As far as I know BMW does not make a direct injection engine yet. So in answer to your question the injector doesn't go directly into the cylinder.

 

Search for HPFP

BMW has made several direct-injection engines. One is the 3.5 litre in the 335 and 535 (which has a problem with the fuel pump). I think some (all?) of the Diesels have it and the bikes may too.

Bluebee, you're right, it's not direct-injection.

 

1) As you can see in the diagram, it sprays just outside the cylinder, behind the valve.
2) Each sprays at its own specific time, in order to have the air/fuel mixture ready for the next intake cycle.

 

1) Each sprays at its own specific time

I think we've established all 7 injectors (9 on the V8) inject into the intake manifold now. Thanks. Now we just have to figure out if all 7 (9 in the V8) injector jets open at the same time or at different times.

In the diagram, which only shows two injectors, both spray at the same time (into the intake manifold).

I understand the diagram may be wrong ... (most diagrams show only 1 injector, plus the start injector, both firing simultaneously).

I'm looking for a published description of the E39 fuel injection system which describes WHEN the 7 injectors open up in relation to each other.

Googling, I've found L-Jetronic and D-Jetronic for Lucas, Bosch, VW, Porsche, etc., but it would help if I knew for sure what type of system we have.
For example, I presume the 2002 525i is Bosch L-Jetronic, but I'm just guessing (it could be D-Jetronic or Lucas or ?).

Do you know which of the following closest describes our E39 I6 seven fuel injector operations (specifically WHEN the jets open up)?
- L-Jetronic Fuel Injection
- L-Jetronic Fuel Injection Systems For BMW E12's
- L-Jetronic Basics
- L-Jetronic 912 Technical Manual
- L-Jetronic VW Workshop Manual

 

As a point of clarification, note that I said the primary injectors spray before the cylinder, not in the manifold. The manifold is the set of plastic runners going to the head, and the primary injectors are not in there. They spray into the ports in the head - before the cylinder, after the manifold.

I would imagine, though I can't say with certainty, the CSI is not linked to any one injector. I can tell you almost positively each injector sprays at its own prescribed timing based on the cylinder firing order. It wouldn't make sense to tie the CSI to an injector since they all have different timing, so it probably has its own rhythm.

 

There is no cold start injector on the e39. It is an indirect injection system where the injectors are located before the intake valve. Upon cold start the engine utilizes the signal from the crank sensor and goes into what is called batch fire. Batch fire is when an entire bank of injectors fire at one time, or all the injectors on both banks fire at one time. This is because the DME does not know where the piston is on it's travel on the number one cylinder. Once the location of the number one piston is established, the car utilizes the signal from the cam sensor, and goes into sequential fire. This is when each injector is pulsed one at a time for each cylinder, now that the DME knows exactly where the number one piston is on it's stroke. The total time the car is in batch fire is only about a half to a full second before it switches to sequential.

I'd also just like to add a quick note on why an engine must run richer on cold start, since the topic of cold start injectors was brought up. Many people simply say an engine must run rich because the engine is cold, without knowing exactly why. Liquid fuel cannot be burned, only it's vapors. So the job of an injector is to atomize the fuel into a tiny mist of droplets, so they can be burned properly. When you start a cold engine, the cold atomized fuel hits the back of the intake valve and condenses partly back to a liquid. So now you are left with a smaller percentage of burnable, atomized fuel vapor. So extra fuel is injected upon cold start to make up for the deficit that was lost due to condensation. Just a little food for thought.

Cold start is simply controlled by the DME lengthening the injector pulse width, or on time.

 

There is no cold start injector (CSI) on the e39.

Interesting. So the L-Jetronic diagram above is wrong for the E39. I guess that means the I6 has only six injectors and the V8 has only eight.

Do we have a good diagram for the E39 EFI?

It is an indirect injection system where the injectors are located before the intake valve.

To me, I was taught everything up to the intake valve is the "intake manifold", by definition. I guess some of you use a different definition of intake manifold. But, the main point is clear; the atomized fuel is available to the very next cylinder that is in the "pull" stage of pull-push-pow-puey in this order for the M54 engine: 1-5-3-6-2-4

Upon cold start the engine ... goes into what is called batch fire.

Interesting. Very interesting. The misfire we're trying to debug ONLY occurs upon a cold start ... so perhaps this bank-firing is a previously unknown clue.

