Fuel pump diagnosis and full test routine.

[nathan8]

Hello All! Its been years since I've been here but now I have some information to share, and need some advice too.

I had what can be described as a surging idle. Sometimes car is fine, but with increasing regularity until it occurred on every journey and then immediately after turning on. At speed there was no noticeable problem. The when idling, the revs would drop till the tacho hits zero, then rev to correct itself causing a surge in power. If in gear this forces the car to lurch which is embarrassing.

LHM came on onece or twice. Sometimes started in low power assuming 6 cylinders. Sometimes the red check engine light came on.

I took the car to a garage on dec 12th, and got it back two weeks ago no better, with a bill for 1x new MAF and 3 hours labour. Joke. This is the test procedures I have taken.

1/ Pull individual fuel pumps fuses
RESULT: pulling the fuse for pump 1 makes no difference, pulling pump 2 fuse stops the car within 3 seconds (on the next idle sag before the high rev)

Because the garages misinformation (telling me DK motors are at fault) I went through the following

2/ Rebuilt and cleaned DK motors. No change. Open and close cleanly.
3/ Swapped Coils over. No change.
4/ Swapped MAFS over. No change.
5/ Garage had seen the CPS sensors were sending a signal, though one went off - presumed it was the aftermarket chips, removed these and then the CPS were both reporting RPM.
6/ Spraying easystart to check for vacum leaks
7/ New fuel filters
8/ Battery check - they are fine I get about 50 starts from a charge without running (during my testing).

Car is in good condition with having the folling parts replaced under my tenure
Rotors & Caps
Sparkplugs & complete spark lead set inc sensors
Intake manifold gaskets
Was chipped, but now BMW standard chips back in
Fuel lines
Injectors professionally cleaned and flow matched
Vacum return hoses + valve (crank ventilation?) Other vac lines to DK motors as well
Banjo bolts
Lambda / O2 Sensors

Ok at this point I'm fairly sure its not the sparking or the fuel flow lines. Something to do with the fuel pumps because it runs fine with one fuse out - when put back and pulling the other one it dies straight away. Runs fine at speed.

Assuming no CPS problems, I cracked on - this is the procedure to test your fuel pump system

1/ Check fuses 23/24 - Replace - try shorting with copper cable in case
2/ Relay checks (blue in EML box) - tested resistances and heard the clicking. Swap over & see if problem swaps sides (mine did not). Run on approx 1.5V in closing the 12v power to pump.
3/ Short relay 30&83 sections (see diagram on side of relay) this provides power to the pumps for sure. Listen individually. Note that one relay also activates the DME cooling fan (scared me).
4/ Check that you get 12v at the fuses from the relays.
(NOTE: Relays only active when running or cranking, sometimes for 1 second on making position 2, but not always)
5/ Remove rear right hand (UK driver side) seat and access plate. Remove and re-insert round 3 pin cable. As you look at it top pin ground (grounded by DME) Left hand pin pump1, right hand pump2.
6/ Check continuity from fuses and relays to these pins.
7/ Put small wire into each pin of the connector and re-attach. Tricky. Use these to check you are getting 12v as expected from the front (in case a wire from front to back is bad).
8/ I have checked resistances across these pins. Ground to either pin is ~1.5 +/-0.5 Ohms, pump1 to pump2 is (suprisingly) about 1 Ohm.

At this point I still had no joy. Everything is as expected. Continue!

9/ Remove pumps
10/ Check the bolts to the pumps
11/ Jury rig fuel pump to a 12v battery. Beware sparks + Fuel!
12/ Compare pump operations - few seconds so they dont burn out. Beware about 200ml of fuel will eject!

At this point I find an anomoly - Both pumps gush from the top mounted ejection port, at visually the same rate. Pressure unk at this point. However! Pump 1 (the one that was detected bad by pulling the fuses) also leaks fuel at the top of the pump, near the electrical contacts. Seems to be coming from the white bearing at the centre of the pump.

