So as a summation:
There’s a misfire only while idle. As the temperature of the engine increases the number of misfires decreases.
Replacing the coil packs, voltage transformer, O2 Sensors, Cats, intercooler MAP, air intake MAP, vacuum hoses, throttle Y-pipe, and a host of other components I’d consider indirectly related to this problem have also been replaced.

I have a guess, it’s worth a shot, and it’s cheap: Whenever you clean (and especially replace!) any throttle body, you need to make sure to reset the adaptations.
That tiny little buildup of crud actually makes quite a difference to a car’s idling performance.

Did you do this?

[edit]

Same for O2 sensors

 

Just to make sure all of us following this are on the same page..

1. Your misfire count goes up when the engine is not up to temperature and the engine is actually misfiring, e.g. lumpy.

2. Your misfire count goes up when the engine is not up to temperature but is not actually misfiring.

Which is true?

 

Also you never mentioned your air filters. Are they by any chance dirty?

 

I think you need to probe specific cylinders. I was a bit curious about there being an issue on random cylinders across multiple banks and it seems there is a correlation. The cylinder firing order is:

..... 4-9-1-12-5....

A significant issue with the number 4 firing can cause the misfire count to propagate to some of the subsequent firings and cause the issues on cylinders 9 and 1 (and less so on 12 and 5). In fact if you look at your misfire counter the misfires are in order of their firing order starting with 4 highest, 9 second.. etc.

You should take a closer look at cylinder 4. I wonder if compression is low when the engine is cold.

 

It's #2. The engine's not actually lumpy when the misfire count goes up. I've compared it against the fuel injector test in SDS when you shut off one injector (gets lumpy). Done it to all 12, and they seem to be working normally.

 

And I was thinking the same thing about cylinder #4 and the compression test. I just ran out of time to do it last night before it got dark, but that's next up.

 

If it's #2 then I would say looking for a mechanical issue is not going to help. If the engine is running sweetly but the misfire count is shooting up then whatever is doing the count is causing an issue. Are you sure camshaft and crankshaft sensors are fine?

I'm not sure entirely how misfires are detected by the ECU but I would have thought some combination of ionic charge measurements from the coils and the signal from the crankshaft position sensor would be the cause. Are you absolutely sure your coil is fully ok?

 

I was thinking about this thread and I came up with an idea I don’t want to forget:

Fuel Filter. The car has a variable pressure fuel pump. I don’t know where it regulates that pressure, but if it’s before the fuel filter, a clogged fuel filter could starve the engine under idle condition when I would expect the variable fuel pump to be at its lowest pressure and could be causing fuel starvation.

 

If it was a case of fuel starvation then the engine would actually misfire and it would feel lumpy.

I had a misfire when one of my coils failed. The counter would get to about 45-50 when idling before resetting to 0. It was very obvious that the misfire was present.

If your misfire count gets to 100+ at idle it would be very obvious and very lumpy. If you aren't experiencing that then it cannot be a mechanical issue with the cylinder actually misfiring.

I'm going to suggest an electronic issue with related to misfire detection.

 

The coils were already replaced. It doesn’t have to be full on fuel starvation, simply partial low fuel pressure. It would also explain the temperature dependence of the misfires.

 

But he has said the engine isn't misfiring. It is only saying that it's misfiring but the engine is actually running smoothly.

If there was a problem with fuel starvation the engine would be actually misfiring and running rough.

 

It could be a partial misfire that wouldn’t be felt but warrant a misfire reading. Time will tell; we have to wait for OP to get back to us.

 

OK folks, I finally got some time to do the compression test. Here are the values, in PSI. The short version: it appears that cylinder #12 is just slightly lower than the others.

1. 195
2. 194
3. 190
4. 194
5. 190
6. 192
7. 198
8. 194
9. 205
10. 200
11. 195
12. 180

Cylinder #12 is at 180 PSI. I did check the torque of the other spark plug, and I did the check in both spark plug wells (yep, with a plug in the "other" well).

I also found that in Cylinder #4, there was some brown gunk in one of the spark plug wells. This gunk had built up enough to prevent getting the socket around the plug. Once I gently and patienty got it out of there, the spark plug came out. However, it sure felt loose. I thought I torqued it down to specs two and a half years ago...but if either I hadn't torqued it down right, or it had come loose over time, I could see some gunk getting through there over time and messing things up. That is the cylinder that was having the biggest misfire counts.

