More HPFP failures but there is light at the end of the tunnel...

[JiveTurkey]

Very interesting read over at E90/E92 forum:

http://www.e90post.com/forums/showthread.php?t=294057

[jeremyc74]

It's interesting, but I'm going to be completely shocked if it's the pressure sensor that's been causing all these problems and BMW has just been replacing the whole pump instead of the part that's failing.

The pressure sensor just screws in like a typical oil pressure transducer, and had a plug on the back.

[JiveTurkey]

More good info from that thread:

http://www.bmwtis.com/tsb/bulletins/...p/B125506g.htm

[Robert]

If the case I hope the new sensor is engineered correctly. Really don't want to band-aid the old sensor with one of the same model and only wait for it to go south as well.

[Absorber]

That was a great read! Sounds promising!

[Lester]

Aha - looks like number 7 on the drawing.

07 Sensor, low pressure 13537547883 $91.91

[jeremyc74]

Originally Posted by Lester

Aha - looks like number 7 on the drawing.

07 Sensor, low pressure 13537547883 $91.91

That's the low pressure side. I'm pretty sure the one they're talking about is screwed directly to the pump and measures the high pressure side. It's part of the pump assembly so that's probably why they're replacing the whole thing every time.

[MadMan77]

Originally Posted by jeremyc74

That's the low pressure side. I'm pretty sure the one they're talking about is screwed directly to the pump and measures the high pressure side. It's part of the pump assembly so that's probably why they're replacing the whole thing every time.

According to Eric@AMS on the e90post thread he is talking about the low pressure sensor. Here is his post referring to this:

http://www.e90post.com/forums/showpo...4&postcount=16

[jeremyc74]

Originally Posted by MadMan77

According to Eric@AMS on the e90post thread he is talking about the low pressure sensor. Here is his post referring to this:

http://www.e90post.com/forums/showpo...4&postcount=16

I stand corrected. I didn't read through the whole thread and at the beginning it sounded like he was describing the one on the high pressure side. I'll REALLY be shocked if that low pressure sensor has anything to do with the failures we've been seeing. The ECU is capable of setting a completely seperate code for that being out of spec, but who knows. Those guys for sure aren't idiots so maybe they're on to something.

As a programmer I'm not sure why BMW would be concerned with the low pressure side, other than to make sure the pressure was sufficient, but maybe there's something strange going on in there.


For anyone who hasn't seen it, here's the technical info on how the HPFP works
(Starts on page 24)

http://www.1addicts.com/goodiesforyou/n54-2.pdf

High Precision Injection (HPI)

High-precision injection represents the key function in the concept for as economic a use of fuel as possible. The new generation of petrol direct injection satisfies the expectations placed on it with regard to economic efficiency without compromising on the engine's dynamic qualities.
High-precision injection provides for amore precise metering of mixture and higher compression - ideal preconditions for increasing efficiency and significantly reducing consumption. This is made possible by locating the piezo injector centrally between the valves. In this position, the new injector, which opens in an outward direction, distributes a particularly
uniform amount of tapered shaped fuel into the combustion chamber. The new direct injection of BMW HPI spark ignition engines operate according to the spray-directed process.

HPI Function

The fuel is delivered from the fuel tank by the electric fuel pump via the feed line (5) at an
“feed” pressure of 5 bar to the high pressure pump. The feed pressure is monitored by
the low-pressure sensor (6). The fuel is delivered by the electric fuel pump in line with
demand.
If this sensor fails, the electric fuel pump continues to run at 100% delivery with terminal
15 ON.
The high pressure pump is driven “in-tandem” with the vacuum pump which is driven by
the oil pump chain drive assembly.
The fuel is compressed in the permanently driven three-plunger high-pressure pump (8)
and delivery through the high-pressure line (9) to the rail (3). The fuel accumulated
under pressure in the rail in this way is distributed via the high-pressure lines (1) to the

piezo injectors (2).

