Looking for any assistance to diagnose a CEL P0087 low Fuel Rail Pressure on E550 M278 with 90K. Owed the car since 20K and still love it. Still get great mileage and have not had any other issues starts right up.

Car will throw the code on heavy acceleration (put the pedal to the floor) turbos kick in pulls hard up to 5000RPM and she then goes into limp mode. If shut down and restarted runs fine. if babied will run fine for quite a while

How can I check the Fuel Pressure sensor on the fuel rail?
Where can I put a T in to check fuel pressure I have not found a place to tap in?

I replaced the following due to mileage, age and low pressure CEL P2540 hoping it would correct the issue
The Right and Left Fuel Pump Assembly - Mercedes-Benz (218-470-00-94) and
the left Fuel Sender Unit - Mercedes-Benz (218-470-09-94)

If fuel high pressure regulator fails would it begin to throw off the engine’s air fuel ratio, which can lead to engine misfires?

 

See attachment to get a better understanding of fuel injection system.

 

Are you getting a corresponding misfire code on any one cylinder? I've always had this come up in the past on other (non-MB) vehicles when an injector was failing and sticking open. It's happened often enough to me that a bad injector is usually the first place I look, especially when you describe the circumstances.

 

no misfire the only code that keeps coming up is the p0087 nothing more

 

On the M272 and M273 there's a black cap on the fuel rail (passenger side bank, forward end of engine), which allows for connection to monitor pressure or performance injector cleaning. Must be a similar service port for the M278, just your dealing with orders magnitude higher pressures (60-ish vs. 3000 psi).

Also something to consider, besides a faulty pressure sensor, is this a low-pressure issue (fuel delivery) and/or high-pressure issue (HP pumps)... konigstiger's doc has good background info.

Ideally you could monitor the pressure via scan tool and verify with mechanical to rule out onboard sensor. Then its a matter of why the low pressure.

 

If you've got the right scan tool you can monitor all that in real time. I can't say for sure which tools do that, I have the iCarsoft and I know it reads a bunch of parameters but I can't recall offhand if fuel rail pressure is one or not.

 

69,

OBDLinkLX bluetooth with torque APP confirmed ( I use it ) to be able to read Fuel Rail pressure from car OBD2 port. M276 3.0 TT is my engine.
On mine at WOT fuel press from 188 BAR goes down to 150 BAR average.
I only have 1 HP pump, not 2.

Find attached my data log of a test drive which you can study fuel rail pressure vs RPM vs throttle opening etc etc.
It is using Banks Gauge called Data Monster via OBD2. This one is a powerful logger and can log 100s of PID, if only 1 Hz speed.
NOTE : There is something wrong with Banks Gauge firmware for AFR data, it is opposite... LOL. Rich become Lean and vice versa.

Hope it can help you.


Easy view

 

Manual gauge on fuel rail test nipple idle has 55 to 60 PSI when i stab the pedal goes averages 65 PSI but has jumped at times to 70 PSI

I used a OD2 scan tool today and now get a second cel P2540 watched the high pressure rail
The document provided by konigstiger thank you that states the M278 at wide open throttle should be 170 bar which it looks like it goes there it now seems to take to long sometimes to recover when floored dropping to 1650.

Now I have no idea if what I see is correct ???

 

Please comment based on the charts below taken with Torque Pro How I understand the normal cruise pressure is suppose to be around 170 bar with a M278 and will dip down then recover on hard acceleration is this correct?

Is it recovering correctly?

Is the charts showing a intermittent failure on the 3rd stab?



Note: I have tried 2 new low pressure fuel pumps both throw a P2540 which is low pressure warning along with a P0087

the original pump only throws the P0087 code the issue shows up when I slam the pedal to the floor


Original Factory Pump with 90k of use throws P0087 code



First new replacement pump immediately throws a P2540 and P0087



Second new replacement pump immediately throws a P2540 and P0087

 

Do these rpm fuel pressure charts help?

