Removing the fuel rail---AHA A4 2.8L 6V

[cparke]

Ok, for those of you who haven't read my previous posts, I have an engine that has been running rich (multiplicative only, more on Bank 1 than Bank 2) since I first used fuel cleaner treatment as a fuel additive. Using more fuel cleaner has just made the problem worse.

At this point, I want to try swapping the fuel injectors between the banks to see if the fuel trim values change banks with the injectors (nice thing about 6V is there are two of everything already installed on the car), if so, then I've found the problem! (if not, they're all fine and that isn't the problem).

To remove the fuel rail, I have to disconnect the fuel lines of course. While one of the fuel hoses going into the fuel rail looks fairly simple to disconnect by loosing the tightening coil and pulling the hose off the rail end, the other hose has some special nut screw-on connection (like a pressure connection) that is really tight. Is it true all I need to do to disconnect the second hose is to carefully loosen the connection nut? Is there any other special advice anyone has about dealing with the fuel lines before I attempt to remove the fuel rail? (other than pull the fuse on the fuel pump and run the engine until it stalls)

Other possible causes for the too rich condition that have been already suggested to me:
- MAF: I finally bought a new MAF, all air readings are now perfect, still engine is running just as rich. Guess I can sell the old one as used once I get out of this fuel trim mess.
- Intake boost leak, vacuum leak (I changed the small vacuum lines recently, but there are others that I did not, no changes, need professional equipment to diagnose further I figure)
- Engine coolant temperature sensor (I have the new sensor, but it is very difficult to get down to where it is to install, )
- O2 sensors (unlikely I could work on a part so deep under the hood, they test out fine in every other respect, but could they really be bad?)
- Plugged catalytic convertor (totally out of my league it would seem, hoping it doesnt come to this)

Thanks for your thoughts/advice.
CP

[hiwords1]

the 02 sensors are not as hard to get to as you think. the passenger side one can be accessed after you remove the intake box. with the box out, you can do the ECT also. the driver's side 02 sensor is accessible if you undo the 3 screws for the coolant reservoir and move it out of the way. there's only 1 MAF for both banks so there's no way that's gonna affect only one bank. just take some time and stare at your engine bay and look around a bit. things are more accessible than most people think on the 2.8 engines. i think you can check how much fuel is going through the injectors with vag com but i'm not totally sure

[MattzWarsteiner]

Fuel system guidelines:

Wipe the connections and surrounding parts off first.

For the pressure connection on the rail, use a flare nut or line wrench. It wraps around the nut further than an open end wrench and will protect the nut from distorting. Use another on the mating side to keep it from twisting.

Cover the lines with a rag to prevent the residual pressure from spraying out. When things are apart, keep the connections clean and covered to keep dirt out.

Since this apparently all started with the injector cleaner, I would check the operation of the FPR while you are in there. All you need is some temporary fuel hoses & clamps, a T and a pressure gauge rated for fuel.

[cparke]

You guys are great! Thanks so much.

[audispeed]

why are you pulling the fuel rail to swap injectors?

[cparke]

Fuel trim tonight on bank 1 -18% and on bank 2 -24%. Swapping injectors between the banks to see if the trim values move too, indicating if injectors are the problem or not. Seems like a simple procedure to me.

[audispeed]

It is simple. You don't need to disconnect the fuel rail from the fuel lines though - at least I didn't. I don't think your injectors are the problem. You are pulling an awful lot of fuel. Have you checked the vac line to the FPR? The O2's can be changed from in the engine bay - just wear gloves. Can you post a log of your O2 and MAF values?

[cparke]

Yes, I can post a log. Which measuring blocks and what driving conditions?

I have a new MAF so that is not the problem.

Swapping O2 sensors was another idea, but they are so far down I don't think I could reach, if I could get down there. The driver side O2 sensor seems so inaccessible, I can't even see it. Don't I need a special tool to remove O2 sensors?

[cparke]

You don't need to disconnect the fuel rail from the fuel lines though - at least I didn't.

I have a V6 with two fuel banks. If I don't disconnect the fuel lines, how can I pull the injectors out straight up?

[MattzWarsteiner]

On those posted trim pcts, both banks are out of spec and the difference is about 1/3. I think the bank differences are due to one worn injector on the richest bank. Rail off with the injectors hanging, you could do a quick drip test to find the bad guy on the -24% side.

