Fuel pump issue????

[Mattypompy]

Hi guys,

 

I have an enduring issue with the glowplug warning light coming on and the engine going into limp mode or cutting out all together as did the other day. After turning off then turning on the ignition it is resolved. The issue occurs generally at higher road speeds and gears under acceleration. The fault code is;

Fault code:

P0251 - Injection pump fuel metering control A malfunction (cam/rotor/injection).

Also I have a squeal/screech from the drive belt area (suspected fuel pump) at idle and under load intermittently.

 

I have read up as best I can and have come up with the following thoughts on diagnosis.

1.  Fuel pump bearings going causing erratic fuel pressure in rail.
2. Injector blockage.
3. Kink in fuel lines.
4. Fuel metering valve defective.
5. Fuel filter (which is new not OEM) incompatible.
6. Camshaft sensor defective.
7. Fuel pressure sensor defective.
8. EGR dirty

I have used some diesel treatment additives Forte and Miller's, no joy.

Anyone have any thoughts on a strategy? Cheapest options first! I'm gonna start by changing can sensor first.

Focus 1.8tdci 2001

 

Cheers.

 

[Tdci-Peter]

1 hour ago, Mattypompy said:

P0251 - Injection pump fuel metering control A malfunction (cam/rotor/injection).

Also I have a squeal/screech from the drive belt area (suspected fuel pump)

On the Mk2 1.8TDCI (up to 2007), the fuel pump is chain driven from the crankshaft. As the engine is common I would expect this to be the same for Mk1. So belt slip will not affect the fuel pump.

P0251 suggests the fuel rail pressure is wrong, probably too low. In higher gears, at max throttle, the turbo has time to build up max boost pressure, calling for max fuel. But rpm may be still fairly low.

It could be:

Clogged fuel filter, or air leaks into the housing: Always the first item to check.

Blocked or kinked pipes, maybe.

Excess injector leakage, testable by leak off test.

Fuel pressure regulator solenoid valve.

Fuel pressure sensor.

Fuel pump: Probably the 1st stage, the transfer pump & the metering system. This stage delivers a metered quantity of fuel (at up to 6 bar) to the 2nd stage of the pump, which then raises the pressure to 200Bar to 1600Bar for the injectors. This stage is rather rpm dependent.

Forscan can monitor fuel rail pressure in real time. See: http://www.fordownersclub.com/forums/applications/core/interface/file/attachment.php?id=42953

This may, or may not (!) help.

 

[Mattypompy]

Hi Peter,

 

Thanks so much for the very knowledgeable reply! You're correct the pump is chain driven, the drive belt pulley just powers the alternator I guess.

I've had a few more thoughts and am going to analyse some data from the ecu and see how to proceed. I have Forscan (think we may have discussed before!) So will look into fuel pressures through the rpm  range and compare to tolerances.Also look at injector metrics.

I have ordered a camshaft sensor £7 so will try that first.

Then I'm thinking new Bosch fuel filter £16. Then perhaps fuel pressure sensor, cheap too.

If that doesn't resolve it, then I'm thinking paying a specialist for a leak off test or replacing the metering valve £52. What do you think? Also where's the fuel pressure solenoid?!

 

Thank you once again.😀😀😀😀

I'm not going to spend out for a new pump and fitting, too expensive for this car.

[Mattypompy]

OK so I have some data from the obd now.

There seems to be a significant discrepancy between actual fuel pressure and required pressure,  around 20℅ @ idle.

Fuel pressure is 260 bar @ idle and around 700 @ 3k.

Fuel demand on injectors range between 6-7.2mg @idle and 15-22mg@ 3k.

Fuel offset ranges between 1 & -0.9.

Cylinder combustion noise ratio 1.4-3.7:1

Not sure if this is suggestive of anything.

Also my Haynes reckons the fuel pressure sensor is integral with the pump and can't be replaced but eBay sells them separately.

 

Any suggestions for a sequence to approach it?!

 

Cheers!!!😀😀😀😀

[Tdci-Peter]

2 hours ago, Mattypompy said:

Haynes reckons the fuel pressure sensor is integral with the pump and can't be replaced

I think the pressure sensor is in the common rail manifold unit (pic below). Yes, I have noticed it being sold now. It may not be easy to remove, very tight or rusted at the thread entry. This is a high pressure part (22,000PSI), the sealing will be metal to metal, usually on a small diameter ring, not at the thread. Also preventing dirt ingress is vital, Ford procedures use a vacuum suction device around HP parts when dismantling or assembling.

The pressure regulator solenoid (or IMV) is mounted in the back of the fuel pump. I am fairly sure this is a low pressure device, working between the 1st & 2nd stages of the fuel pump. So sealing is less of a problem (can use rubber o-rings & seals), but cleanliness is still vital.

You have gone deeper into the data than me, I have only looked at the fuel rail pressure. 200 to 700 Bar sounds reasonable for an unloaded engine. It needs near to full power to get to the maximum pressure of 1500 Bar or so. It may need road testing to see that. You should be able to monitor DTC count, which would show if & when the error happened.