Other questions that can be meaningless

Does the low pressure pump get most of it fuel from the small rubber valve on the bottom of its housing or the overflow pipe?

I ask because The new sending unit overflow pipe is about an inch shorter than original the original pipe ended in the center of the low pressure pump housing the new one ends right at the edge of the housing I would think at hard accel all of the fuel is not going into the middle of the housing

It feels like the low pressure pump is starving on hard acceleration. Can the fuel being shifted to the back of the tank cause this?

 

According the excellent document from konigstiger the limp mode is actuated when the high rail pressure fails I guess to maintain some set pressure level. This can happen because there is a pump problem but you have replaced both so pumps are (were not) not the likely problem.

The other reason for low rail pressure could be too low pressure from the low pressure pump in the fuel tank but you have replaced this twice and now get the other trouble code P2540 with them both on top of the only code P0087 you only got with the original low side pump.

Then, of course, there could be a leak but that would be easy to notice by just smell of gas.

There are two different fuel pressure fault codes. Is one of them (P2540) for the low pressure side and the other (P0087) for the high pressure side?

Anyway, by the document, the limp mode is caused by the high pressure side being too low so the low side reading should only set the trouble code and possibly lit up the CEL.

So, after replacement of all pumps, low pressure pump even twice, I have hard time to believe the low high pressure side condition is true. I think it is a reading issue at the sensor B4/25. The low reading from this sensor sets the limp mode and the code P0087, I believe.

But, there is the code P2540 too. If this is caused by the too low pressure reading from the fuel tank pump sensor B4/7 then it is a strange coincidence that both pressure sensors fail at the same time, but it is possible.

Would be interesting to see what this pressure reading does together with the high pressure side pressure. It should be possible to plot the same way as the high side pressure, I think.

It is also possible to connect a pressure gauge in the low pressure side service port 6 to see what the pressure does and compare to the electronic reading from the pressure sensor.

If it proves out that the low side electronic pressure reading is wrong then it could be due to bad sensor, of course, but as it after all component changes is likely that the reading from the high pressure side is also wrong it raises a question about the voltage feed to the sensors. I don’t know if both sensors are fed and read by the same module. It likely is the case so could it be a problem with failing fuel pressure control/monitor module? It could be failing but not setting any code for itself but causes wrong readings from the sensors.

As a next troubleshooting steps I would plot the two pressure readings together to see what the low side pressure does when high side changes with engine speed. I don’t think low side pressure should change hardly at all.

Then, if thinking the two pressure sensors are going bad at the same time then replacing those should fix it but as I said earlier it would be quite a coincidence for that to happen.

…just my thoughts…

 

Arrie

I also just finished putting a Manual gauge on fuel rail low pressure side service port 6 and read the following 55 to 60 PSI at idle. Stays at 60 PSI in gear not moving. when i stab the pedal for a extended period 4 to 8 seconds it will jump some times but settles in and averages 65 PSI. It has never dipped below 55 at idle and never dipped below 60 in gear

Yes I get both (P2540) for the low pressure side and the other (P0087) for the high pressure side now

The B4 low pressure sensor on the fuel sender unit was changed 3 times as of today no change

Would the sensor B4/25 on the high pressure side rail cause a P2540? I can see it causing the P0087

I saw a video that showed a good sensor should read pin 1 to pin 4 128k ohms pin 2 to pin3 infinite pin 4 to pin 3 38k ohms and pin 1 to pin 3 12k ohms


Has anyone tested the B4/25 sensor Bosch 2789050100 with a meter?

 

55 psi should be plenty pressure to supply fuel to the high pressure pumps. I connected my iCarsoft scanner yesterday evening and for a big surprise I could not find the low side pressure reading in anywhere with it. I found the high side pressure and at idle it was reading around 119 bar. When I gave it some small gas pedal bumps the pressure max was 124 bar and min 116 bar. I did not go driving it so I don’t know if it would go to about that 170 bar that it should go to.

Also, my scanner does not show any kind of fuel supply module when I go to the “BENZ” and select my car and engine. I can read live data on the “OBD” side. I wonder if there is a problem with my scanner. I do have the latest software in it though.

