In desperate need of help here
 

I am going out of my mind with a misfire. It's a 1995 E320 with 120,000 miles on it. A few weeks ago the car began to run roughly just once when I ran over a few bumps, kind of like a misfire. No CEL illuminated, the problem resolved within three-minutes. A few weeks after that without warning, it happened again for the second time. The same thing happened, a few minutes later it resolved. Now this misfire is permanent. The wiring harness was bought brand new three-years ago. The spark plugs are all Bosch F8DC4. The ignition coils were all proactively replaced along with the rubber plug boots three-years ago. The only code I ever get is #23, "ignition misfire/coil with cylinders 3 & 4". I have replaced the coil over number 4 with three known good ones, I have replaced the plug wire and boot that leads to plug number 3. I just tried replacing both spark plugs for cylinders 3 & 4 with brand new Bosch F8DC4 plugs. It's always the same, the car shakes violently, the exhaust smells awful (obviously raw fuel is escaping), the shaking is worse in R or D. The code is ALWAYS the same, #23. I did an extensive search on this very common problem. Usually by replacing the coil or switching to the correct plugs solves everything. There is no gas in the fuel regulator vacuum line, I disconnected the MOT valve briefly, no change. I cleaned the MAF sensor with CRC-brand MAF sensor cleaner. The rough idle is instant when the engine is started as is the code (#23). WHAT IS WRONG WITH THIS?

The OVP has been replaced (two-years ago) with the MB-brand, the 15 amp fuse is still intact. When swapping the coils, the fault does NOT follow the coil. I just swapped two of them, I still get code #23, cylinders 3 & 4.

If you can, get some new plug wires and try them. I had a problem that could not be detected by sight but plug wire(s) were both arcing at different times causing a rough idle. Or, a plug wire could be loose and you think it is on tight...also happened to me. If I could only see it in person! Did you happen to swap or take off plug wires and reattach wrong ones in wrong order (I have done that as well!). Sounds like something that has a very easy solution (it usually is)!

And you're sure the code is erasing - that is you're getting one flash after it's erased? And you're getting no codes on any other pin? Swap the connector with the coil when you move them around. At some point you have to start suspecting the wiring harness even though it's a replacement.

Yes, code #23 erases successfully every time. The moment I start the engine, it returns. No other codes found on pin 8, just one flash of the light. I hope to hell it's not the harness. That harness has not been bent/moved since I installed it three-years ago.

My bet is bad MAF. On my 124 it took months of misfire troubleshooting before I got a MAF related code, but that was the culprit. It was also around your mileage.

Another question I just thought of, does the rectangular metal "cap" with the six long screws have to be installed every time the engine is started? I'm not doing that since I would have to constantly remove and reinstall it while chasing this problem. I know that cap holds the coils down in place. I've even pushed down on the coils and plug connectors while the engine is running and it changes nothing. The fact that this cap is not attached wouldn't effect this issue would it?

I just tried (with the engine running) unplugging just the low-voltage side coil of cylinder two and the engine stalled immediately (as expected). I restarted the engine and re-connected #2, I unplugged the low-voltage feed to coil #6 and the engine stalled immediately (as expected). I reconnected everything and tried that experiment with cylinder #4 - no change at all in the engine. I'm thinking the coil isn't to blame so much as it is the low-voltage wire that is feeding it. It's not giving it any power at all. Does anyone know what that voltage reading should be on coil feed #4 with the engine running? I even tried (with the engine running) moving the low-voltage feed wires around to cyl. #4 to see if the engine smoothes out - no change. This is so FRUSTRATING.

Another update. I just removed the two brand new Bosch F8DC4 plugs that I installed in cylinders 3 & 4. I installed those yesterday just for the hell of it to see if this problem was the plug. Both new plugs I removed today were completely clean, the tips/electrodes were as clean as they were when they came out of the box. Which means that neither cyls. 3 or 4 are firing. Obviously cylinder 3 gets its command to fire from the coil, #4. I just now reinstalled the same older Bosch F8DC4 plugs back into their cylinders. I also swapped out the OVP with a used known functioning one - no change, still the only code I get is #23...

I would assume you would need to put a scope on it in circuit to see the signal. I don't know that you will read anything with it disconnected if the control line is an open collector type. That just means that the coil itself pulls the line high. I don't know how those work however so I am just guessing. I think you are on the right track though. It might be worth paying a good indy mechanic to check that, if you can find one. Very few are any good with electronics unfortunately.

You could check the coil circuits to the engine control module for continuity and resistance - coil 2 to socket #10, coil 1 to #9, coil 3 to #21.

engine performance circuit 2 of 2 here: http://www.autolib.diakom.ru/CAR/Mercedes-Benz/1995/E320/SYSTEM%20WIRING%20DIAGRAMS/

That's a great idea to test continuity, I can't believe I didn't think of that. I just lifted off the left side of the wiring harness at the brain behind the battery, unfortunately, none of the spades or receptacles are numbered that I can see. To compound it, with one probe in the low voltage wire of coil #2, I got continuity on numerous different sockets of the wiring harness head...

The 2 plugs you believe are not firing should be wet with unburnt fuel. Is that the case?

Yes they were, numbers 3 & 4 were clean but reeked of gasoline.

Do you have a timing light to verify no spark impulse at 3 or 4? Otherwise, you can check for spark visually with the plug out. I would do this before going any farther on the wiring.

The socket numbers showing on the wiring diagram at the link are at the engine control module behind the battery, not the wiring harness plug. They should both be numbered though, on one or both sides, although it may take a flashlight and magnifying glass to find them. Those Mitchell diagrams on the Russian site are great for free but don't have the detail for the wiring harness plug, detail which is available on the factory 124 service CD. Maybe someone will look it up and post it for the wiring harness plug.

Continuity for coil 2 on numerous sockets at the plug is not a good sign.