Once the location of the number one piston is established, the car ... goes into sequential fire.

Interesting. So, the first second or so, all the fuel injector jets open; and after that, the fuel injector jets open in the order of 1-5-3-6-2-4 (for the M54).

So extra fuel is injected upon cold start to make up for the deficit that was lost due to condensation.

This makes sense that the pulse is lengthened when the engine is cold to compensate for fuel lost to condensation. I guess that simplifies things as we don't need to consider the non-existant cold start injector (csi) when debugging.

It sure would be nice to see a diagram of the E39 fuel injection showing what the pulses look like for each of the cylinders on the same time scale.

 

Once in sequential mode there is only one cylinder firing at a time. In a 6 cylinder engine that means that there is combustion once every 120 degrees of crank rotation.

On hot starts the engine still goes into batch fire until it establishes the position of the number one piston.....just like it would on cold start. But since the motor is already warm injector pulse is decreased because extra fuel is not needed.

And everything up to the cylinder is not the intake manifold. Air/fuel passes through ports in the cylinder head before it reaches the cylinder.

Also, the diagram you have is incorrect. It is for the older, L jetronic systems. Ours are motronic, meaning spark and fuel are controlled by one engine management system.

 

Great info Ryan, I didn't realize that.

 

Can we carry this fuel-injection quest for understanding a bit further to better understand the cutoff that occurs (in the misfire thread).

In that misfire thread, there is a fuel cutoff which, like an autoimmune disease, I believe "may" be the real problem more so than the initial misfire antigen that caused the severe fuel-cutoff reaction. (my hypothesis, not yet proven if it's even feasible).

That fuel cutoff "goes away", by all accounts, in the very next ignition cycle. It does not come back anytime soon (by all accounts), and it even goes away without an ignition cycle (by some accounts) simply by clearing the OBDII P1349 single-cylinder-misfire-with-fuel-shutoff code.

I don't understand how the fuel is shut off and more importantly, what turns it back on.

For starters ... does an ignition cycle always turn the fuel shutoff back on?

And, how severe must be a cold-start misfire be to seemingly immediately cause a fuel shutoff in the first place?

 

For starters ... does an ignition cycle always turn the fuel shutoff back on?

And, how severe must be a cold-start misfire be to seemingly immediately cause a fuel shutoff in the first place?

For the latter question, I believe that the fuel cutoff is initiated by the OBC as a result of some other electrical signal, such as detection of a misfire in a specific cylinder. It then cutoffs the fuel to that cylinder to preclude bigger problems. The severity of that type of deficiency needed to generate the fuel cutoff is probably defined as some threshold value (e.g. % of failed spark ignitions) buried in the OBC software. Your car is intermittently surpassing that threshold value. The intermittent nature is not too surprising as electrical problems (e.g. coil failure) are known to come and go.

Not sure about your first question as since you are getting a code, the OBC is remembering the code. Consequently, it should continue to preclude the problem (continued injection of fuel into cylinder 4). That the problem cures itself is a bit perplexing but so are all electrical and computer problems.

 

Fuel is cut off when the vehicle experiences a catalyst damaging misfire. A catalyst damaging misfire is categorized as a misfire that occurs at least once every 200 rotations of the crankshaft. Misfire detection is handled by the crankshaft position sensor mounted on the ring gear of the flywheel/flexplate. It contains a permanent magnet and when each tooth of the ring gear is passed by the magnet, an ac analog signal is produced. A misfire causes the crankshaft to slow down, so the signal to the DME is interrupted. This is how the DME recognizes a misfire. How much the crank slows down is how the DME determines the misfires severity. Also, there is a small gap between 2 teeth on the ring gear, this is where the number one cylinder is. So when the DME receives no signal for a brief moment, it knows the gap has just passed the sensor, and the number one piston is at top dead center. How long after this gap the misfire is detected, allows the DME to realize exactly which cylinder the misfire is occuring. That is why 99% of the time you can get the exact cylinder misfire via the OBD port.

The fuel is cut off by the DME by not providing a ground for the injector to that cylinder or cylinders. When the vehicle is shut off and turned on again, the short term fuel trims are reset and the vehicle starts from scratch. If the misfire has not been corrected then the whole cycle starts again.

 

So very informative and well presented. I know you did not get all that from the Bentley manual, so where did you learn it?