The required fuel delivery pressure is determined by the engine-management system as
a function of engine load and engine speed. The pressure level reached is recorded by
the high-pressure sensor (4) and communicated to the engine control unit.
Control is effected by the fuel-supply control valve (7) by way of a setpoint/actual-value
adjustment of the rail pressure. Configuration of the pressure is geared towards best
possible consumption and smooth running of the N54 engine. 200 bar is required only

at high load and low engine speed.

Index Explanation Index Explanation

1

High-pressure line to injector (6) 6 Low-pressure sensor

2

Piezo injector 7 Fuel supply control valve

3

Fuel rail 8 Three pluger high pressure pump

4

High pressure sensor 9 High pressure line (pump to rail)

5 Feed line from in-tank pump

Index Explanation Index Explanation

1

Thermal compensator 8 Supply passage, pressure limiting valve

2

Low pressure non-return valve (check valve) 9 High pressure non-return valve (x 3)

3

Connection to engine management 10 Pendulum disc

4

Fuel supply control valve 11 Drive flange, high pressure pump

5

Return, pressure limiting valve 12 Plunger ( x 3)

6

Supply from electric fuel pump (in-tank) 13 Oil filling, high pressure pump

7 High pressure port to fuel rail 14 Fuel chamber ( x 3)

High Pressure Pump Function and Design

The fuel is delivered via the supply passage (6) at the admission pressure generated by
the electric fuel pump to the high-pressure pump. From there, the fuel is directed via the
fuelsupply control valve (4) and the low-pressure non-return valve (2) into the fuel chamber
(14) of the plunger-and-barrel assembly. The fuel is placed under pressure in this
plunger-and-barrel assembly and delivered via the high pressure non-return valve (9) to

the highpressure port (7).

The high-pressure pump is connected with the vacuum pump via the drive flange (11)
and is thus also driven by the chain drive, i.e. as soon as the engine is running, the three
plungers (12) are permanently set into up-and-down motion via the pendulum disc (10).
Fuel therefore continues to be pressurized for as long as new fuel is supplied to the
high-pressure pump via the fuel-supply control valve (4). The fuel-supply control valve is
activated by means of the engine management connection (3) and thereby admits the
quantity of fuel required.
Pressure control is effected via the fuel-supply control valve by opening and closing of the
fuel supply channel. The maximum pressure in the high-pressure area is limited to 245
bar. If excessive pressure is encountered, the high-pressure circuit is relieved by a
pressure-limiting valve via the ports (8 and 5) leading to the low-pressure area.
This is possible without any problems because of the incompressibility of the fuel, i.e. the
fuel does not change in volume in response to a change in pressure. The pressure peak
created is compensated for by the liquid volume in the low-pressure area.
Volume changes caused by temperature changes are compensated for by the thermal
compensator (1), which is connected with the pump oil filling.

Pressure Generation in High-pressure Pump

The plunger (2) driven by the pendulum disc presses
oil (red) into the metal diaphragm (1) on its upward
travel. The change in volume of the metal diaphragm
thereby reduces the available space in the fuel
chamber. The fuel thereby placed under pressure
(blue) is forced into the rail.
The fuel-supply control valve controls the fuel
pressure in the rail. It is activated by the engine
management system via a pulsewidth modulated
(PWM) signal.
Depending on the activation signal, a restrictor crosssection
of varying size is opened and the fuel-mass
flow required for the respective load point is set.
There is also the possibility of reducing the pressure

in the rail.

Limp-home Mode

If a fault is diagnosed in the system, such as e.g. failure of the high-pressure sensor, the
fuel-supply control valve is de-energized; the fuel then flows via a so-called bypass into
the rail.
In the event of HPI limp-home mode, turbocharging is deactivated by an opening of the
wastegate valves.
Causes of HPI limp-home mode can be:
• Implausible high-pressure sensor values
• Failure of the fuel-supply control valve
• Leakage in the high-pressure system
• Failure of the high-pressure pump

• Failure of the high-pressure sensor

There's some more info there, but it's a PITA to copy and past all the diagrams that go with it.

[oasis3582]

No way it is that easy. No way.

