Hi,

I have an ML320 CDI (non bluetec) with an OM642 engine that does not want to start. It cranks, and after a few seconds times out and cranking stops.

History so far:
The car "ran out of fuel" due a really bad repair some knuckle head "performed", and the fuel line from the tank output into the rigid tube got disconnected. The driver at that time tried for quite a while to start the car, until giving up and calling to a tow.
After temporary repairing the fuel leak (rigid line is on the way) the engine started only three more times, enough to move the car to a storage location (my garage). It took quite a while to start it the first time, but after jumping the battery with another car, it started "easily". Next start was done after leaving the engine running for 30min to heat up (assuming bad glows), and it was really hard to start, but once it started we drove it into the garage.
One week later I tried to start it, and only by jumping it and lots of tries it ran, and only once, and since then I have not been able to start it, be it jumped, after leaving battery with a charger, etc.

I have changed both air filters, they were full of debris, and fuel filter. The tank does have fuel, scanner reported 47L, but just assumed it might be a bad reading (a few months ago the car had an issue were it was not reading fuel tank level), and added ~10L more, and scanner now reports 55L, so seems to be that there is fuel.

When trying to start the engine, DAS reports battery voltage goes under 11.x (DAS turns red); this does not happen if jumped.
I also checked waveform on camshaft sensor and it is working as expected.

One thing I noticed, is that when I try to start the engine, the DAS reports rail pressure at barely 120bar, most of the time 110~119bar, while saying minimum required is 120, interesting; and also quantity and pressure valves show a value between 30 and 50% each while cranking. Shouldn't pressure valve value either be at 100% or 0%, whatever it is for fully closed?, and quantity sort of "open" not sure this one.

I have not measured the leak from pressure valve, but the number I see on DAS while cranking makes me wonder if there is something else, like CDI computer opening it on purpose.

Also, no errors are reported at all. The vehicle has a couple errors, parktronic , rear SAM window heater, and top window issue, but nothing else.
If I try to start it with pressure valve connector disconnected, the CDI computer notices it, the same case with quantity value; the error is reported.

I am really out of ideas, if anyone can give me a help I will really appreciate it,

Thanks,
Ivan

 

At a high level it sounds to me like you are not getting enough fuel to the engine. It's just a guess but based on everything you are saying it seems to me that there is not enough fuel pressure

I can't remember if there is a port to add a fuel pressure gauge on the diesels? If not DAS can tell you what the fuel pressure is (and should be)

There are two pumps I believe.. one in the tank under the back seat, and one under the hood. One of them is a booster pump. If either fails you can have fuel pressure problems and car won't run.

I'd run all of the DAS fuel pressure tests, and follow any guided tests that it suggests after that. It seems weird to me that the CDI computer would do something funny right out of the blue. This sounds related to starving the engine of fuel..that is usually when fuel pumps burn out.

 

The fuel pump under the right rear seat is the Low Pressure unit. It is the one most likely to start failing.

 

The car ran out of fuel once, empty tank. It was due to the fuel tank level sensors not working supposedly, that is what the owner said; and at that point it was taken to the repair shop that damaged the car for fixing.

I will ask for help to measure the low pressure pump side, as I do not own a fuel gauge, but know someone that has one
Will update as soon as I have info about it.

 

Update so far on electric pump:
1. generates more than 7bar pressure after a couple seconds it is running.
2. flow is about 56mL/second. I measured 1.22L in 21.7 seconds

I measured these two number on the fuel filter output. Also it took like 18h for the pressure to go nearly zero after I powered off the pump, I wonder if this is normal.

In this page https://diag.net/msg/m42gozknibrjhlt701d3th52f5 the 642 fuel pump should at least have a flow of 500mL/9s, which is basically what I got. I cannot find in WIS what the right numbers should be. Also, the pressure gauge I used goes up to 100psi (~7bar), so I am not sure how much over 7bar it really produces.

Still engine does not start.

 

Another update on the saga.

I new high pressure fuel pump with quantity valve and temperature sensor, and pressure sensor later, the car "starts". What we did is alter the voltage the pressure sensor sends to CDI computer, and make it higher, hence increasing the "pressure" CDI module sees and now when the car is in "on" position, CDI reports rail pressure as 160bar. Then we crank and it starts, and as soon as it starts "reverse" the pressure sensor output voltage trick. It takes more than normal, but starts pretty reliably. And yeah, if we turn it off, the trick needs to be applied again...

This way it always starts; this is not a solution, but at least engine starts now.

I will really appreciate it if anybody could hook an scanner and go to the "CDI4 module->current values -> engine does not start -> engine rotates" and provide the values as the engine cranks:
* rail pressure: at what value engine starts? 120bar? 140bar? 150bar?
* quantity valve %, in "on" and while cranking
* pressure valve %, in "on" and while cranking


For the voltage trick we used a potentiometer between +5, sensor output and wiper to computer, to raise the output. If anyone is interested I can upload a schematic of it.
For the "reverse", I short the potentiometer between sensor output and computer input.

 

Ivan,

I am having the same issue, interested if you found a permanent resolution/cause and if you could post your temporary schematic. It would be much appreciated. Thank-you

 

I forgot to update the thread...
In the end, we measured the return of the six injectors, and only two of the six had some acceptable numbers. We pulled them out, and sent them to a lab, and three were leaking into chamber like crazy, one was kind of stuck, and I do not recall right now the others. Bought 6 new ones (yeah not cheap) put them in and car started.

Currently it runs "OK", but if you accelerate "too hard", and pass the 3000rpm in engine, check engine comes on, and looses power. The error, I do not have in mind right now, but is CDI computer reports too much pressure in rail. Also, it passes government inspection, but in my opinion it still has too much diesel smell in exhaust.

Regarding the error, all I have found is that the return might be obstructed. I still need to check that.

For anyone that reads this, does anyone know a "easy" way to measure rail pressure, with some external unit, not the rail sensor itself. I have my doubts that the sensor might be off, and reports higher than it really is.

 

Remove the eng. cover and locate the test port on the fuel rail. With a cold eng. attach fuel pressure tester, start eng. and pressure should be more or less around 50 psi and should graduate with throttle advance.

 

Hi IVAN
Have the same issue with no start but Didn't understand how you did ti trick the voltage . would you please give more details ?
4 months Now and still have the same issue
Thanks a lot