The Mk1 Focus seems to have Delphi system, the Mk2 has Siemens / VDO, so there will be differences. Without some really good comparisons, I can not think of a good way to differentiate between IMV & Pressure Sensor errors, or other pump errors.

But a failure to achieve set-point (discrepancy between measured fuel pressure and required pressure) would suggest IMV or pump. (20% error in sensor would just give the wrong real pressure, but measured pressure should still match required. OBD & the pcm both see the same, wrong pressure.) Especially if the measured pressure was low, suggesting the pump / IMV could not deliver

[Mattypompy]

I take your comments about the pressure sensor being a HP device and the criticality of the mating faces and torque.

Perhaps that's why Haynes warns against replacement for safety reasons for a DIY mechanic. Picture below of a fuel pump for mine. You can see where the IMV inserts into it and I think the pressure sensor is there too above it. But this is a Mk1 so the FIP is probably different to yours.

I'm gonna replace,

Cam sensor,

Fuel filter

and then IMV If that fails to resolve.

After that I'll think about PS or then cut my losses and sell maybe.

Thanks for the info Peter.😀😀😀😀

 

 

 

[Tdci-Peter]

1 hour ago, Mattypompy said:

Perhaps that's why Haynes warns against replacement for safety reasons for a DIY mechanic. Picture below of a fuel pump for mine.

Yes, though it is not from Haynes, it was, and probably still is, Ford policy not to sell the sensor separately. It could be they are being overcautious about leaks. Ford say the HP pipes are single use only, but I am sure lots of people have re-used them without problem. (Provided they have not been bent, damaged or abused in any way).

I am still a little confused about the valves, both on your Delphi system & mine.

Lots of images show the mk1 pressure sensor fitted to the "spider" or common rail accumulator. I have an image that shows this in addition to what looks like two valves on the pump. My wiring drawing shows connections for 2 valves, a pressure regulator and a flow regulator. But for the Siemens system I can not track down replacement flow regulators, only the pressure regulator. For the Delphi, I can only track down the flow regulator (IMV) valve.

I am fairly certain that the flow regulator works at the intermediate pressure to control fuel entering the HP stage, and that the pressure regulator valve works on the HP outlet as a back-up, quicker acting regulator or dump valve. But good information on this is scarce.

So if there is a "lump with wires" screwed into the spider, then that will be the pressure sensor. I think the second connector on the pump may be a HP pressure regulator valve.

 

[Mattypompy]

Strange, the confusion over the location of the PS. I have just consulted Haynes in the fuelling bisection looking at the common spider rail and the removal process refers to removing the connector for the PS! The image shows it partially too. But, it also definitively states it's located on the rear of the pump! Either there are two PS's or it's a printing error and the one on the pump is a valve not a sensor!

 

All very unclear, but I would really like to replace it if possible to cover all my bases and have peace of mind. They only cost £23 for a Delphi but I need to know where to start! Think I'm going to look for a schematic of the pump.

Cheers😁😁😁😁

[iantt]

the pressure sensor is the one on the spider arrangememt, the pump has inlet metering valve and fuel temp sensor.

with your symptoms i would first be changing your fuel filter, and drive it hard under full throttle to see if fuel pressure drops or not.

[iantt]

also stationary with engine on, observe fuel pressure while wiggling the pressure sensor wiring and connector checking for spikes in fuel pressure.

[Mattypompy]

Cheers Ian. I was going to do the filter as first port of call. Obviously Haynes has erroneous info in it then. Thanks for pointer!😀😀😀

[Mattypompy]

Was going to do an electrical test on the sensor.

[Mattypompy]

Hiya,

 

Changed fuel filter and cam sensor today. Piece of p***! Will road test and try and recreate fault.

Looked at IMV and temp sensor. Seem OK to get to. Live data still indicates disparity of about 60 bar at idle required vs actual pressure. Wiggled sensor wiring, no erratic reading on live data. Diesel specialist wants £70 for diagnosis. Which is the exact cost of new IMV and Temp sensor!

 

What approach would you guys take? Garage or component replacement?

 

Cheers!!

[Tdci-Peter]

1 hour ago, Mattypompy said:

Which is the exact cost of new IMV and Temp sensor

Does Forscan not have a PID for fuel temperature? That would check that sensor fairly easily. I think temp sensor would have to well out to cause your problems. It is mainly for fine adjustments to injection timing.

If the actual (sensor) pressure is 60bar lower then the required pressure, I think I would do a leak-off test first. This should be within the scope of diy, no high pressures, not so much worry about dirt, as it is all done in the return lines. Just a tenner or so for some pipe & bits off eBay, I would guess. There are several youtube vids of this. Excess leak-off will drag the fuel rail pressure down, if it is more than the pump can deliver.