First thing I would check is the FPR before pulling the rail, tho.

Give the FPR vacuum hose a visual then pull the hose on the other end and see if it can hold a vacuum. Since hoses were off before, I would double check the routing especially this one. Fuel in this hose = FPRs diaphragm failure.

Rig up a temporary gauge and see if your fuel pressure is within spec. Probably about $20 worth of brass and hose to do it right.

Engine at idle, AC off:
55-61psi with the FPR vacuum line off
46-55 and FPR connected

Key off.

Residual PSI, assuming a good fuel pump check valve:
30 cold to 40psi warm engine after 10 minutes is good. Anything less
points to injector leakage.

[cparke]

Give the FPR vacuum hose a visual then pull the hose on the other end and see if it can hold a vacuum. Since hoses were off before, I would double check the routing especially this one. Fuel in this hose = FPRs diaphragm failure.

Oh, I was very careful about that particular vacuum hose. I can say with 100% certainty that the vacuum line from the FPR that I replaced had a leak in it for perhaps a very long time. Could that have caused the FPR to get damaged? How could fuel cleaner have worsened the situation?

Engine at idle, AC off:
55-61psi with the FPR vacuum line off
46-55 and FPR connected

Ok, I don't exactly follow this very helpful procedure of yours.

How do I measure FPR pressure with the FPR connected to the vacuum line (step 2 above)? Also, am I measuring vacuum going into the FPR, or fuel "pressure" coming out of the FPR into the vacuum line?

[audispeed]

Our cars have a rising rate FPR. If its not seeing vacuum, its putting out higher fuel pressure then it should. Now that you have fixed the lines drive around for 50 miles or so and see what the trims look like. Or disconnect and reconnect the ecu and drive for a while to see what they look like. And yes, I have a v6 as well and you dont have to disconnect the fuel rail. There is enough slack in the lines to remove the injectors.

[cparke]

Our cars have a rising rate FPR. If its not seeing vacuum, its putting out higher fuel pressure then it should.

Higher fuel pressure would mean richer mixture I presume?

Now that you have fixed the lines drive around for 50 miles or so and see what the trims look like. Or disconnect and reconnect the ecu and drive for a while to see what they look like.

That has already been done. No change. Maybe the fuel check valve is bad. Maybe there is a leak somewhere in the lines I didnt fix so what I did doesn't matter anyway. But the fact fuel cleaner caused this to happen, and to worsen, still makes me think the problem is in the fuel delivery system.

And yes, I have a v6 as well and you dont have to disconnect the fuel rail. There is enough slack in the lines to remove the injectors.

I guess if I lay it upside down over the airbox I could keep it connected. I might be swapping the whole fuel rail with injectors and regulator with another one instead, however.

BTW - can I remove the FPR without first removing the fuel rail from the manifold?

[audispeed]

Higher fuel pressure would mean richer mixture I presume?

Yes

That has already been done. No change. Maybe the fuel check valve is bad. Maybe there is a leak somewhere in the lines I didnt fix so what I did doesn't matter anyway. But the fact fuel cleaner caused this to happen, and to worsen, still makes me think the problem is in the fuel delivery system.

I guess if I lay it upside down over the airbox I could keep it connected. I might be swapping the whole fuel rail with injectors and regulator with another one instead, however.

BTW - can I remove the FPR without first removing the fuel rail from the manifold?

Yes, the FPR can be removed with the fuel rail still in place. I don't think the problem is the fuel rail though. Do you have any logs of the O2's.

[MattzWarsteiner]

The cleaner could have done something to the FPR, yeah CP. The check valve in the fuel pump is really there to make it easier to start a hot car or car that has been sitting and would not cause a rich condition. I agree with Audispeed, the rail is just piping. Its not going to affect your trims.

A leak you would smell and would have to be massive to affect the trims and when it did, they would go lean instead of rich I think.

Sorry about the ambiguity on the FPR test. The pressures listed are for Fuel in PSI and this is out of the service manual. You will need to break in after the FPR and get some flare nut to nipple fittings for the gauge rig. < < No OEM hose cutting required

AudiS, I think the reason the book says to remove the rail is to avoid stress on the pipes leading to the injectors, but a careful tech hehe might not have any trouble with it.

