AWR-fix: e53 Fuel pump / siphon pump details

[andrewwynn]

@ scourt: when below about 30L total the right side should read about 1.2L +/- 0.3L.

Indications are the right sending unit is also defective. You can pull it out and measure the full sweep range of Ω. I forget what the range is but Google is your friend. Too bad I just threw out a spare sending unit. I had a pump fail, replaced under warranty and of course the other parts were fine.

The problem can be mechanical, like the float getting stuck, it's not uncommon that the bottom of the pump is not locked into its groove and at an angle. The pump is spring loaded to ensure a tight fit to the bottom but that can backfire because the bottom can get caught on top of the groove and not seat down all the way and that will throw off the fuel measurement.

Pull out the pump (starboard side) and it's likely the same problem you found on the port side sender.


Sent from my iPhone using Tapatalk Pro

[andrewwynn]

@ black: genius to make an o-ring groove. I did exactly the same for my food processor that would send powder out the top when running fine items.

I used a round Mill end bit in my Dremel to make the groove. I used nylon zip ties to hold my siphon pump together once I got the new o-ring in. (not around like your hose clamp) vertically around whatever I could find to hold it tight.

If you take out the pump side you may be able to pull the siphon outlet out of the holding clamp and confirm if fuel is coming out in volume.

I'm just realizing that as you pointed out there are two seams one with one without a seal. I forgot what I did to seal that extra seam. I think I may have used an axial seal (think washer vs o-ring).

There are no moving parts in the siphon pump, it's nearly impossible to have a failure in the foot but the head unit is stupidly assembled with no seal just nylon on nylon so it will fail. You could likely fix with just the hose clamp without even taking apart. Surely with a few wraps of Teflon tape to make it snug first.

The "simple test" for proper pump operation is to drive down to single digits distance to empty (DTE). There is zero chance of getting to single digits when the pump is worn out.

How are you confining the siphon is not working? Fuel starve when fuel in port side of tank?


Sent from my iPhone using Tapatalk Pro

[semcoinc]

FWIW, my IBUS app in my Eonon GA6166 shows my full fuel at 82 liters. It will hang there for a little while before dropping to 81 liters.

Mike

[andrewwynn]

Tank has approximately 93L capacity. You are seeing about a 10L error if you see 82L full.

The floats at pinned to the top of the tank for the top 10L so there is no difference in the float reading between 82L and 92/93L.

This is also why the fuel gauge doesn't move for the first 100km or so: the needle doesn't move because it's accurately representing the fuel float level.

[scourtaud23]

I will look into the fuel pump sender unit (maybe tomorrow if I find the time).

I also wondered if backpressure could be built up inside the tank when filling up if the Fuel tank vent valve had failed (I don't know but when I bought the car it was unplugged from the throttle body so I wonder).

[andrewwynn]

During fueling the filler neck is open to atmosphere. I couldn't find the official values for the sender Ω range. ;-(


Sent from my iPhone using Tapatalk Pro

[BlackMW]

@andrewwynn: I checked the OBC test #6, the left tank shows around 30 litres, the right tank shows 8 litres. Running like this for days.

I was driving with the pump running dry without noticing since I run on LPG,
than I got 20-30 second crank times. I replaced the pump but didn't realize the cause.
Half a year later LPG broke down and I got a reminder of the issue again when it didn't start one morning with tank showing 1/3.
Now second pump got longer crank times (3-8 seconds) so I know it is damaged already plus the return hose is quite soft.
Tomorrow I will get and install a brand new pump and recheck the OBC for fuel levels when running. Hopefully that will be the end of story

I had the issue for 1.5 years every two months or so, thought the tank is venting or water is getting in during winter, been putting $hit to remove water from gasoline without success

[scourtaud23]

Originally Posted by andrewwynn

During fueling the filler neck is open to atmosphere. I couldn't find the official values for the sender Ω range. ;-(


Sent from my iPhone using Tapatalk Pro

I'm thinking of filling up the car before I do the test so I can compare values in tank (via the plug) and on the bench (true full) I'll at least see if the tank is full or not and if the values are correct.

[BlackMW]

@scourtaud23
You should be able to read sender resistance in ohms in OBC hidden menu #6.
Using ohmmeter, my 2000/6 M62B44 senders read 030 ohm at empty and 500 ohms at full.
However this might be different for newer models as different forums say 980 ohm is empty and goes down when filling up.
But You can compare to your other side sender that You've replaced, left and right are the same.

[andrewwynn]

If you are forced to use polluted gas like we do (ethanol added) there will be no water in your tank ever. The ethanol absorbs the water. Even without it's not likely a cause of your grief.

The right side sensor should settle on about 1.3L (there will actually be about 5-6L the last 4 or so are purposely below zero so you will never run down to a dry pump even going to zero DTE.)

The long crank sounds like a different problem. I would definitely measure the fuel pressure to make sure you are getting full pressure. You could have a failure in the pressure regulator or the one way valve at the pump.

Let the pump finish priming before attempting to start.

If you have fuel pressure then you could have bad spark or dirty injectors or a could other things.


Sent from my iPhone using Tapatalk Pro