davegsm82
Go Kart Newbie
- Location
- Blyth, Northumberland - England
- Car(s)
- MK7 Golf and T3 Tdi
My Golf (2.0 GT TDi 150 - CRB) died after driving to work, fortunately the fault presented itself 1/4 of a mile away and it limped/kangaroo'd there, albeit unable to get above 2500rpm and 30mph.
It restarted immediately after stopping in the car park but after it had cooled again a couple of hours later I couldn't get more than a cough out of it.
I plugged in VCDS and the following codes presented themselves;
5211 P0087 - fuel rail system pressure - too low
5212 P0087 - fuel rail system pressure - too low
5231 P0087 - fuel rail system pressure - too low
10058 P3045 - Fuel pump - electronics defective
10062 P3044 - Fuel pump - short circuit.
TLDR: Fuel pump controller under back seat had died. Read on for testing and analysis;
The first 3 fault codes were related to the fuel rail while the engine was (just) still running and didn't come back after clearing. The second 2 remained or immediately came back after cycling the ignition and point NOT to the HPFP on the engine, but the in-tank 'lift' or Low Pressure Fuel Pump.
Looking at the live fuel pressures during cranking they showed no significant rise in fuel pressure, which would usually indicate the High pressure Fuel pump on the engine has started to, or has failed completely, some details of this kind of fault at the end.
Running output tests of the Low pressure pump (in the engine module) without the engine cranking allowed me to hear an unusual sort of knocking/clicking noise, whereas I would usually expect a whirring/whining noise. Removal of a hose from the fuel filter in the engine bay showed little fuel flow and supplied in pulses. After being trailered home I did some tests and found that the pump controller under the rear seat wasn't drawing much power and with the rear seat removed you could hear the pump trying to run again in pulses.
The symptoms and fault codes all pointed to failure somewhere in the rear pump so I replaced both with parts from a scrapped car, contrary to advice I read online the pump controller does NOT need to be the same revision number, the original was part number 5Q0 906 121G, I replaced it with the same part number but revision 'P' which works perfectly.
Something to be aware of, is that the fuel pumps on these cars are NOT like standard pumps, you can't just feed them 12V and make them run. VW in their infinite wisdom (and probably something to do with Euro emissions rules) have made the pumps variable speed, demand increases with engine speed, so the ECU sends a PWM signal to the controller which spins the pump faster or slower. The following video shows an oscilloscope plugged into pin 6 of the pump controller, with the engine running and the accelerator being pushed you can see the width of the pulses increase with engine demand;
Running the output test will give a PWM signal on this pin to try and run the pump, if you aren't getting any activity from the pump there's a few things to try before pulling the trigger on a new controller or pump;
(With the controller unplugged, ignition on)
1 - check you have ~12V on pin 5
2 - check you have continuity (almost zero ohms resistance) between pin 4 and ground (the seatbelt latch bolt is a good clean ground).
3 - if you have an oscilloscope, look for the signal similar to the previous video on pin 6 while cranking or during output test.
4 - Check resistance between pins 1-2, 2-3 and 1-3, it should be pretty low, less than a couple of ohms in each case.
If all of those are correct, then you can look at the output of the controller using an oscilloscope;
This is the signal from any of pins 1, 2 or 3 and ground (pin 4). The frequency between bursts is 200Hz, this represents the rotation speed of the pump.
If you were to zoom in further into the signal you will find that the 'carrier' frequency of the PWM signal is about 15KHz, you may or may not be able to hear the pump singing if you have young ears and it's out of the tank.
If you have all of these signals (same signal on pins 1-3) and no/poor pump rotation then I would suggest that your individual problem is possibly a jammed/seized in-tank pump.
Failure of the High Pressure Fuel Pump
One of the failures attributed to low fuel-rail pressure is the failure and disintegration of the HPFP which is driven by the timing belt on the 2.0TDI's (and 1.6's). If this happens then it's usually uneconomical as the entire fuel system requires replacement or serious flushing to remove the metal particles which are distributed by the return line to the tank. A quick and simple check for this catastrophic damage is to remove the metering valve (N290) on the top of the pump, (2x Torx bolts) and look for any contamination, any shards of metal inside indicate pump failure.
