A device to give the incorrect fuel mixture at all engine speeds." Pfff, they clearly never met EFI. EFI is a carburetor with bureaucratic red tape to really screw it up. The boost control is, imo, at least as bad as the EFI.
Say you're cruising at 60 @ 1800rpm on level ground and load is 25%, or ~50ftlbs? Boost will be 0 and the DC will likely be 50 or less, but it could anything that nets 0 boost. Then you give it some gas, but not enough to down shift. The first thing the ecu does, for "safety", is drop DC, which is the opposite of what you want but it is what it is. Then when it sees everything is safe it will increase DC, maybe to 60, 70, even 90, who knows. Then as boost nears max torque for that throttle position and rpm, it will lower DC to lower boost and thus torque. It also does the same with your throttle, which is super annoying, so you have to watch both. Once time I forced the turbos to make max boost no matter what, and the ecu responded by holding the throttle at ~25% when I had it floored. It was not doing this to hit the target torque, it was a safety feature. Getting the throttle to open up under that condition is not easy, but I managed to do it, eventually. Since the ecu is slow to respond, you can overboost despite that safety. You simply floor it at the right moment, where the turbos can spin up faster than the ecu can respond, and overboost. Even when mine was not allowing more than ~8psi boost at best, I could get it to 17 if it hit it just right.
So bypassing the solenoid can overboost if you're not careful, and it happens fast. Will it hurt anything? Probably not, if you only do it once, but it will test your intake gaskets and all the inlet connections, hose clams etc. It will also stress your drivetrain and may break something that was on the way out anyway. I've overboosted many times and broke nothing, but I may just be lucky?
Bypassing the solenoid is a tool to check if the turbos are physically capable of making max boost, but that throttle will fight you, as mentioned. The throttle may simply limit power to oem levels, or it may go into safe mode like I described. Only one way to find out, but it will tell you if the turbos are capable of making boost. So to do it carefully, you just give it gas more gently and let it build boost more slowly. If you floor it it can overboost in a half second, if you hit it right. So use the app to see what the boost is at xx rpm and note how xx boost feels like. If it overboosts, you'll feel it because it makes power, lots of power. When mine was not making power it couldn't spin the tires to save it's life, but if I hit the gas at the right moment, in 2nd gear at ~40mph and get that 17psi before the ecu could stop it, it would break the tires free. So one way or another, I think you need to verify the system is physically capable of boost to eliminate from the list. Because sometimes the problem could be the turbo(s), or wastegates, or maybe clogged cats. Stuff you can't see but if you can prove the turbo can easily make full boost then you can rule all those problems out.
Hopefully I didn't make things more confusing, but I can better explain anything that doesn't make sense. Start with the solenoid bypass at idle, then the vac gauge, it'll tell you a lot. And to use the Torque Pro app you just need a cheapo OBD dongle like this:
so between that, the app ($5?) and gauge etc it's pretty cheap.
Using Torque Pro you'll want to add Engine RPM, Engine Torque, Accelerator Position D, Throttle Position SAE, Turbo DC, Boost Pressure, Manifold Absolute Pressure and Engine Load. I think there are three Loads to choose from. I think Absolute Load SAE, Calculated Load and Engine Load. Go with Engine Load. If not there, use SAE. Calculated Load is useless.
Optionally, you could skip the vacuum testing and just get the app and dongle to see if the ecu is trying to make more boost and can't, or does it think boost is fine? It's one or the other, and mine was the latter. I think for everyone else on the planet it's the former, but the app and dongle are cheap and you need it, or something to read it, either way...
Vac gauge hookup: Using a T fitting, which I can't recall exactly other than it was 1/4" OD male tubing, I plugged the T into the oem vac to the turbos, and used a short piece of vac line to connect the other side to the solenoid. I used a simple 2" vac gauge I got on Amazon for ~$10. Then, as mentioned, a line from the T to the gauge inside the car. Or I suppose you could mount it on the wiper blade, as long as you can see it.
With this test, and watching data via an app, I discovered my power loss/low boost was caused by the ecu commanding low boost. I never found out why btw. Since it doesn't look at boost, it was actually controlling torque, but boost as a result. From what I can gather it's calculating the torque incorrectly. So, apparently, when the engine is making somewhere ~250-300 ftlbs, it's calculation says 400, so it thinks it's at oem max and won't allow more boost. Since I was unable to find the cause, I simply reprogrammed it to make a buttload more torque until my actual torque was where I wanted it. Fyi in case anyone comes to the same conclusion.
Chevota, thanks for a detailed write up.
I just ordered a new solenoid which is the next part i can throw at before doing the vacuum test, which i was going to do by taking it to a shop but after reading your post, i might try to do some myself.
Well, if the solenoid is not the culprit, i need a next move so i might do what you suggest.
The P0299 most of the time is an issue with the actual physical wastegate not closing all the way. This is why it shakes, it's loose. Turbo bank 1 is passenger side and turbo bank 2 is driver's side. Remove the Catalytic converter assembly and visually check that the wastegate is properly working. If you find the wastegate is the issue, new turbo assembly will be needed.
I have 2014 s550 with p0299 code only, I noticed oil leaking under the car ver little bit still leaking, want to get input before I take it to a shop thanks in advance.
gladerider; any updates? I'm sure I'm not the only one who's curious about your issue, and fix. And I/we are here if you have questions.
I would check vacuum before/after the solenoid to check it, but it doesn't hurt to have a spare. Before/after vac check is super easy and will tell you a lot. If only doing it at idle, which should be good enough, just be sure it's fully warmed up. It's programmed to be weird when cold.
I went and got it checked, I noticed it had a broken plastic connected to three hoses, I got the part number and turned out to be something Called “cooling water change over valve” was told this feeds coolant to both turbos, they told me ima have to change both turbos most likely, but said to relace that part first and hope the turbos are still ok, I replaced that part and filled the coolant tank on the center under the engine cover…reset the code and drove it and still came back, is there any way since it says passenger side turbo in the code that the drivers side turbo still is good? Was also told I had a small leak on the radiator and the oil leaks is possibly the oil pan and gasket that need to be replaced. Was quoted $7,000 usd and to me that crazy considering I can get a new engine close to that price.
03-10-2023, 05:30 AM