[cparke]

How difficult are the 6 injectors to remove from the rail? I'm a little concerned because the gasket on the engine coolant temperature sensor was so tight! and I'm not looking forward to pulling 6 injectors attached to the metal rail out the same way...

[MattzWarsteiner]

I could be wrong, but I think the fuel pressure may be too high on this car. Both banks are too high, so unless this fuel additive is really bad news, I don't the injectors were damaged but the FPR may have been and this is what I would want to check.

[cparke]

Forget the original problem with both banks too rich; I had a very stupid repair shop (but with very smart mechanics working for it) fix whatever the problem was without even knowing they fixed it. Really! I took the car away from them because they had the car for 3 days and all they could tell me is "everything looks fine, it may be the MAF even though you say you installed a new one, and we are requiring you to buy another MAF through us in order for us to further diagnose." Like hell! Turns out the car is now actually running correctly but you wanted to start changing parts (non-refundable) based on guess-work that I already considered before I brought it to you?

Obviously, the shop can't/won't tell me what they did (or they don't even know what their own people did on the car), but when they gave me my car back "unrepaired" on my demand, it has been repaired somehow and the fuel trims are now within specified limits and the check engine light is not coming on anymore and not going to come on. Not sure what the fuel cleaner did, maybe it just screwed up some difficult-to-reset learned value in the ECU or something like that (OT: what is the proscribed method to fully clear all ECU learned values, and what is the proper procedure to subsequently cause these values to be re-learned?).

However, I'm not entirely out of the woods yet. While Bank 2 multiplicative fuel trim is now normal at -2.3%, Bank 1 multiplicative fuel trim is moderating between -8.7% and -10.3%, slightly rich but just within specified limits. This is probably the way the car was when I originally bought it. The problem is not as urgent now anymore, but it also smells to me more like a bad fuel injector issue on bank 1 than before, so I still want to pull the fuel rail at some time in the near future to swap the injectors between the banks. That is why I am now asking how difficult it is to actually remove the injector from the manifold with its gasket on.

Any of you with experience actually removing the fuel rail, please tell me if it really will just pop out when pulled upward or if each injector needs to be pried out with a tool, and upon the re-install, if the injectors will just pop back in or if I need to use some lubricant on the injector's gasket to get them in?

[MattzWarsteiner]

Keep driving it. The trims will probably rise more if the prob still exists.

[cparke]

Keep driving it. The trims will probably rise more if the prob still exists.

Yes, I'm keeping my eye on it. If it is the FPR, you'd expect the rise to be very slow gradual like this? However, FPR would not explain the spread in trim between Bank 1 and Bank 2, only a bad fuel injector comes to my mind to explain that.

[MattzWarsteiner]

I agree on the bank diff. Probably one injector causing the imbalance.

[cparke]

I agree on the bank diff. Probably one injector causing the imbalance.

1 week later, fuel trims are the same as they were last Friday and CEL has never lit. The intermittent load calculation cross-check is still a pending DTC in the ECU, how much longer do you think it will be before that goes away on its own? Idle air ctrl learn value (group 55) also has remained within spec. the whole time (this value had also been going high previously).

As to the bank 1-2 significant trim difference (bank 1 trim has fluctuated from as little as -8.7 and to as much as -11.7 this week, not enough to throw a too rich DTC at any point), besides a bad injector isn't it also possible that the bank 1 O2 sensor is giving out bad readings but not enough to throw an O2 code? Should I worry about fuel trim at those values even though no CEL?

It is very bizarre that the repair shop unknowingly fixed my problem by just doing their computer diagnostic and then telling me to try change my MAF again, which I refused to do and took the car back 'unrepaired' according to them. They may have done something else too, but they won't tell me any detail on what else they might have done and it couldn't have been anything major.

[MattzWarsteiner]

O2 sensor failure modes....

fatigued sensors usually report lean or excess O2, which would make your trims go to the + side not -.

damaged sensors caused by overly rich mix or burned oil/coolant in the exhaust stream or a failed heater will often under-report O2 and this would make the trim go negative in the early stages, but when really clogged, the ECU will notice the lazy response and flag them, lighting the CEL.

How fast the trim values change depends on the weighting used in the ECU program and Im not sure how that is done in our cars. I do know that the smaller the difference between commanded mixture and reported O2 levels, the longer it will take for the trim values to change.

