Fuel supply problem - W203 220 cdi

[nujon]

Problem is that once the tank gets to just under 1/2 full, air is being drawn into the fuel line and the engine inevitably stops. Connecting a vac pump to the fuel filter outlet with a half-full tank produces only air wheras blowing down the line and listening at the filler cap - one hears a noise akin to blowing down the straw of a just finished milkshake! - clearly there is a leak path in the tank which is exposed when the level drops sufficiently.

Anyone any ideas

 

[alibally]

1- Does your car have an electric lift pump or a mechanical one?
2- When was the fuel filter last replaced?

The reason for asking is that if the fuel return pressure is low the fuel won't transfer from one side of the tank to the other.

 

[nujon]

mechanical lift pump - the problem started after changing the fuel filter - how do the two sides of the tank work together do you know?

 

[alexander patie]

the return flow from the front is routed through a nozzle in the tank. the nozzle is placed in such a way as to generate a vetruri effect which scavenges the fuel from the O/S of the tank into the N/S swirl pot. the pipe work in these tanks has been know to fracture causing problems like this, although i would suspect a problem with the filter as it has just been changed - silly question, but are the pipes on the right way round?

 

[nujon]

A search of this and sister sites reveals the problem I describe to be not uncommon - and often emerges after breaking the integrity of the fuel supply/return system in the engine compartment (i.e. to change fuel filter, replace leaking pipe etc.). Noting that the sender units have absolutely no moving parts and that a 'venturi spray pump' fed by the fuel return is intended to create the vacuum to pull fuel from the passenger to driver side, failure of this function must be one of:-

• a failed pipe/cracked sender casing
• an airlock
• insufficient fuel return rate to 'drive' venturi pump

- noting reports from those who have removed the sender units only to find the interconnecting pipes in apparent good order - and noting the problem-onset to commonly follow work on the fuel system, I suspect an airlock

Any views / solutions welcome

 

[nujon]

I have established that the 'spray-pump' (eductor) system that sould transfer fuel fom the n/s to o/s side of the fuel tank is not doing so. I am in the process of removing the sender units to inspect and would welcome some advice:-

(i) how does one unscrew the (very-tight) black plastic rings that hold the senders in place?

(ii) what to look for on the sensor units that might explain the fault? (I ask this cos a couple of years ago a member was exploring the self same problem but did report what his solution was)

 

[nujon]

After 4 hours trying all ways, finally managed to unscrew the 150mm dia. HDPE nut that holds the n/s fuel guage sender/transfer spray pump unit in place ! (too dark now so I will defer lifting out and removing and inspecting the unit until tomorrow) The nut was very tight - attempts at removal using a self-locking, 3-leg wrench and then a strap wrench would not work - the plastic simply deflorms and grips even more. Finally succeeded using a hard wood drift and a Manchester screwdriver to 'tap' (slight under-statement) the nut round via its gripper bands - 1 + 1/2 full turns required before hand turnable and then with considerable effort - will definately be re-assembling with a smear of PJ on the threads!

 

[nujon]

Member alibally kindly replied to my post regarding the above issue and raised the possibility that insufficient fuel was returning to the fuel tank spray (venturi) pump to transfer fuel from the n/s to the o/s - this now seems to be a strong possibility since:-

1. fuel tank interior and contained fuel nice and clean - no solids, gum, gunk etc.

2. removing the inspecting the spray pump module has reveals no problems viz:

• spray pump module sqeaky clean - no fouling / debris of any kind
• spray pump nozzle and venturi body in good visual condition - no erosion or distortion evident
• nothing broken or cracked - no internal leak paths

3. running the engine and monitoring the fuel returning to the tank reveals a constant albeit 'lazy' flow - judgementally no where near enough to induce venturi effect at the spray pump

What might cause an inadequate fuel return rate wherin there is no external leakage of fuel from the return line - anyone any ideas?

 

[nujon]

No nearer solving the problem - removed the driver's side in-tank module ('swirl pot') a non-dismantlable multi-chambered plastic thing to which the two hoses from the n/s spray pump connect - no obvious failure - so:-

- if fuel in the drivers side of tank, engine will start and run normally until driver's side empty - but passenger side of tank remains full! - inspection suggests nothing wrong with passenger side module (spray pump) - therefore what?

 

[nujon]

Latest theory is that the pressure regulator function (sensor and / or valve) is defective in that it is allowing too high a proportion of the fuel take-off from the rail to be recycled to the HP pump - to the extent that the fuel return flowrate is too low to drive the spray (venturi) pump in the n/s of the fuel tank - anyone any views?

 

[zuffenhausen]

May be a leaking injector allowing a drop in pressure. See my issues here:

http://www.mbclub.co.uk/forums/electronics/90433-help-car-just-died.html

 

[sadmerc]

just read your posting as we are having the exact same problem, and wondered if you had managed to resolve this problem yet?
It has been suggested to us that it maybe the air vent filter, we are therefore going to have a look at this possibility tomorrow, and will let you know the outcome. IF however you have managed to resolve this problem any advice would be greatly appreciated.

 

[nujon]

Thanks for the reply - still on with the problem and have just replaced the rail P/R valve in case it is not opening fully on warm engine - got a full tank at the moment so will need to see what happens when the tank approaches the half-full, critical point. I remain convinced the problem is inadequate fuel return rate but this of course has a number of possible causes:

- air entering fuel feed lines
- chronic wear of fuel pumps
- pump shaft seals leaking
- excessive fuel suction line resistance
- fouled fuel filter
- defective P/R valve
- defective rail pressure sensor
etc.

For various reasons, I suspect the the regulator

I'll keep you posted

 

[nujon]

Thanks to Zuffenhausen also - I did a home-made leak-off test which did not reveal any obvious problem - i would note in any case that the leak-off manifold joins the fuel return line at the rail outlet

 

[nujon]

HELP - no further forward. Since last post, have renewed all the plastic and rubber pipes in the engine bay, the fuel filter, rail pressure switch, regulator and lift pump - no different - fuel not transferring from near to off-side. All I am left with as a possible cause is excessive injector leak-off causing the regulator valve to throttle the fuel return in order to enable sufficient rail pressure for the engine to run.

Any views please?

 

[jaymanek]

You must have used a non OEM fuel filter?

Only Mb, Bosch or Hengst will do... The others will cause the exact problems you are experiencing.

They have a valve like system within which controls the return flow to the tank.

We have seen quite a few of these in our workshops...

If you take your car to Mb they will charge you for a set of pumps/senders and also a new fuel filter...

 

[nujon]

Thanks - I initially replaced the filter with a FRAM then replaced in turn with an OEM Bosch as another reader had alerted me to the problem you raise. Sadly, fitting Bosch filter has made no difference.

 

[jaymanek]

but you will need to fill the tank again for the process to be able to work smoothly... have you done this since changing to the Bosch?

 

[nujon]

Thanks - yes - now several tank loads later

One factor I've never 'confessed' to earlier was putting valvoline injector cleaner into the tank literally on the eve of all my problems starting - could this of itself 'mess-up' the fuel supply / injector systems.

Think I'll get a professional leak-off test done next - I did a 'home-made' one at tickover and noted some leak-off @ No. 3 - maybe gets worse at higher rpm (read somewhere that the transfer system does not work much below 2000 rpm)

 

[jaymanek]

the design of the system is really poor i think. Initially there was an electric pump on both sides of the tank.. when the supply side of the tank got low, it pumped fuel electrically from the other side...
they changed this and it relies on certain pressures to force the fuel into the other side... got to be a recipe for disaster.