I'm assuming this is the problem, but before I order the £400 assy (both pumps, housing and level meter) from bmw (they only have the full thing) - I'd like to ask if anyone has experience of this kind of fault with a pump.

If no one knows I will swap the pumps over, re-install the pumps and then see if the fuse that upon movnig causes the car to die swaps as well. For my own refernce its currently fuse 24 that I remove to cause this death.

Best regards to all. I'll try to add a picture of the top of the pump in a moment.

Nathan

[nathan8]

Picture of the fuel pump. Pump one is at the back, two in foreground.

Red circle shows the area petrol is leaking (squirting) from on pump1. Pump 2 does not do this.

[nathan8]

when fitted will report results

TA

[Huddo]

By now having replaced fuel pump your problem is solved, right? When the pump gets older, it seems to work fine, but cannot maintain consistent fuel pressure when idling especially as it warms up. The DME detects insufficient pressure = no gas, and signals to cut-off engine. The fuel pump burps back into life and the engine sometimes just catches again, but sometimes dies.

To check if this is problem, hold accelerator at about 1,000 rpm (of course with foot on brake) and see if this surging still happens. It's the sign of non-servicable Bosch pump seals getting old

Another indication of same problem is - perfect cold and lousy hot start after few minutes of shutdown. The fuel refluxes back out of fuel pump leaving gas bubble. You then need to crank engine for a good few seconds to get engine going again and until you load the engine by moving off, it will continue to be very cantankerous.

[nathan8]

They sent me something with a top plate etc attached. I'm going to have to return it and try elsewhere, or get on sent from America I think they have just the pumps.

Really need to get it sorted as I've just bought at 360 and between that , 8 series and my monster truck its getting a bit ridiculous.

Cheers for your help and encouragement

[nathan8]

And no change!

I'm going to go and check now - I think perhaps I've taken the wrong pump out so I'm going to go and check by swapping over the wiring for pump1 and pump2 and see if the error swaps sides as well (checking this by pulling fuse 23 and 24 again

Its never an easy fix - especially when your working away from your tool box.

Report later I guess

N

[nathan8]

Gets stranger.

If I swap the cables at the pump so that pump1's electrical feed goes to pump 2 and vice versa the fault also swaps.

I think this means that its got to be a pump not an electrical problem

(brain a bit addled by petrol fumes)

I took the pumps out again and the pump that does not seem to be working is the new one! If the other (older pump) is disconnected it dies immediately. If the new one is disconnected it makes zero difference to the engine.

This is driving me batty. Next tests are

1/ Check basic pump operation on bench with 12v source.
2/ Flow test - put some longer tubeing onto the pump and measure how much comes out in 10 seconds etc.
3/ Pressure test - get a pressure test kit and check the pressure at the rails.


ponderings
------------------
I was wondering if it could be the fuel lines from the pumps to the fuel rails, or the rails, or the regulation valves being shot on one side? Doesnt the fault following the electrical swap over preclude this?

I suppose not.

I might see if I can run a line direct from the pumps to the fuel rails cutting out the fuel lines.

The fuel filters have been replaced. Perhaps a pipe has bent or corroded closed?

[rocket.science]

i bought two after market fuel pumps with 255lph fuel rate.
they work fine... but what i had to do is upgrade the internal wiring of the fuel pump carrier. to a larger gauge wire. the fuel pumps have a larger current draw and the thin ground wire could not handle the flow and heated up quite a bit.
each pump cost me $60,- ... it's well worth a bit of engineering

[nathan8]

Ok so the new fuel pump had reverse polarity.

Problem still not cleared.

I think I might have two problems actually. Swapping the electrical supply for the pumps (ie cable 1 to pump2) and testing by pulling fuse 23 and 24 causes indicates that the pump is still at fault. The engine dies when 24 is pulled on a 1-1 and 2-2 connection, when the connections are crossed 1-2 and 2-1 then it dies when 23 is pulled.