Anyway, another thing I checked is the supply voltage to the fuel control unit. It read 12.5 - 14.3 V during my bump test.

Because it seems you are getting low pressure from both sensors I would really look into the possibility of the control unit going bad or voltage supply to the control unit not working properly.

But hey, you have changed pumps numerous times so changing the low and high side pressure sensors should not be that big of a deal to do on top of it. If that would not solve the problem then it would really point to the control unit.

 

Arrie,

My icarsoft V2.0 also can not read low pressure fuel pressure or temperature of high pressure fuel in the rail


69,
Below looks good in my opinion



Since the pressure sensor is downstream of HP fuel pump , I think the HP fuel pump and its built in pressure regulator is okey since the value can be read and a good pressure value it has.
However the B4/25 pressure sensor is also supposedly a temperature sensor. If the temperature sensor section is defective either by its element or wiring issue, the code you are having may pop out.
Remember, OBD2 is generic and it is not Xentry, so one can't swallow 100% those OBD2 codes.
B4/25 is a 4 wires/pins sensor. Ref 5V (pin4) , signal pressure (pin2) , signal temperature ([pin3) and ground ( pin1) [b]<<< Pin numbers based on N10-3 ECM my engine version.
Becareful when you messing with 5V reference sensors.

 

I read the data again today while driving the car. Fuel rail pressure went to 169 - 171 bar. It stayed at just below 170 but when pressing hard on the pedal pressure went to 171.2 bar at one time. At idle it went down to 120 bar again.

 

I want thank everyone for your help.

Here is the fix and diagnosis for the M278 CEL P0087 Fuel Rail Pressure Low / Fuel Pressure Sensor and P2540 Low Pressure Fuel System Sensor Circuit Range/Performance

The Pressure A Temp Sensor B4/25 was also causing the P2540 CEL Intermittently this is what hung me up

The fuel sender (the fuel filter assembly A2184700994, 2184700694, 2184700194 which includes the low pressure sensor) and the Pressure A Temp Sensor B4/25 part number 2789050100

With a warm engine the right side low pressure pump will provide range of 55 to 65 PSI to the 2 high pressure pumps. Your high pressure rail will measure approximately 11900 kpa (119 bar) in gear if you put your foot to the floor it can jump up to 22000 kpa (220 bar)

Here is how you can troubleshoot the Pressure A Temp Sensor B4/25. It is so quick and simple I wish I went straight there but the low pressure DTC threw me off. You can remove the sensor in 10-15 minutes it is located on the passenger side on the high pressure rail under the air filter.

Below is a comparison of measured values from the old defective sensor and the new one using a ohm meter set to 2M Ohms

A good sensor will have the following values pin 1 to pin 4 you will have a reading around .562 M pin 2 to pin 3 should be infinite / NC. Pin 1 to pin 3 will read .007 M and finally pin 3 to pin 4 will measure .572 M again these are approximate values each sensor will vary a bit.

The bad sensor read as follows pin 1 to pin 4 you will have a reading around .022 M pin 2 to pin 3 was 1.099 M Pin 1 to pin 3 will read 1.135 M and finally pin 3 to pin 4 will measure infinite / NC again these are approximate values.

The car is like brand new the rear end now breaks lose with ease my sensor must have been cratering slowly making the performance loss undetectable

Cheers

Crunch

 

Great news 69. After all work it had to be the sensor(s) causing the issue. Luckily it was not the control module...

Happy "fish-tailing"...

 

so it is the temperature section of the B4/25.
This is a lesson learnt about generic OBD2 code and common OBD2 scanner. It is made for many cars in general.
There is a specific code for it in OBD2 : P0180 is a generic OBD-II code that is triggered by a malfunctioning fuel temperature sensor 'A.'
I wonder, is this MB own doing that for generic OBD2 data reporting, MB won't issue a very detailed one and with Xentry, MB will issue a full blast rich details ?