Hey man, hope it stays good. Never know, right?

[cparke]

O2 sensor failure modes....
How fast the trim values change depends on the weighting used in the ECU program and Im not sure how that is done in our cars. I do know that the smaller the difference between commanded mixture and reported O2 levels, the longer it will take for the trim values to change.

Thanks Matt for your thoughts...

I will say this much... when I had the problem, when I reset the trims, initially the trim was good, but while driving it would suddenly jump up (like double) in a single sample and then never return. I still have the log from when that happened. Based on your description of how fuel trims are adjusted, sounds like an intermittent leaky injector, yeah?

[cparke]

Keep driving it. The trims will probably rise more if the prob still exists.

Last week, I had two cylinders (paired on the coil) start to misfire at the same time and check engine blinking. Had to change the coil even though it is only 4 months old (warranty covered the part). Not sure if the car;s problems somehow damaged the coil, but let's not dwell on that just now.

Computer had to be reset following the repair, of course. Now my fuel trims are again on the rise. Whatever the computer diagnostic at that repair shop did, somehow their procedure must have just given the computer certain learn values that it liked to keep under the car's condition but which are hard to get it to re-learn with normal driving. Having lost those, I am now back where I was before with rising trims, as I feared would happen. I sure wish I knew what those learn values were and how to get them back!

Anyway, let's re-cap for a minute everything I've now done with the car to try to address this problem, each unsuccessfully:

1) Full tune-up: fuel filter, coil, wires, spark plugs.
2) New OEM Gas cap changed
3) MAF changed
4) Vacuum lines changed
5) Engine coolant temp. sensor changed
6) Fuel injectors swapped between banks, and 1 new fuel injector installed on the more troubled bank (bank 1) too.

My short-list of remaining things to try to resolve this too rich problem, not sure what order to go next:
- Fuel pressure regulator
- Spark plug gap (shouldn't be bad because it is new & professionally installled)
- Spark plug wires (shouldn't be bad because it is new & professionally installled)
- Compression test & possibly valve/valve gaskets
- Oxygen sensor.

The basic symptoms are, following a computer reset, multiplicative fuel trims continually become more rich each day, while additive trims remain normal. Bank 1 is always more rich than bank 2, usually by a factor of 3:1 to 2:1. Both trims do rise however. Fault codes are thrown eventually.

Please help me find this problem! Even if I could get the computer to be happy with a -10.2 trim again on bank 1, I would be very happy. This has been going on for much too long already, and I really need to put this problem behind me!

[cparke]

Keep driving it. The trims will probably rise more if the prob still exists.

Last week, I had two cylinders (paired on the coil) start to misfire at the same time and check engine blinking. Had to change the coil even though it is only 4 months old (warranty covered the part). Not sure if the car;s problems somehow damaged the coil, but let's not dwell on that just now.

Computer had to be reset following the repair, of course. Now my fuel trims are again on the rise. Whatever the computer diagnostic at that repair shop did, somehow their procedure must have just given the computer certain learn values that it liked to keep under the car's condition but which are hard to get it to re-learn with normal driving. Having lost those, I am now back where I was before with rising trims, as I feared would happen. I sure wish I knew what those learn values were and how to get them back!

Anyway, let's re-cap for a minute everything I've now done with the car to try to address this problem, each unsuccessfully:

1) Full tune-up: fuel filter, coil, wires, spark plugs.
2) New OEM Gas cap changed
3) MAF changed
4) Vacuum lines changed
5) Engine coolant temp. sensor changed
6) Fuel injectors swapped between banks, and 1 new fuel injector installed on the more troubled bank (bank 1) too.

My short-list of remaining things to try to resolve this too rich problem, not sure what order to go next:
- Fuel pressure regulator
- Camshaft position sensor G40
- Spark plug gap (shouldn't be bad because it is new & professionally installled)
- Spark plug wires (shouldn't be bad because it is new & professionally installled)
- Compression test & possibly valve/valve gaskets
- Oxygen sensor.
- Something else?

The basic symptoms are, following a computer reset, multiplicative fuel trims continually become more rich each day, while additive trims remain normal. Bank 1 is always more rich than bank 2, usually by a factor of 3:1 to 2:1. Both trims do rise however. Fault codes are thrown eventually.