As much as I would like to believe that some guys and a dyno can solve a 3 year old problem before all of BMW, come on...

[MadMan77]

Here is the BMW SIB that is listing some of the troubleshooting in regards to this issue:

http://www.bmwtis.com/tsb/bulletins/...p/B125506g.htm

It mentions both the low and high pressure sensors amongst a few other items to troubleshoot.

[oasis3582]

Originally Posted by JiveTurkey

More good info from that thread:

http://www.bmwtis.com/tsb/bulletins/...p/B125506g.htm

Originally Posted by MadMan77

Here is the BMW SIB that is listing some of the troubleshooting in regards to this issue:

http://www.bmwtis.com/tsb/bulletins/...p/B125506g.htm

It mentions both the low and high pressure sensors amongst a few other items to troubleshoot.

Is there an echo in here?

[Yearofthe1]

Originally Posted by oasis3582

No way it is that easy. No way.

As much as I would like to believe that some guys and a dyno can solve a 3 year old problem before all of BMW, come on...

These aren't just "some guys." Those "guys" over at CP-E are ENGINEERS, you know, the type that go to college and get a degree and everything. Have you even seen what they've developed for the direct-injection market. These aren't off the shelf parts they're slapping together. For f-sakes, they've managed to tune a Mazdaspeed3 to hit the 400 hp mark.

As the Jamaicans so famously state - "Respect mon, respect!"

[oasis3582]

Originally Posted by Yearofthe1

These aren't just "some guys." Those "guys" over at CP-E are ENGINEERS, you know, the type that go to college and get a degree and everything. Have you even seen what they've developed for the direct-injection market. These aren't off the shelf parts they're slapping together. For f-sakes, they've managed to tune a Mazdaspeed3 to hit the 400 hp mark.

As the Jamaicans so famously state - "Respect mon, respect!"

I do have tons of respect for CP-E, but I just have trouble believing they could isolate an issue that has eluded the OEM for years. It would be extremely embarrassing for someone to solve this from under BMW's eye.

As an engineer myself, I know how failure investigation workscopes pan out. If it is possible for someone else to solve this, that just means BMW has been too cheap to issue a true fix. Neither is a positive scenario for them (but good for us!)

[Citizen Z]

Originally Posted by Yearofthe1

These aren't just "some guys." Those "guys" over at CP-E are ENGINEERS, you know, the type that go to college and get a degree and everything. Have you even seen what they've developed for the direct-injection market. These aren't off the shelf parts they're slapping together. For f-sakes, they've managed to tune a Mazdaspeed3 to hit the 400 hp mark.

As the Jamaicans so famously state - "Respect mon, respect!"

+1 - I've spent enough time on the phone with a few of their engineers to quickly realize they arn't just a bunch of smart guys with plenty of automotive experience and know how. They are a collection of mechanical and electrical engineers. At the end of the day, they are more of an engineer and design firm than an actual conventional auto tuner firm. I doubt they have out-of-the-box experience to build a competitive Outlaw 10.5 class car without climbing the same learning curve as everybody else. But they could probably put together their own custom turbo design from CAD.

[RPM90]

Originally Posted by jeremyc74

That's the low pressure side. I'm pretty sure the one they're talking about is screwed directly to the pump and measures the high pressure side. It's part of the pump assembly so that's probably why they're replacing the whole thing every time.

Hey it might be that simple. It would great if it is!

According to your tech post on how the system works, the low pressure sensor is what controls the high side fee/demand.
So, it's possible that it's affecting the HPFP in a negative way.
It could be calling for more and more pressure until the pump eventually fails.
It does seem plausible as we've speculated that it's not the pump itself, but something else. I was thinking in one post that perhaps it has something to do with the overall pressure control.
This could vindicate what we've been thinking all along, that it's not just the pump.

Even if this isnt is, obviously this is being seen by a lot of people, and with more and more people looking into it, something will be found.
I'm still sure BMW is working at this. I mean this isn't a creaking dash or squeaking seats. The "M" in BMW is for Motor afterall.
Even though I prefer "engine", but that's a different matter.