The following Rosstech page has more information on this failure;
http://wiki.ross-tech.com/wiki/index.php/16471/P0087/000135
It restarted immediately after stopping in the car park but after it had cooled again a couple of hours later I couldn't get more than a cough out of it.
I plugged in VCDS and the following codes presented themselves;
5211 P0087 - fuel rail system pressure - too low
5212 P0087 - fuel rail system pressure - too low
5231 P0087 - fuel rail system pressure - too low
10058 P3045 - Fuel pump - electronics defective
10062 P3044 - Fuel pump - short circuit.
TLDR: Fuel pump controller under back seat had died. Read on for testing and analysis;
The first 3 fault codes were related to the fuel rail while the engine was (just) still running and didn't come back after clearing. The second 2 remained or immediately came back after cycling the ignition and point NOT to the HPFP on the engine, but the in-tank 'lift' or Low Pressure Fuel Pump.
Looking at the live fuel pressures during cranking they showed no significant rise in fuel pressure, which would usually indicate the High pressure Fuel pump on the engine has started to, or has failed completely, some details of this kind of fault at the end.
Running output tests of the Low pressure pump (in the engine module) without the engine cranking allowed me to hear an unusual sort of knocking/clicking noise, whereas I would usually expect a whirring/whining noise. Removal of a hose from the fuel filter in the engine bay showed little fuel flow and supplied in pulses. After being trailered home I did some tests and found that the pump controller under the rear seat wasn't drawing much power and with the rear seat removed you could hear the pump trying to run again in pulses.
The symptoms and fault codes all pointed to failure somewhere in the rear pump so I replaced both with parts from a scrapped car, contrary to advice I read online the pump controller does NOT need to be the same revision number, the original was part number 5Q0 906 121G, I replaced it with the same part number but revision 'P' which works perfectly.
Something to be aware of, is that the fuel pumps on these cars are NOT like standard pumps, you can't just feed them 12V and make them run. VW in their infinite wisdom (and probably something to do with Euro emissions rules) have made the pumps variable speed, demand increases with engine speed, so the ECU sends a PWM signal to the controller which spins the pump faster or slower. The following video shows an oscilloscope plugged into pin 6 of the pump controller, with the engine running and the accelerator being pushed you can see the width of the pulses increase with engine demand;
Running the output test will give a PWM signal on this pin to try and run the pump, if you aren't getting any activity from the pump there's a few things to try before pulling the trigger on a new controller or pump;
(With the controller unplugged, ignition on)
1 - check you have ~12V on pin 5
2 - check you have continuity (almost zero ohms resistance) between pin 4 and ground (the seatbelt latch bolt is a good clean ground).
3 - if you have an oscilloscope, look for the signal similar to the previous video on pin 6 while cranking or during output test.
4 - Check resistance between pins 1-2, 2-3 and 1-3, it should be pretty low, less than a couple of ohms in each case.
If all of those are correct, then you can look at the output of the controller using an oscilloscope;
This is the signal from any of pins 1, 2 or 3 and ground (pin 4). The frequency between bursts is 200Hz, this represents the rotation speed of the pump.
If you were to zoom in further into the signal you will find that the 'carrier' frequency of the PWM signal is about 15KHz, you may or may not be able to hear the pump singing if you have young ears and it's out of the tank.
If you have all of these signals (same signal on pins 1-3) and no/poor pump rotation then I would suggest that your individual problem is possibly a jammed/seized in-tank pump.
Failure of the High Pressure Fuel Pump
One of the failures attributed to low fuel-rail pressure is the failure and disintegration of the HPFP which is driven by the timing belt on the 2.0TDI's (and 1.6's). If this happens then it's usually uneconomical as the entire fuel system requires replacement or serious flushing to remove the metal particles which are distributed by the return line to the tank. A quick and simple check for this catastrophic damage is to remove the metering valve (N290) on the top of the pump, (2x Torx bolts) and look for any contamination, any shards of metal inside indicate pump failure.
The following Rosstech page has more information on this failure;
http://wiki.ross-tech.com/wiki/index.php/16471/P0087/000135