OBD2 trouble codes are produced/reported by ECM and OE has the rights to set their own parameters where the threshold should be at and what trouble code should be issued on the OBD2 port.
OBD2 is primarly an emission "see-thru" for general public , initially enforced by California in the 80s I believe.

ECM knows the fuel temperature reading is too high or error, hence LIMP mode executed for safety.
ECM can use Engine Coolant Temperature and the calculated engine oil temperature as 2nd logical references or cross check.
As to how stingy or how generous MB would allow its OBD2 reporting, which can make its own Xentry semi useless or desperately needed..... that is for us to find out.

69 wrote
Below is a comparison of measured values from the old defective sensor and the new one using a ohm meter set to 20M.

A good sensor will have the following values pin 1 to pin 4 you will have a reading around .562. pin 2 to pin 3 should be infinite / NC. Pin 1 to pin 3 will read .007 and finally pin 3 to pin 4 will measure .572 again these are approximate values each sensor will vary a bit.

The bad sensor read as follows pin 1 to pin 4 you will have a reading around .022. pin 2 to pin 3 was 1.099. Pin 1 to pin 3 will read 1.135 and finally pin 3 to pin 4 will measure infinite / NC again these are approximate values.

I have to assume 69 data is x1000 ohm. What is not stated is, the reading is taken with hot engine or cold ambient temperature engine at fuel rail ? This is a thermistor, so temperature of the sensor tip must be known when reading takes place.
So good temperature sensor pin 1 to 4 = 562 ohms. Bad sensor = 22 ohms ......... at X ambient temperature and assumed fuel rail is already at that ambient temperature.
Usually ( not always ) temp sensor in an engine is a negative coefficient type, less resistance reading = hotter. It is not linear, so must look at its temperature curve.
So probably this sensor reported some super high temperature well beyond 130C - 150C, if the values read is referenced to typical Bosch or Delphi 2.5K ohms at 20C thermistor.

Delphi : https://pe-ltd.com/assets/coolant_temp.pdf
Bosch : https://www.bosch-motorsport.com/con...or_NTC_M12.pdf

 

S-P,

Are you saying the wrong fuel temperature reading can also cause the Limp Mode?

That document that konigstiger posted does not mention fuel temperature for it. The only thing mentioned is that Limp Mode is set if the fuel pressure does not reach the value it should reach. This value is at around 170 bar and perhaps a few bar higher with WOT. In my car (2013 S550) fuel pressure is at around 120 bar until the car is driven when the pressure immediately jumps to 170 bar. I’m thinking the OP’s problem was the pressure reading not going up or not maintaining it at 170 bar with WOT. If the new sensor fixed this then it obviously was a bad pressure sensor reading.

 

Yes Arrie, I believe so. In my opinion it was the temperature section of the B4/25 sensor which is more responsible for the trouble code, in comparison to the pressure sensor section.
The pressure side doesn't look so bad for the new HP pump, not the greatest though...agree and if it is based on minimum 1 hz data logging and represented as 1 hz graph.

I have to correct the way I interpret OP ohms reading of my previous post : I now believed it is Mega Ohms value and not Kilo Ohms, also I misread the pin numbers
NOTE : I am kind of confused with OP resistance readings. In todays auto ranging digital DMM resistance function, why would one choose 20M ( 20 mega ohms scale ) ?
Secondly, is there such a scale today 20M ? Its either Ohm, Kilo-Ohms or Mega-Ohms. I think he meant Mega Ohms is set manually and his meter can only read to 20 mega ohms ?? 69 can you verirfy ?
Lets assume Mega Ohm scale is being used.
Let's assume ECU is removed/disconnected when measurement made, or measurement is made at sensor pins directly, hence no inteference from ECU 5V voltage regulator circuit. 69, can you verify ?
Good sensor Pin1-ground to Pin3-Temperature sensing = 0.007 x 1,000,000 = 7 Kilo Ohms. <<Make sense.
Bad Sensor Pin1-ground to Pin3-Temperature sensing = 1.125 x 1,000,000 = 1.125 Mega Ohm <<<< defective sensor reading , as though as cooler than -40C / -39F


Today out of curioisity I just verified, B4/25 is a 10K ohms at 25C and is a NTC ( negative coefficient ) thermistor.
This is 99% certainty, a 100% until I can actually find the temperature curve of the said sensor to be able to cross check 1 Celcius accuracy.