[jeremyc74]

Originally Posted by RPM90

Hey it might be that simple. It would great if it is!

According to your tech post on how the system works, the low pressure sensor is what controls the high side fee/demand.
So, it's possible that it's affecting the HPFP in a negative way.
It could be calling for more and more pressure until the pump eventually fails.
It does seem plausible as we've speculated that it's not the pump itself, but something else. I was thinking in one post that perhaps it has something to do with the overall pressure control.
This could vindicate what we've been thinking all along, that it's not just the pump.

I think you're misreading something in that post.

The fuel is delivered from the fuel tank by the electric fuel pump via the feed line (5) at an
“feed” pressure of 5 bar to the high pressure pump. The feed pressure is monitored by
the low-pressure sensor (6). The fuel is delivered by the electric fuel pump in line with
demand.
If this sensor fails, the electric fuel pump continues to run at 100% delivery with terminal
15 ON.

You can't monitor how much pressure a pump is producing from a sensor on the inlet side. There are two sensors. One checks the inlet side of the HPFP to make sure it's being fed at the right pressure, and one that's used to modulate the outlet pressure. The one these guys are saying they're working with is on the inlet side, and I can't figure out any logical reason it would have anything to do with the problem.

The low pressure pump isn't a variable pump. It's either on or off, and the only time you should see low pressure at the inlet side of the HPFP is if you're out of fuel.

I don't see any way the engineers that designed this engine and fuel system from the ground up could miss something this obvious.

On the other hand it did take a few of their field reps several days to diagnose a slipping clutch when we had it nailed down in about 5 posts with nothing more than a description of the problem, so I guess anything's possible.

[HondaGoneRogue]

Also from the posts, it seems more likely that the failure would be caused by something BEFORE the component rather than after the component . Of course, I have no idea about this issue I just figured I would throw my .02 in the pot. Here's to hoping we get a damn fix!

[RPM90]

Originally Posted by jeremyc74

I think you're misreading something in that post.



You can't monitor how much pressure a pump is producing from a sensor on the inlet side. There are two sensors. One checks the inlet side of the HPFP to make sure it's being fed at the right pressure, and one that's used to modulate the outlet pressure. The one these guys are saying they're working with is on the inlet side, and I can't figure out any logical reason it would have anything to do with the problem.

The low pressure pump isn't a variable pump. It's either on or off, and the only time you should see low pressure at the inlet side of the HPFP is if you're out of fuel.

I don't see any way the engineers that designed this engine and fuel system from the ground up could miss something this obvious.

On the other hand it did take a few of their field reps several days to diagnose a slipping clutch when we had it nailed down in about 5 posts with nothing more than a description of the problem, so I guess anything's possible.

I could be misreading it.
Here is what prompted my post:

HPI Function

The fuel is delivered from the fuel tank by the electric fuel pump via the feed line (5) at an
“feed” pressure of 5 bar to the high pressure pump. The feed pressure is monitored by
the low-pressure sensor (6). The fuel is delivered by the electric fuel pump in line with
demand.

They were suspect of the "low pressure sensor" right?
If so, then it does have an effect on the HPFP as stated above.

I was speculating that perhaps that "feed pressure" may have something to do with this.
Again, speculating.

My original thought though, as you clarfiy, would be more related to the pressure coming AFTER the pump. In that, perhaps that sensor is demanding more than is needed.
I may be misreading/understanding how much effect the low pressure sensor has on the high pressure sides output. Unless it does?

I read that thread, and someone wrote that BMW was suspecting that the ECU may have a fault in this. According to your tech post, the ECU controls the "high pressure" demand.
So, maybe there is something there as well.

Here is what the tech post says:
The required fuel delivery pressure is determined by the engine-management system as
a function of engine load and engine speed. The pressure level reached is recorded by
the high-pressure sensor (4) and communicated to the engine control unit.
Control is effected by the fuel-supply control valve (7) by way of a setpoint/actual-value
adjustment of the rail pressure.