However pulling the kingpin that is accessable (left hand side of the engine as you are sitting driving) always causes the engine to die, quickly. If pump wiring is swapped over then the engine still dies when the left hand kingpin is pulled. Pullin the other one (harder to get - near the computer port plug) makes no difference ever.

I've swapped over the coils (and I had a spare) no effect
Swapped over the DME conections - no effect.

There is a new maf on one bank, but this is the bank thats not running. Is it possible the new MAF is... naff? I guess I need to swap them out.

The bank that is running poorly is not getting any where near as hot as the other bank.


Plan of action

1/ Get pressure test of the fuel at the rails to remove that possibility.

2/ Swap mafs

3/ shout at it.


Could really use a hand. What could cause one side to hardly work. Remember its not getting as hot as the other.

[nathan8]

Hi

Where did you get them from? I just had to spend UK£ 200 on a BOSCH unit

TA

[nathan8]

[nathan8]

The garage that I used to trust (not any more) said that they thought it was the DK motors and I should replace them in pairs.

1/ Can DK motors cause an engine to run like this (IE one bank hardly running, and surging and then almost dying at idle)
2/ Is it true that they need to be replaced in pairs.

[rocket.science]

hello nathan
the pumps i got were from TRE perfomance.

TRE performance bmw fuel pump

[mlweaver]

Nathan,
I've read all your posts, but there are too many variables to try to answer each question.
Let me offer my thoughts about your problem.
If the car runs well at speed then the fuel pumps are probably fine. A small leak inside the gas tank doesn't mean the pump is not delivering adequate fuel pressure. Poor fuel pressure would be more likely to cause a problem at higher RPMs.
A rough idle is most likely to be caused by MAF, DK motor, or electronic engine management.
All three of those can be swapped side to side to try to isolate the issue.
I know you said you have tried that, but please consider doing it again, slowly and carefully observing the exhaust temps before and after the swap. It is easy to be fooled (ask me how I know this!).
If the check engine light comes on or the problem triggers a LHM, then it should generate a fault code. It would be worth the effort to find a good code reader (or BMW dealer) to read those fault codes.

Best wishes.

Mark W in mid-MO
2008 335xi coupe
1993 850Ci 6-speed

2 turbos ..................or 12 cylinders, no waiting!

[TxGR8White]

I too doubt you have a fuel-pump problem. Have you studied the way these components work with one another? You need to have crank position sensors AND cylinder impulse sensors that are in working order. You also need DK's that do not stick on the closing cycle, and MAFs that are also in working order.
You need to systematically eliminate each component, in order to find the problem.
If you have access to a BMW GT1 or MoDic, then that will at least give you a very good starting point.
Have you done a "stomp test"?, If so, what were the codes that came back - these will point you towards the electrical, or the emissions portion of the engine.
Are the O2 sensors in working order?, are they connected correctly to their respective plugs?
Are the fuel filters clear (I had one clogged filter that put the car into Limp Home Mode the second I would start it)?

You can start with the pumps. Pull the electrical plug on top of the pump carrier - get a 12Volt source and put power to the individual pins (the 2 pins next to one another are the individual positive terminals, the single pin is the common ground). Do you hear each pump running when powered? If so, the pumps are fine. Next check the wiring to the carrier - is the wire hard and brittle?
Check the fuel pump relays and the fuel pump fuses - if you still have the stock BMW fuses in your car, do yourself a favor and get rid of them and replace with the new style fuses.

It seems to me that your car is sensing a "problem" with one bank, and this is the cause of the pump not firing. You need to start swapping components from one side to the other, one at a time, and see if the symptom jumps sides. As soon as it does, you should have the culprit in your hand.
The key is to only change ONE thing at a time and see if it made any difference.

Good luck, and keep us posted.

[850idontno]

When swapping electrical feed from pump 1 to pump 2 the fault also changes.There is your 1st clue that it is something other than the pumps.As TxGr8White suggest,check related electrical componants,also check for corroded connections. Good luck

[nathan8]

Hi There Mark I thought I'd reply here.