Please help me find this problem! Even if I could get the computer to be happy with a -10.2 trim again on bank 1, I would be very happy. This has been going on for much too long already, and I really need to put this problem behind me!

[MattzWarsteiner]

It's interesting that after a reset & coil change, the ECUs back to reporting rich. Wish I could offer more insight into why that is especially if it digital electrickery turns out to be the real problem.

Anyway, you know what I would want to check first.

[cparke]

Anyway, you know what I would want to check first.

Changed the fuel pressure regulator and the camshaft position sensor (G40) tonight; no change in fuel trim learning. I'm at the end of the line here, nothing left seems relevant to change except the ECU itself. Even after changing that, I still have doubts the problem will be resolved. What a nightmare!

[MattzWarsteiner]

Changing the regulator ruled out the valve itself but not the system, which is why I was interested in the pressures on the rail. Are you getting a good vacuum at the FPR? You said the line may have been broken for awhile, maybe the vacuum port leading into the manifold is gunked up somehow.

Here's a thought, the fuel cleaner totally clogged the return line to the tank and the regulator can't keep up. One way to find out if this is the cause or it is vacuum related is with a fuel pressure check.

[cparke]

Changing the regulator ruled out the valve itself but not the system, which is why I was interested in the pressures on the rail. Are you getting a good vacuum at the FPR? You said the line may have been broken for awhile, maybe the vacuum port leading into the manifold is gunked up somehow..

I changed the small vacuum lines a few months ago. I also recently ran the car with the FPR vacuum line disconnected to check this very thing. No fuel came out of the out FPR while car was running with the vacuum line disconnected, and yes there was air getting sucked down the FPR line. As for strength of the vacuum, no I didn't test it, but I could block it with my finger without any pain. The vacuum line from FPR has a check valve to protect the rest of the system from getting gunked up.

Here's a thought, the fuel cleaner totally clogged the return line to the tank and the regulator can't keep up. One way to find out if this is the cause or it is vacuum related is with a fuel pressure check.

I don't have an appropriate guage to measure fuel pressure. Disconnecting the fuel line from the fuel rail to put a guage in the middle is another messy task too. I also don't understand why one bank is consistently more rich than the other, even with a new injector installed and the other two swapped.

[cparke]

Latest theory is the engine is burning oil with the fuel, causing the sensors and ECU to think the engine is running rich and to cut back on fuel. How can I confirm if that is actually the cause, and what is involved to fix that if so?

[cparke]

Rig up a temporary gauge and see if your fuel pressure is within spec. Probably about $20 worth of brass and hose to do it right.

Engine at idle, AC off:
55-61psi with the FPR vacuum line off
46-55 and FPR connected

Key off.

Residual PSI, assuming a good fuel pump check valve:
30 cold to 40psi warm engine after 10 minutes is good. Anything less
points to injector leakage.

Hi Matt,

I finally found an appropriate fuel pressure gauge and got the nerve up to disturb the factory installation of the fuel lines in the Engine Bay and install a fuel pressure gauge to diagnose my ongoing fuel trim too rich multiplicative problem. Yes, I still have the same problem even after all this time.

The results are quite strange, and pressure does appear too high.

Fuel pressure
without the vacuum line connected to the FPR: 64 PSI
with the vacuum line connected (approx. 18inHg Vac) 56PSI.

Fuel pressure does jump as expected whenever the engine is initially revved.

At engine shutdown, fuel pressure usually (but not always) drops immediately to 42PSI, but then over the next 10 minutes rises to 58-60PSI! How the hell does that happen with everything shut off? Then it will slowly fall over the next few hours, after 24 hours it will reach 0PSI.

Okay, this doesn't sound right to me. It was a rainy day with a LOW PRESSURE atmosphere outside too, if that matters any. Bentley and you said the specified value without the vacuum hose connected to the FPR should be approx. 3.8-4.2 bar (55.1-60.9 psi). I'm a bit higher than that. It also talks about residual pressure 10 minutes after shutting off the engine being 31.9PSI on a cold engine and 43.5PSI on a warm engine. I have residual pressure rising actually and then taking hours to reach those levels.

Here's a thought, the fuel cleaner totally clogged the return line to the tank and the regulator can't keep up. One way to find out if this is the cause or it is vacuum related is with a fuel pressure check.

Is that what my fuel pressure results seems to be saying?