[RPM90]

Originally Posted by HondaGoneRogue

Also from the posts, it seems more likely that the failure would be caused by something BEFORE the component rather than after the component . Of course, I have no idea about this issue I just figured I would throw my .02 in the pot. Here's to hoping we get a damn fix!

Well, if it's before the HPFP, then the low pressure sensor is suspect.
If it's after the HPFP, then it's the high pressure sensor and the ECU which takes and interprets that reading.

If the low pressure has something to to with overall high pressure, then it could be part of the problem.
I would guess if the low pressure sensor is causing the HPFP to be under supplied with fuel, that would cause a problem in that the HPFP would be working harder to keep fuel pressure up with an insufficient supply.

I remember reading a C&D, I think, article where they had the low pressure fuel pump replaced. It didn't really make sense what they meant as most failures were with the HPFP. Maybe they wrote the wrong thing, or maybe they just had a one off problem with their low pressure pump, the one that feeds the HPFP.

[jeremyc74]

Originally Posted by RPM90

I could be misreading it.
Here is what prompted my post:

HPI Function
The fuel is delivered from the fuel tank by the electric fuel pump via the feed line (5) at an
“feed” pressure of 5 bar to the high pressure pump. The feed pressure is monitored by
the low-pressure sensor (6). The fuel is delivered by the electric fuel pump in line with
demand.

They were suspect of the "low pressure sensor" right?
If so, then it does have an effect on the HPFP as stated above.

I was speculating that perhaps that "feed pressure" may have something to do with this.
Again, speculating.

My original thought though, as you clarfiy, would be more related to the pressure coming AFTER the pump. In that, perhaps that sensor is demanding more than is needed.
I may be misreading/understanding how much effect the low pressure sensor has on the high pressure sides output. Unless it does?

I read that thread, and someone wrote that BMW was suspecting that the ECU may have a fault in this. According to your tech post, the ECU controls the "high pressure" demand.
So, maybe there is something there as well.

Here is what the tech post says:
The required fuel delivery pressure is determined by the engine-management system as
a function of engine load and engine speed. The pressure level reached is recorded by
the high-pressure sensor (4) and communicated to the engine control unit.
Control is effected by the fuel-supply control valve (7) by way of a setpoint/actual-value
adjustment of the rail pressure.

The low pressure sensor MONITORS the feed pressure, which is held steady at 5 bar. The low pressure pump is inside the fuel tank, and it's the same thing that's used on the 128i. It's a typical fuel pump that's not variable pressure.

The ECU determines what the pressure should be on the HIGH side of the HPFP, and then adjusts it based on the reading from the HIGH PRESSURE sensor.

You can't control the outlet pressure of a variable PD pump based on it's inlet pressure. The inlet side is a constant 5 bar, and the low pressure sensor is just making sure it's where it's supposed to be.

Again, I can't come up with any logical reason the low pressure sensor would have anything to do with this. It's just there to make sure there's proper pressure feeding the HPFP, not to control anything.



Edit:

Ok, I'm taking another look at this, and now I'm questioning whether or not the low pressure pump is variable speed or not. There's another diagram in the tech document that shows some of sensors, and they're showing a transistor where the electric pump is. They also mention that if the low pressure sensor fails the electic pump will continue to deliever fuel at 100%, which suggests that it's capable of something less than that.

IF that's the case, and the LP sensor were failing in such a way that caused the pump to run at a much slower speed than it was supposed to maybe the HPFP is being starved, which is leading to the repeated failures after a short period of time.

[HondaGoneRogue]

Reading about the failsafes and the fact that we have two pumps reminds me of the systems on F-16 aircraft. (I am an Aircraft Fuels Systems Craftsman) On a system that has redundant components, problems are almost always masked by obscure symptoms and over-logical thinking. I have seen a lot of problems be solved by replacing components that no one would have thought of. And this is on a 30 year old 28 million dollar aircraft with all of Lockheed's engineers and cross-tells. So I deff can believe that CP-e or anyone for that matter could come up with an idea or solution. No company is infallible.