Is the fuel offset the ratio between what is being actually injected, and normal (mapped) expectations? If so, the sensor must be ok. If actual (real) pressure was so high that the pump could not cope at idle, and dropped 60bar (sensor reading) below the requested, then surely it would show up. Remember Forscan & the pcm both see the sensor readings, not real pressure. There is no easy way to read real pressure (unless you happen to have a 2000Bar sensor & some 2000Bar fittings handy)!

But - If the actual (sensor) pressure was 60bar higher than required, it all changes, and the sensor almost has to be faulty. Like the pump can not drop down to the real pressure that the faulty sensor is requesting.

If leak-off is ok, I think I would do the IMV, since it is not too expensive. If that fails it almost has to be the pump. Then its decision time: Sell (as a semi non-runner), quotes from specialists & pay up, or 2nd hand pump, with a big, not very easy diy job & no real guarantee.

---

But of course, all this may be irrelevant, and the fuel filter fixed it - we live in hope?!

 

[Mattypompy]

Sorry, my error I meant Pressure Sensor in both quotes! Duh! The one coming off the 'spider accumulator rail' that you mentioned previously. Just supplant Temp for Pressure!

Yes I have read about the leak off test. Doesn't look technically demanding I may go down that route.The diesel specialist said that would be included in the £70 diagnosis.

Don't know what the fuel offset is, maybe is actual vs expected in mg/s. I should check!

I think I understand your logic behind whether the real pressure is higher or lower than required for what the implications are for the sensor. I would have thought if it was reading too high then it being ostensibly lower anyway, the car would surely stall if the discrepancy was significant. If it's reading low then that could be a fault in the sensor causing the limp mode. If it's accurate then that suggests the fault is elsewhere. I can eclectically test it with the voltage gradient increase across the resistor and hopefully rule in or out. Going to check fuel lines too.

Yes I'll get a  quote if the pump needs replacing, if proscriptive, then try a PX maybe. Might check out old filter and use a magnet to see if any metal has got in from pump. Apparently that can happen.

 

Cheers.😀😀😀😀

 

Firstly I will burn up and down the M27 first to see if today's work has helped!

[Mattypompy]

 

Definition 'Injector Offset': The delay beteween when the injector driver fires and when the injector actually starts to open 

Cheers.😀😀😀😀

 

Firstly I will burn up and down the M27 first to see if today's work has helped!

 

'The offset value (often referred to as the dead time, lag time, or latency value) is used by the ECU to compensate for the loss of flow during the opening and closing period of each pulse, where full (aka constant or linear) flow has not been achieved. While the numerical values needed for data entry in the ECU are listed in the “Injector Voltage Offset” table at the bottom of the Data Match Technology information sheet, the accurate matching of this characteristic for each injector ensures that the ECU, which uses only one correction value for all injectors, is able to instruct the injectors to dispense exactly the right amount of fuel.

WHY? While offset matching will improve the result with all injectors, it is especially critical for injectors with a large flow rate (for example1000cc and larger) at idle or light throttle. This is because the offset correction length applied by the ECU represents a much higher percentage of total pulse width, so any error is greatly amplified.

For example, a 2150cc injector may need a 1.1msec total pulse width to dispense the right amount of fuel at idle, of which 0.7msec is the offset correction at 13 volt, resulting in an effective pulse width of 0.4msec. If there is a 0.2msec difference in offset value from one injector in that set to another, the effective pulse width could vary by 0.2msec/0.4msec=0.5 or 50% within the same set, which would easily cause enough variation in AFR to result in a misfire. Using the same variance in a smaller injector, like a 650cc, which requires a 2.3msec pulse width and has a 1.0msec offset value, results in an effective pulse width of 1.3msec. If the 650cc injector set has the same 0.2msec variance from one injector to another as the 2150cc injector in the example above, it is only a 15% fueling error, which is much easier for the motor to forgive.

Also at play is the fact that in large performance injectors the effective pulse width range you are operating in (0.3 to 0.8msec) is SO short that the injector is only performing an incomplete valve opening, which is an action that is much more challenging to repeat consistently than the fully settled opening at 6 or 7msec. Therefore a well-matched injector set is easy to tune and performs well in idle, cranking and other light throttle conditions.'

[Guest]

All very interesting

[FatHead1979]

12 hours ago, Ladiesman020 said:

All very interesting

We love a good "Necro Poster" on here 

[Guest]

I'm having me own fuel pump issues

[Guest]

What has happened to the car in regards to the fault.

[Daniel260491]

Hello all I know this thread hasnt been replied to in a while but just wondering what did the problem end up being as I have been getting the exact same symptoms with my jaguar xtype 2.0d . 

[nicam49]

It's usually a weak pump. They're getting quite long in the tooth by now. If you've got a code reader that runs on Forscan, then you can test the pump in this way: disconnect the electrical connectors to all 4 injectors and to the IMV., then monitor the frp (fuel rail pressure) while cranking the engine for <30secs. If the pump is good you should see a min of 1000 bar. Mine is weak and I get 850 bar and the same error as you're seeing.