TEMPERATURE SENSOR SECTION OF B4/25
Probed at Pin 1 and 3 B4/25 wire harness's connector. ECU removed/disconnected.
Temperature : T2 is reading at sensor metal body ( fast). T1 is reading at fuel rail pipe ( slow ). Response in Ohms reading is not immediate, because B4/25 sensor tip temperature is immersed in gasoline.
I need to use 2 thermocouples to take their average as the B4/25 sensor tip temperature.



TEMPERATURE SENSOR SECTION OF B4/25. Heated with hair-drier.



TEMPERATURE SENSOR SECTION OF B4/25. Cooled down with cold water. I believed only 23C at best at sensor tip and not 21C.



Typical 10K Ohms at 25C NCT thermistor temperature curve
https://www.bapihvac.com/wp-content/...stor_10K-2.pdf
https://www.dwyer-inst.com/PDF_files...TURE_TABLE.pdf









The pressure sensor is guaranteed a piezo type, because it is slim shaped and its high operating pressure.
What we don't know is, if the piezo is resistance based output or voltage based.

TEMPERATURE SENSOR SECTION OF B4/25



Pin 3 to 4 is no use to read for any resistance reading.
Pin 2 to 3 is also no use to read for resistance reading.


PRESSURE SECTION OF B4/25
Pin 1-Ground to 2-Pressure Sensor sensing can be read, if the piezo is resistance based. I measured it anyway for fun and get 274.9 Kilo Ohms if at approx 35 Celcius.



Heating the pressure sensor to approx 47-49 Celcius. I get 249 Kilo Ohms or drift of 74.1 Kilo Ohms at +13 Celcius temp rise over the 35C. Piezo does have DRIFT when exposed to changing temperature.



The thermal drift characteristics of piezoresistive pressure sensor
https://core.ac.uk/download/pdf/82013807.pdf


.

 

S-Prihadi

Sorry for the confusion the flaw of a old electronics tech. I assumed everyone would understand what the "M" stood for. For those who do not know "When a "M" is used for a ohm meter reading that would stand for Meg Ohms. If measurements are taken in Kilo Ohms it would be a "K" and for measurements below 1K it would be just Ohms" If you have a auto ranging ohm meter this may be different due to the logic built into the meter making that decision for you.

Now to answer the questions


Measurements were taken at room temp 75 F directly on the sensor pins with the meter set to 2 Meg Ohms (correction for above meter was set to 2M not 20M)


The opinion that Pin 3 to 4 is no use to read for any resistance reading. Pin 2 to 3 is also no use to read for resistance reading and the 2nd logical reference. Either one of these 2 readings alone 100% validate a bad sensor.


The reason I measured at 2 M was to uncover any flaw in the sensor and did not switch to Kilo ohms for other measurements was to standardize measurements to prevent confusion which I guess did not work. Where this works for example the measurement from pin 2 to pin 3 (the logic pins) should have NC no connection between them. If this was measured with a meter set to kilo ohms it would not show this defect and would appear ok. Just this one measurement alone tells you the sensor is bad. low impedance on Pin 3 and Pin 4 again would tell you the sensor is bad


I don't think the ECM could use engine coolant or oil temperature for a 2nd logical reference or cross check when the temperature is being based on fuel pressure. Pressure is directly proportional to temperature. If you increase the fuel pressure and compress the molecules, they start to move faster thus raise the temperature.

I am 100% certain this failed due to heat. The discoloration yellowing of the plug shows the signs of heat. I wish there was a way to remove the heat away from the sensor and rail.