 

Donna what is the problem you are trying to diagnose? If there is already a thread on it sorry I must have missed it. I might be able to give you some help if you'd like.

 

what is the problem you are trying to diagnose?

Hi Ryan,

The original thread is titled:
How do diagnose a sudden and disturbingly unsettling cold-engine misfire?

In that thread, I'm not asking (per se) for the solution; I'm hoping to come up with a debugging procedure, to help myself and others in the future. To that end, I'm first trying to understand how the various parts work. The knowledge you've imparted has helped immensely already!

Many E39s, (based on searches), have a random single-cylinder misfire upon cold start only which causes fuel to shut off to a single cylinder (in my case, cylinder #4) until the next ignition cycle (or a clearing of the P1349 code that occurs as a result of the misfire and fuel shutoff).

Most of those E39 owners immediately attack simply by throwing parts at it, some religiously, others quite haphazardly (all with good intentions).

When this cold-misfire-fuel-shutoff-cycle first happened to me, I thought the car was going to shake itself to death, it's that severe. Idiotically, I drove for 15 miles, painfully attaining highway speeds, and the engine never smoothed out - as I was unaware at the time of the simple "fix" of cycling the ignition (or clearing the OBD II code P1349).

The problem has not returned since I had it (some say it's temperature and humidity related) yet researching the threads, there are very many people with the same rare intermittent cold-start single-cylinder misfire with fuel shutoff.

Therefore, I feel it's worth the effort to come up with a diagnosis, more for the edification of all than to solve the problem (which has a simple fix which is to shut the car off and turn it back on).

The exasperating thing is, that, in almost every one of those E39 threads I've found, something (almost anything related to the cylinder gas/air/spark/compression/timing cycle) is replaced and the rare cold start misfire "miraculously" goes away, making that particular E39 owner a rabid believer in the part they almost whimsically replaced (since, in most cases, as in mine, it would have gone away anyway due to the random nature of the problem).

Surprisingly, the number of single-fix solutions touted is astoundingly large including vanos, ccv (aka pcv/cvv), maf, tps, cps, icv, cts, 02s, dme, ecu, etc., and of course, the coils, plugs, and boots and just about every filter in the engine compartment.

I've taken a different approach than everyone else - I've done nothing, and the problem also just as "miraculously" has gone away, (so far). Obviously the intermittent nature of the problem makes any one proposed solution less valid, but, what is more valuable than a proposed solution is a sensible debugging plan to test.

The problem, as always with random problems, is to catch it in the act prepared with the battle plan.

So, I'm simply trying to come up with an understanding of how the E39 works (under single-cylinder misfire conditions) so we can come up with a sensible debugging plan to immediately spring into action the next time the random misfire occurs to any one of us.

 

swap the coil to another cylinder, if the miss follows it, its the coil. If it stays in cyl 4, swap the plug also, if it still stays in cyl 4 do a compression test.

 

If compression is good swap the #4 injector to another cylinder.

I would like to be able to turn the injectors off on a cold start, so I could crank the engine for about a second to get the oil going.

 

what bluebee wants to know is, what would cause a misfire, with fuel cutoff on a cold start.

 

While doing research to help someone today in another thread:
- need to clean fuel injectors and check connections

I realized we really don't have very good links for fuel injectors themselves in the VERY best of E39 Links.

This is the best we have so far (and it's not all that good):
- How to understand (1) (2); remove (1) (2) (3); inspect (1); test (1); and clean (1) (2) (3) (4) your BMW E39 fuel injection system; & where to buy fuel injectors (1) (2) (3).

Do you know of BETTER reference threads for fuel injectors?

 

This is a great thread, with Bluebee asking some good questions, and RyanM dropping some serious knowledge on us.

 

A somewhat related question was asked today...
> E39 (1997 - 2003) > Hypothetical idea

Not that I am going to do it, I was driving to work this morning and was wondering. If I put a two way kill switch on two injectors, so at crusing speeds I could save more gas by turning off the two cylinders manually. Could this cause harm to the engine? What two cylinders would I be able to do this on? Once again, I am not going to do this. It was just a pre work idea that came to mind. I enjoy the v8 two much to ruin it by turning it into a v6.