After checking the compression, I pulled all spark plugs and had a look at 'em. Noticed that some of the spark plugs had what looked like some light oil staining on the bottom of the white insulator section of the plug...and a few others did not. There was a bit of still-gooey gunk on about 2/3 of the spark plugs, which came off easily enough with some kerosene and a wire brush, followed by some brake parts cleaner. Furthermore, on some plugs, there was a bit of dried-up gunk on both the side electrode and the iridium center electrode, so I lightly scraped that off. They're pretty clean now. All 24 plugs are back in and torqued to 25 N-m; I must've checked each one of them three times after installation.

So, do these cylinder pressures look right? And if so, then is the slightly low pressure in cylinder 12 anything to worry about?

 

Also, I replaced the fuel filter a couple of years ago, to address that (I'd say appropriate) question.

 

I would say as long as they are within 10% of each other it's fine, and they are.

What are the red insulator boots like?

 

Red insulator boots are also looking good; they came new with the ignition coil pack assemblies a few months ago (they are the Beru OEM ones, too--Clark was thorough). Just to be sure, I pulled 'em out and inspected each one, since the ignition coils were already out.

I've put everything back together except for the pre-throttle-body MAP sensor, since I accidentally cracked the nipple off the MAP, so now I've got to order a new one. Ugh...well, I needed a couple of other parts anyway for one of the other cars, so might as well do 'em all at once.

 

Realized I hadn't addressed these two questions previously.

Camshaft sensors are 2.5 or so years old. Crankshaft sensor is two months old, replaced pre-emptively as part of all this work (car has 175K on it, so seemed to make sense). Camshaft sensors are Genuine-MB, and the crankshaft sensor is OEM Bosch.

Well, the coils are from Clark over at v12icpack.com. I was getting misfire count increases with the original (Genuine-MB) coils as well, so now I'm suspecting that the original coils are actually OK!

Soon as the replacement MAP sensor arrives, I'll fire up the car and see if this last weekend's work made any difference.

 

Quick update: new MAP sensor is installed. Got the computer on the car, then started it up. Looks like cleaning the plugs indeed helped some. The misfire count is really only going up now on Cylinder #4. There's a very slight bit on 9-12, but it's slight; nothing like it had been previously.

Even with Cylinder #4, the misfire count slows down once we get past about 45 Celsius, slows down a lot more by 60 deg C. (almost to nothing), and no counts going up whatsoever past that temperature.

Pics to follow.

 

Have you resolved this problem yet? My 65 has similar cold start misfire issues on both banks. Restarting a few times before driving will clear them. It NEVER misfires at WOT.
Did you happen to upgrade the throttle body gasket with the larger diameter one? The original gasket is a few MM smaller than the TB which occludes free airflow.
I upgraded to this version to improve WOT performance along with some other work at the same time and have had this issue since.
Is it possible that smaller gasket with the protruding lip creates turbulence with an air/fuel mixing effect which promotes better combustion until the engine warms up.
In a DAS troubleshooting session for cold start Misfire of cylinder 4 and 10 it asks you to verify that the correct part# gasket is installed at throttle body.
Another thing to check out if you have not already. Some day I will swap original back to see if it makes any difference. Good luck.

 

Good to hear back.

Is it possible your valve cover gasket is leaking? This causes all sorts of issues including the resulting failure of the coil packs when the oil by capillary action crawls up the coils into the electronics.

How long since you replaced the spark plugs? With a leaking valve cover gasket, the oil could creep between the spark plugs and the engine along the threads. This would explain why the spark plug might have been under-torqued and/or came loose. It would also explain the oil on the plugs themselves. This should be really easy to inspect visually though and should immediately stand out.

Regarding the oil leaks,
I was able to fix my leaking valve cover gasket with 2-cans of liqui-moly oil stop leak, as the gaskets had recently been replaced.

 

You just described my misfiring symptom once larger TB gasket installed. I have misfire at cold start then it will go away when car up to normal temp....im thinking the gasket is the issue.

 

TB Gasket?

 

Throttle body gasket.

This is interesting because I am seeing something very similar. Although mine aren't bad enough to trigger a hard misfire code my fault counter goes up to about 5-6 on one or two cylinders on cold start idle.

I have the original gasket on order from the dealer and plan on putting it back in.

 

throttle body.

 

I would like to add this find I made recently
as I am on a M275 mega build, I got spare Voltage, I have 2 engine ecus adapted so i can test and swap ect i'm doing all kind of test these weeks

My latest find is that you can also get startup misfire if you have a weak battery
thing to try out is leave your battery charging overnight. then in the morning try a startup and monitor in Xentry fault counter if anything changed
if not, you can eliminate this