Mysterious misfire cyl 7&8 after TCG work.

[M5 Jed]

Hoping the M62 experts chime in here.

I read Joe Kitsch’s thread of agony regarding the bent valve of cyl 3. I thought maybe I’m in the same boat. But now I don’t know.

I completed the TCG work. Car fired right up but it was shaking a bit. Idle was stable but the shaking was noticeable, like the engine mounts were bad, they are new btw. I scanned the car and there are multiple misfires across all of bank 2. I took compression readings of both banks. Below are the results in the posted pics. Bank 1 is perfect, bank 2 is a bit low. So I figured that I didn’t quite time the engine correctly. Ok fine, I’ll retime it again. Took apart the valve covers and upper timing covers. I’m using G.A.S. M62 timing tools to do this. I even had my Fluke Meter on audible tone during the whole timing and torquing process. At every step I checked, double checked and triple checked for continuity. I also turned the engine by hand 2 rotations and checked the continuity again. Everything looked good. So I assemble everything back up, and fire it up. It’s not misfiring on cyl 5 & 6 like before, but it’s still misfiring on cyl 7 & 8. I’m like, WTF???!!!

I do a second compression test. Results looked even better. Compression on bank 2 is now up to normal, albeit still a little lower than bank 1. To do compression tests, I remove fuses 17 & 31, clear out some excess fuel, then rotate each cylinder 8 cycles. Gauge usually tops out at the 6th cycle.

I swapped spark plugs between banks, no change.
I swapped coils between banks, no change.
I swapped cam sensors, no change.
I swapped Vanos solenoids, no change.

I haven’t bothered to look into the injectors or intake because that is all new. I had put in an M60 intake manifold and renewed all the gaskets 3 months ago and the car has been running flawlessly. I really doubt any of the wiring have issues, like I said, car ran flawlessly before. The only part of the intake I touched to do the TCGs was the main air duct.

Car has 140k. TCG work was done as a preventative measure. A new liquid cooled Bosch alternator was also added during the TCG work. Car also has a full Dinan CAI and MAF, along with a large bore TB, running on Dudmd’s Tune.

If any of the bank 2 valves even kissed the piston, no way the compression would even improve right? What am i missing?

I did notice that the brake booster hose from the booster going into the valve is rock hard, no cracks, but it is rock solid.

I’m about an inch away from torching this car and moving on. Never in my life have I come across such a time waster as this one.

[JimLev]

Might need a set of those headers you make to answer your questions. Just kidding.
Only a few questions at a time so we can move on after you answer.
1) What did you find was wrong to get the compression back up? Valves should be OK.
2) Your latest bank 2 reading look OK. We're all the plugs out and the TB wide open when you did the comp test?
3) Are you sure the dots on the cams were pointing up when you locked the cams down?
4) Check the injector connectors on 7 and 8.

Even a new brake booster hose is hard.

[M5 Jed]

Answers to your questions in the same order:

1. The engine was misfiring on all bank 2 cylinders. This gave me the suspicion that I did not time the engine correctly, and perhaps some valves were slightly open. So I retimed it again, ran the engine for a couple of minutes, let it cool and took a 2nd compression test. And what you see in the pic are the new readings. Seeing the bank 2 compression readings improve after the retime, gives me confidence that now I have the engine timed correctly.

2. All the spark plugs were removed in both compression tests. The TB was closed in both compression tests.

3. All the cams were in the correct position, with the engraved markings pointing up on the square bosses then I installed the cam lock boxes. I never paid attention to the dots. The other confirmation I look for both cyl 1 cam lobes pointing at about 45 degrees to each other.

4. I haven’t checked the injector connectors since they weren’t disturbed during this whole time. I remove the lids of the wire boxes and just move the entire coil out of the way, I don’t even disconnect the coils.

[Mannyf540]

Answers to your questions in the same order:

1. The engine was misfiring on all bank 2 cylinders. This gave me the suspicion that I did not time the engine correctly, and perhaps some valves were slightly open. So I retimed it again, ran the engine for a couple of minutes, let it cool and took a 2nd compression test. And what you see in the pic are the new readings. Seeing the bank 2 compression readings improve after the retime, gives me confidence that now I have the engine timed correctly.

2. All the spark plugs were removed in both compression tests. The TB was closed in both compression tests.

3. All the cams were in the correct position, with the engraved markings pointing up on the square bosses then I installed the cam lock boxes. I never paid attention to the dots. The other confirmation I look for both cyl 1 cam lobes pointing at about 45 degrees to each other.

4. I haven’t checked the injector connectors since they weren’t disturbed during this whole time. I remove the lids of the wire boxes and just move the entire coil out of the way, I don’t even disconnect the coils.

Are all your fuel injector clips accounted for?

I found one and a piece of timing chain guide in my valves.

[M5 Jed]

The injector clips were never touched. All the injector clips are correctly placed as well as the injector connector. When I remove the valve covers, i first remove the lids to the wire boxes. This allows you to lift the complete coil w/ connector and grommet and move it away so you can lift the valve cover straight up without snagging on the coil connectors.