Sent from BimmerApp mobile app

 

For the record, the following related exchange occurred here today ...
> E39 (1997 - 2003) > Pinpoint question: Do we know what the double VANOS camshaft advance curve looks like

Everyone gets upset when I say nobody knows how the VANOS works; and I know it's a hard pill to swallow, but I'm only telling the truth (which is why it bothers everyone so much).

If I lied, to make you feel better, and I said everyone knows how the VANOS works; and I just want to ask some trivially simple basic questions, could you answer them?

Simple questions, like ...
Q: Does the vanos do anything at warm idle, or not?
If yes, what?

Q: We know the ECM takes in coolant temperature sensor and intake air temperature sensor inputs; but what does it "do" with them with respect to VANOS operation?

Q: At what RPM does the VANOS turn off?

Q: At what RPM (on a warm engine) does VANOS turn on?

Q: What is the BASIC (fundamental) shape of the intake cam curve (i.e., the number of degrees by RPM)?

Q: How does pedal position (i.e., load) fundamentally change that curve?

Q: What is the BASIC (fundamental) shape of the exhaust curve under those same conditions?

etc.

These are extremely basic questions that anyone who actually understood how the VANOS works, would be able to accurately answer in a split second.

I know it's a hard pill to swallow. I have no problem saying when I am clueless; so I understand that when I say the whole reason there is so much VANOS discussion is because nobody understands how it really works, it bothers people.

People THINK they know how the VANOS works (and, at the superficial level, EVERYONE understands how the VANOS works).

But, saying you understand how the VANOS really works is like saying you know how the fuel system works, but you have no clue why a 14:1 ratio is important.

It's like saying you know how our fuel system works, but you have no idea how the unused fuel gets back to the tank (or doesn't, as in the case of the E39).

It's like saying you know how a fuel system works, but you have no idea how advance before top dead center needs to change in terms of engine parameters such as RPM and load.

It's like saying you know how a fuel system works, but you have no idea what happens to engine vacuum at WOT versus at idle speed, versus midrange cruising.

It's like saying you know how our E39 fuel system works, but you don't even know how many injectors there are on the E39.

It's like saying you know how our fuel system works, yet you have no idea of the timing of when the injectors fire or if they all fire at once.

It's like saying you know how our fuel system works, but you don't even know where the fuel is injected into, whether it's the cylinder or the intake manifold.
etc.

If all you know about the fuel injection system is that fuel, at oh, say, somewhere below 50psi is injected at a 1:14 ratio with air, and somehow, someway, it gets to the cylinder that's firing, and then the unburned fuel somehow maybe perhaps makes it back to the fuel tank where somehow there are carbon filters and fuel regulators and fuel filters and sending units scattered about - well then - yep - you'd know VANOS about as well as this hypothetical person understands the fuel injection system.

If you UNDERSTOOD the fuel injection system, these trivially simple to answer questions would be known:

Q: How many injectors are there on our engines?
Q: How many of those are cold-start injectors?
Q: When do the cold start injectors fire, and why?
Q: Do the injectors fire into the cylinder or into the manifold?
Q: Do they fire at the same time, or a different time?
Q: Where does unburned liquid fuel go?
etc.

For more details on the fuel system, see a similar thread:
- How does the E39 fuel injection work (1)

There are, on average, 6.5 injectors per M54 engine.

They all fire at the same time, except under full-throttle, then they fire in banks of 3.

The excess fuel weeps out of the fuel system; this is known as the "angel's share."

'87 325i e30 answers:

These are simple questions not on par with what you're asking about vanos - slightly more interesting questions would be asking about how the DME manages temperature/advance curves, catalyst warm up times / when does the O2 sensor get factored in, what's the idle speed vs engine temp curve look like, etc.

BTW: why is not a good question to ask, because that answer is always "because the computer is programmed to do that based on a curve".

Where do you folks find all the information on injector batch firing in the e39?

I had read somewhere long long ago that cars built in the late 90s onward had mult-point injection systems that fired one at a time, in sequence.

It should be here:
- How does the E39 fuel injection work (1)

If not, let's put it there.

So, if I understand that thread and what Ryan M is saying, it's sequential after all, and only batch for the initial second or so at engine startup (because the engine brain doesn't know where cyl 1 is yet....).

Somehow reading through -this- thread, I felt that you guys were saying the e39 runs in batch mode with all injectors firing simultaneously.....