I've got a BMW electrics mobile mechanic coming on monday with his test machines so we can do a joint diagnosis.

I have been changing things one at a time to see if the trouble bank swapped with the item. Only doing one at a time, then re-testing.

Dk motors have been cleaned and they operate like each other (if I feel their movements while someone pushes the pedal).

MAF's I've swapped over but will do so again to check

DME's input cables have been swapped over

I still want to swap the fuel lines over at the back of the rails and to eliminate fuel as a problem. I've noticed that at the front the small pipe on the left hand side rail (as you stand in front of the car and look at the engine) is kinked, perhaps the fuel cant get through to cycle round? Worth a check.

Could lambda sensors cause a whole bank to fail? Physically taking these out and swapping over is worth a go, but I'll wait till said man pops round so we can see what his machine says.

Exhaust - mine has a cross over pipe in the middle which makes things trickier but now I know about it I'll be checking the temps at the manifolds when testing.


(The garage that I now dont trust have some diags machines, he did mention something about a CPS not showing RPM at some point, but this was only once and he thought a glitch. Could be the problem? We'll find out on monday)


Thanks for everyones input it helps to talk though my experiments. Working on my own with limited tools, after a days work and chasing a fault around under the influence of petrol fumes, well we've all been there but it really makes a difference to have some e-help

Nath

[nathan8]

Howdy! Just to follow up these questions

If you have access to a BMW GT1 or MoDic, then that will at least give you a very good starting point. // Man with these is booked to visit monday


Have you done a "stomp test"?, If so, what were the codes that came back - these will point you towards the electrical, or the emissions portion of the engine. // Going to do this now and get the results posted here


Are the O2 sensors in working order?, are they connected correctly to their respective plugs? // They are correctly wired side wise (will reverse to check) I've had these the wrong way round before and the symptomps were hesitation then power, idle was still without this almost die then surge - but will check (probably monday). I wouldnt think that a bad lamda sensor could stop an entire bank from doing anything at all!



Are the fuel filters clear (I had one clogged filter that put the car into Limp Home Mode the second I would start it)?
// New fuel filters

You can start with the pumps. Pull the electrical plug on top of the pump carrier - get a 12Volt source and put power to the individual pins (the 2 pins next to one another are the individual positive terminals, the single pin is the common ground). Do you hear each pump running when powered?
// Yes

If so, the pumps are fine. Next check the wiring to the carrier - is the wire hard and brittle?
// No - the wiring seems ok I've checked from the fuse box to the pump end and run the pumps with a hard wired manual 12v source to eliminate the harness.

Check the fuel pump relays and the fuel pump fuses - if you still have the stock BMW fuses in your car, do yourself a favor and get rid of them and replace with the new style fuses.
These have been swapped from side to side, and in the case of the relays, shorted out to eliminate. The fuses have been swapped over and shorted out. Will change fuses now as recommended.

Thanks for your help!

Nathan

[mlweaver]

I see that you have replaced the distributor caps and spark plug wires (very good idea), but your post mentions swapping the coils ...
If the coils are old I suggest that you also just replace them. They do fail with age and can cause disasterous problems with the DME modules. I don't know that they are the cause of your problem, but it is a fairly inexpensive way to avoid even more confusion as you continue the search.
Thanks for posting your findings here - it will probably help many other owners in the future.
Best wishes,
Mark W in mid-MO
2008 335xi coupe
1993 850Ci 6-speed

2 turbos ..................or 12 cylinders, no waiting!

[TxGR8White]

approach it very systematically.
The next thing I would look at are the cylinder impulse senders (the little donuts on the #6 and #12 sparkplug wire). If you have access to a know working #6 or #12 sparkplug wire with donut attached, you can use this to test on the "dead" side.
Just reading the resistance across terminals doesn't always expose a bad plug/wire. I chased a bad CPS sensor for a couple of weeks- the resitance was within specs across the plugs, but once replaced with new, everything worked.
Do keep us posted on your progress, we are eliminating possibilties everytime...