The TCG job was preventative. All the old guides were inspected upon removal and they were all perfect, no pieces missing when compare to the new guides. So the possibility of debris in the lifters is nil.

[peter540]

I've never heard of anyone using your technique of pulling off the valve covers with the coils still in place. I see the logic, but it also sounds like something that might not work every time.

Have you tried pulling the coils on the cylinders in question and reseating them? Perhaps when you you replaced the valve covers with the coils still attached they didn't seat well.

[philly98540]

Maybe bore wash due to first go round with timing off. Doubt any valves are bent as pressure are high on bad side. Maybe timing still off? Also, TB should be held open for compression tes.

[M5 Jed]

I've never heard of anyone using your technique of pulling off the valve covers with the coils still in place. I see the logic, but it also sounds like something that might not work every time.

Have you tried pulling the coils on the cylinders in question and reseating them? Perhaps when you you replaced the valve covers with the coils still attached they didn't seat well.

Maybe I wasn’t clear in my description of my method. I do remove the ignition coil prior to removing the valve covers. But I don’t separate the coil from the connector. I remove the lid covers of the wire box (black plastic box on top of the injectors) this allows you to lift the ignition coil still connected to the connector, and lift it completely out of the way when it comes time to lift the valve cover off.

Yes I verified that the coils are seated correctly. They can only go in one way.

[M5 Jed]

Maybe bore wash due to first go round with timing off. Doubt any valves are bent as pressure are high on bad side. Maybe timing still off? Also, TB should be held open for compression tes.

It’s possible. I’m leaning towards that too, that no valves are bent.

Hmm, I’ve never done compression tests with the TB open before. I’ll try that next time.

Just thinking about it logically. With TDC and flywheel pin in place, and all 4 square bosses on the cams pointing face up, then loosen the cam torx bolts to wiggle the cams a bit and slide in the cam lock blocks, it’s damn near impossible to screw up the timing between the cams and crank.

[M5 Jed]

Here’s an update from this afternoon...


So the misfires were occurring when engine was cold and on the lift, right after the first timing attempt. I would start the car run it for 10 minutes and scan for codes. Seeing the misfire codes, I was assuming engine wasn’t timed right, so I never bothered to drive the car, fearing further engine damage due to mis-timing. While running the car for 10 minutes, I’d put my hands on the dash and on top of the valve covers to feel how much the engine was shaking. It definitely felt like it was misfiring. I swapped the coils, still bank 2 misfires occurred. I then swapped the plugs, still bank 2 misfires occurred. So on and so forth.

After re-timing the engine, i started it up again from cold, and the misfires would occur again. After this I did the 2nd compression test and when the new readings came out, that told me that engine is probable timed correctly. I continued to fire up the engine and again, same misfires would occur. This is when I wrote the original post above.

The low fuel warning light came on so I decided to take the car to the gas station and fill it up. It’s about a 10 mile round trip. I cleared all the fault codes again just before I left to get gas. When I headed out, as expected, SES light comes on. Engine seemed to run ok but I didn’t drive it hard, drove it rather gingerly to the gas station. Put gas in and headed back. At every stop light I would put my hands on the dashboard to feel the engine vibration, it seemed very very minor, almost non-existent. I get back and scan for fault codes, same misfire codes on bank 2. I clear the codes, restart the car while it’s still hot from the gas station trip, scan the car again and no more codes come back! WTF???!!!

I restart the car like 5 more times, scanned each time while running, no codes come back. I stare closely at the engine and put my hands on the valve covers to feel for the misfire and it felt normal, smooth. WTF???!!!

At this point, I still don’t trust the car. I’ll let it cool down overnight and fire it back up in the morning to see if any misfire codes come back when engine is cold and report back.

[JimLev]

Not sure at what fuel level the DME is suppose to disregard misfires for low fuel level.
That seems like this was your problem. Guess you didn't check the fuel pressure.
We always keep the TB wide open when checking compression. It will pump up quicker.
You may get a few more PSI with it open if you do it again.

[M5 Jed]

The fuel wasn’t that low. The light had just come on this afternoon, at least a gallon or so left. It wasn’t like running on fumes.

I’ll redo the compression test in the morning with the TB open. Interested in trying this out.

[JimLev]

A gallon isn't much left, still bet that was your problem.
Not all code readers give you the correct reading.
Many years ago I kept getting a crank sensor code. I replaced it with a BMW part from the dealer, still got the code. The car ran fine. It turned out to be the bank 2 cam sensor.

[M5 Jed]

Well, looks like car is ok now. From cold start, no codes came back. Took it for another 10 mile road test and car seems fine, no more funny vibrations or shaking from engine, and no codes came back.

Hard to believe that low fuel could have caused all those codes. Low fuel definitely doesn’t not affect cylinder compression readings. I think if if the Vanos end stop even backs out a hair while you’re torquing it will be out of time. I’ll keep monitoring and report back.

[JimLev]

Thanks for the update. Good to hear it's OK.
You always could run the fuel down past the low warning to see if you get the code back.