So I drove my C55 to work today and it ran fine with no issues (60 miles each way), but when I got in it to get lunch it started right up as usual, then sputtered and died. It now cranks over but will not start.

I don't believe it is a CPS since it happened when the car was cold. My first guess was bad fuel and/or a clogged fuel filter (car has a little under a half tank), but then when I plugged in my MB 2.0 scanner, it won't communicate with the ECU or TCU but will scan everything else in the car. Tried my Bluetooth obd scanner and it won't connect at all.

Ideas?

 

Well, after checking everything else (MAF sensor, cam position sensor, etc) before committing to spending nearly $500 on a new fuel pump and sender, I had to admit defeat and bite the bullet. So, new fuel pump and sender are in (not as bad a job as I anticipated), car fired right up and runs great. That being said:

Now my fuel gauge doesn't work. Floated around a half tank for a while until I filled it with 14.3 gallons, then settled down to reading empty (no light on though).

I checked resistance on pins 1-3 and pins 1-4 and they read 155 and 350 ohms respectively (consistent with a nearly full tank and checked against old pump sender assemblies). Edit: The full range of both the original senders is about 50 Ohms (full) to 1k Ohm (empty). The actual pump is on the same connector and there are no problems there. I suspected maybe my car's connector had an issue with the sender common (pin 1), but it reads fairly high resistance (in the kilohm range) as expected. There is no mention of a fuse or SAM control over the fuel sender. Ideas?

 

I encourage you to invest 500 dollars in a scanner that reads all codes — at minimum Autel maxicom mk808. The 500 dollar investment will let you read ALL your manufacturer specific codes. It’s a GODSEND to be able to pull all the codes.

Since it happened after the swap it’s still likely to be the sender. Also, sometimes a manufacturer specific code error needs to be cleared to make things function again and this can be done with the correct OBD2 tool.

 

I scanned codes this morning and it said fuel level sender 2 (whichever one that is) was stuck. I'm not sure how the car derived this since both sides (judging by my ohm meter) read relatively close to full, but it cleared itself when I plugged the old senders back in and played with the float levels, watching the gauge go up and down. I then plugged the new ones back in and voila! It now reads 3/4 tank (even though it is full). I'll pull the senders back out tomorrow after burning down a half tank or so - it probably is a mechanical issue. Worst case, I put the original sensors on the new pump/sender assemblies.

I'm sure your Autel is a good scanner, but for $500+/- I'm holding out for a Star DAS/Xentry setup. I'll probably pay closer to double that, but it'll be worth it to be able to code retrofit features such as TPMS, rain sensor, sun shade, etc. This is a car I will be keeping long term.

I have a $125 iCarsoft MB 2.0 that works fairly well and will read and (usually) clear codes from 29 different modules in the car. The interface is clunky though and comms are flaky and the live data doesn't seem to work. I just get N/A for everything. When the fuel pump died, it wouldn't communicate with the car AT ALL until I pulled the battery cable and reconnected it, which of course cleared any codes it threw.

I also have a $170 Autel scanner (MD802 pro if memory serves) that will only access (and clear codes) from 4 modules. What it does, it does well, but it's woefully inadequate. It reads live data well from the ECU, but some of the readings are wonky, such as the MAF that reads WAY lower than it should or other scanners indicate (my VCDS setup for Audi reads the Mercedes MAF in the correct range).

 

Be sure to replace the fuel pump relay too if it wasn't done along with the pump. Have had similar issues with mine over the last few years and think the relay was involved

 

Just an update: fuel gauge issue solved. Totally my bad - the driver's side sender unit's float looked like it might get hung up on the hoses that cross between the two tanks, so I clocked it about 30 degrees away from the hoses, unwittingly pinning the float against the front of the tank. I loosened the retaining ring and straightened it out. Problem solved. Crappy thing is that I had everything out of the tank at least 3 times and kept repeating the same mistake. The gauge reading totally empty was an error condition that was cleared with my scan tool that actually detected the stuck float. I guess the moral of the story is don't second guess Mercedes engineers...

Just for general info, (in case anyone is interested) it looks like the sender unit on the driver's side is a simple siphon pump. It will nearly completely empty that side while keeping the passenger side with the actual pump full. The two floats are simple variable resistors that range from 55 ohms in the "full" position to about 950 ohms in the "empty" position. They are wired independently but with a common ground. The rear SAM than sums them together and runs them through an analog to digital converter with a combined full range of 100 ohms empty to 1900 ohms full. If you measure them at the connector, pin 1 is common and pins 3 and 4 are the two sides. Pins 2 and 5 are the fuel pump power. The drivers side tank will always read lower because of the way the siphon pump is designed, but since the computer adds them together, any slosh from one side to the other is negated at the gauge.