 

69,

Fuel is liquid, it won't get any significant temperature change at all when compressed, because liquid is incompressible.
Air yes, it will get hot or very hot depending how fast we compress it and to how many times its original pressure.

I used to use this kind of Sanwa multimeter in the 80s, we got to zero its Ohms value before use . So I roughly know what you mean when you stated the DMM setting as 20M ( 2M corrected ).
I do not know how your meter shows the data on its display though.


When we state a value of the resistance we best write X.XXX Ohms or Kilo Ohms or Mega Ohms for the other readers, because that is how it supposed to be for no misunderstanding of the value.
"The bad sensor read as follows pin 1 to pin 4 you will have a reading around .022. pin 2 to pin 3 was 1.099. Pin 1 to pin 3 will read 1.135 and finally pin 3 to pin 4 will measure infinite / NC again these are approximate values."
To the younger audience and not too informed on multimeters use, the numbers you stated might confuse them. So please don't take offence on why I am discussing this.
Any succesful fix to a problem is a good database/case-study we all can use. This is important.

Your pressure sensor showed readings, so this combo sensor pressure side may not be very accurate , but it is not defective....pressure wise.
I would not use Ohm meter to test the pins I mentioned as no-use to test because the target is the thermistor ( Pin 1 to Pin 3 ) which your original sensor ( defective one ) showed 1.135 Mega Ohms.

Therefore it is a good time your case study/post, becomes a test for me/us to find out what value range is the B4/25 sensor thermistor section on mine,
when a known good new unit is 7K* ohms ( *at 79F/26C updated by OP ), and I can share with the forum the result, while my batteries were already disconnected that night because I was messing with my Rear SAM.

Heat on fuel rail, HP fuel pump and the sensor B4/25 is inevitable.
Any metal attached on our engine cylinder head or block ( but away from the exhaust or turbo ).... will be near coolant temperature, at least.
I think the damage to your sensor B4/25 is not from temperature overheat, it maybe from fuel leak into the thermistor section.
Don't throw your sensor, let it dry for a week or two and let air circulate into its tip, test again Pin 1 to Pin 3, if value is lower than 1.135 mega ohms by much, the fuel leak may be the cause. Thanks.
I too am curious..... and will make B4/25 as an item to occasionally measure its thermistor section if it is the weak part of this combo sensor.

You know, after the test I did, I waited a few hours to get my fuel rail to cool down ( from hair drier test ) back to ambient temperature.
I measure B4/25 thermistor section again at 30C and also the AFTER aftercooler intake air temperature sensor.
I was curious if it would register the same Ohms value, assuming these two share the same resistance table/curve.

B4/25 thermistor section 8.37 Kilo Ohms at 30C / 86F
B17/8 AFTER aftercooler intake air temperature sensor / Charge Air Temperature sensor , 7.43 Kilo Ohms at 30C / 86F
Thermistor developed drift overtime is not uncommon.

If indeed my B17/8 has drifted and its the same 10K ohm type like B4/25, my engine fuel trim maybe effected , because air intake temperature assumed by ECM hotter than actual.

 

Hi
I have very similar Trouble with my GLK 350 12' , receive p0087 almost everytime i start but..... once i look on live data from my obd device ( kts bosch, thinkdiag ) i see that start Fuel pressure stays more or less at 20-30 Bar (290-435 PSI) level, almost exactly after 15 sec does 50 Bar and after another 5 goes up to 150 Bar and stays there.

Have you seen something like this before ??

Engine runs like crap

I've checked almost (a lot) everything , did some changes for new parts (like brand new HPFP, pressure sensor, also new Crankshaft Position Sensor), checked all 6 fuel injectors and have still no idea what else i can do.....

 

wrong please delete

 

Yoka,
Something is probably defective at either your High Pressure fuel pump or the low pressure fuel pump at tank or its filter ( source ).
The delay u mentioned for high pressure fuel pump pressure to hit 150 bar is way too slow........
If you read this thread carefully, you would have spotted how a healthy High Pressure fuel pump should behave within 2 seconds after engine start.....
Below is 1 second per data point