I haven't noticed this PDF mentioned. Page 40. the M54 is sequential or semi-sequential depending on the condition.

 

Geez, the Vanos is a mechanical variable system that responds to oil pressure. As I know it, there is no electrical/computer controlled alteration of the cam timing. There are no electrical solenoids altering advance/retard. Any electrical alteration of timing takes place ultimately at the spark plugs, this altered by the input you have listed above but does not effect the advance/retard of the cams themselves. Therefore, the advance curve of our cams is quite predictable.

You seem to be confusing how it works with what are the data points of advance/retard of cam timing with the vanos system. I believe someone already posted a diagram of the advance curve.

By posting how advance/retards effects hp/torque in a motor, I hoped it would help you predict what the cam is doing at any given time.

doc

Oh, I believe the answer to the question you proposed starting this thread is, neither. I think the injectors inject around the intake valve portion of the head, which is neither in the manifold, nor in the cylinder. There are motors that inject directly into the cylineder, but this requires another orifice into the cylinder, thus limiting valve size. The third common way of inducing fuel is a throttle body where the mix of air and fuel is induced pre-manifold and fuel is carried by air flowing into the cylinder.

 

I believe someone already posted a diagram of the advance curve.

Nope. Wrong curves for the question.

By posting how advance/retards effects hp/torque in a motor, I hoped it would help you predict what the cam is doing at any given time.

What's the camshaft timing advance, in degrees, at 700 RPM (idle throttle, and warm air & coolant) compared to the default camshaft position?

But, this thread isn't about the VANOS.

The only reason VANOS was brought up above, was to give the cross reference context.

This thread is about the fuel system, and how it works.

 

'Bee, why did you quote me?

 

Um. I thought it was relevant to this thread.
is it not?

 

No, I was being a pain! :rofl:

A 6-cylinder engine can't have half an injector.

The cylinders don't fire in banks of 3. Ever.

"Angel's share" is actually whiskey that evaporates from the barrel during the aging process.

 

No, I was being a pain! :rofl:

A 6-cylinder engine can't have half an injector.

The cylinders don't fire in banks of 3. Ever.

"Angel's share" is actually whiskey that evaporates from the barrel during the aging process.

Actually, they do - even on the e30s. If you unhook the ignition(cam) timing sensor, they revert back to batch fire.

Bee, was this an early thread? L-Jet died off before 1990 if I'm not mistaken. Also worth noting that batch/bank fire on the straight motors gives only minor economy/efficiency boosts, since the duration of the injector is often longer than the valve opening duration - and putting in large enough injectors to do a "whole shot" while the valve is open more often than not negatively affects fuel vaporization (talking MPFI here, not DI).

 

The question of the number of fuel injectors came up again today ...
> E39 (1997 - 2003) > NCT fail Exhaust Emissions High Idle Lamba @ 1.34

You are bringing back very fond memories of my 83 GTi, by mentioning a "cold start injector".

However, our cars don't have "cold start injectors". That was used on a much older Bosch mechanical fuel injection system from the 80s.

Our cars have one fuel injector for each cylinder. 6 cylinders = 6 fuel injectors.
Look back at the original code/description: "P1351 Misfire Cylinder 5 With Fuel Cut-off"

Once the DME sees the misfire, it simply does not "turn on" that fuel injector. This is to prevent damage to the cat.

Since there is no combustion in cylinder #5, the oxygen in the cylinder does not get consumed as it does in the other cylinders burning air/fuel and is that air is pushed into the exhaust making that bank lean / lambda > 1 / due to excess air.

That's my theory. (for now )

 

Interesting... As far I as know the "cold start injector" was only used on mechanical fuel injection and that was in the 80s.

Once the industry moved to electronic fuel injection, we didn't need a "cold start injector" as the computer/DME would lengthen the fuel injector pulse to add more fuel for cold start conditions.

BlueBee I love and appreciate all what you add to this forum, but why do you think the E39 has a "cold start fuel injector"?

 

Interesting... As far I as know the "cold start injector" was only used on mechanical fuel injection and that was in the 80s.

Actually, the early electronic systems (eg L-Jet) still used the cold start injector. I vaguelly recall there is some advantage to throttle body injection during cold start-up, but I forget what. And it is probably horrid for consumption/emissions.

I'd better stop rambling, since none of this is relevant to the E39!