Hello.

Recently replaced valve covers and spark plugs after buying a used 2007 G35 coupe.

After completing the maintenance, cyl 3 started misfiring when it hadn't before. Seems completely unrelated.

Anyways, I ran through the gamut of diagnostics. Smoke check for intake leaks, verified fuel inj 3 is indeed functioning, and finally came across the no spark issue. Although, through some testing I was able to see that coilpack 3 and spark plug 3 together work just fine. When plugging them into the connector for another cylinder, it sparks just fine (verified outside of the cylinder). This suggests a harness problem. The harness connector for coil 3 properly gets ground, power, and has continuity to the ECM for the "spark signal." However, when it starts misfiring, the "spark signal" drops to 0V, which I am fairly certain causes the misfire... Strangely enough, after resetting the ECM, the misfire does not seem to be present, however after driving a few miles it will come right back... Am I crazy to think this is an ECM issue??? Trying to decide if I should have it checked and/or invest in a replacement.

I know some cars will "shut off" a cylinder after detecting harmful misfires, but I assumed that if this was the case, only the injector would be turned off. Is that a bad assumption? I am wondering if resetting temporarily allows cyl 3 to run, but then it detects a misfire and kills the signals to both fuel & spark, which would only really leave one more check to look at... compression

Please let me know if anyone has any helpful feedback, this strange and elusive misfire has been killing me!

 

Have you swapped coil pack/spark plug and left them in another cylinder, it might work fine cold but when hot it starts to malfunction.

This ECM won't drop spark on an individual cylinder as a failsafe.

 

Yea I've swapped coil/spark with cyl 1 and the problem stayed in 3. Tested both inside and outside of cylinder. Thanks for the info. on the failsafe.

I recently took my ECM out again and sprayed a bunch of Electronics Cleaner on the pins. I've driven like 10-15mi since and the misfire hasn't returned yet (usually it does by now). I am not expecting it to remain fully functional, but potentially promising. There doesn't seem to be any corrosion or anything on the ECM pins, but I thought screw it why not try a $15 cleaner before I try a $500 ECM and so far so good I guess. Again, really don't have a lot of faith that it will last, but I am going to keep testing the car.

 

When looking for corrosion make sure to look at the BACK of the harness also, not just the front of the harness. You might have a clogged sunroof drain that's dripping water on the F108 main connector, or dripping on the wire which then runs down to the connector. You can also pop the back of the connector off to get a better look at it.

 

Interesting.

I already established continuity from F108 to the coil pack connector so I assume the connector itself is okay. You are suggesting removing the ECM connector and checking both sides right? When you say "popping the back off," is that like in addition to removing it from the ECM?

The place where I sprayed the cleaner was upwards onto the ECM pins themselves (not in the connector). I figured if there is continuity from F108 connector to coilpack, doesn't necessarily mean there is continuity from ECM to F108 (although I assume continuity issues there to be extremely rare given how shielded it is).

 

F108 is the ECM connector with the lever. If you have continuity from that harness to the coil pack then it's looking like a bad ECM at this point unfortunately.

Did you take that F108 ECM harness off? If not I would disconnect the battery, pump the brakes to drain residual charge, then take that harness off and inspect those pins thoroughly, including inside the ECM. It's easier to remove the ECM to get a good look at it, just two 10mm nuts holding it to the chassis.

If you get a replacement ECM just make sure the MEC number matches EXACTLY. Schedule ahead with Nissan to perform the ECM/BCM/key pairing, tell them you're going to swap the ECM in the parking lot, they'll usually give you an exact place to park that works best for the technicians (i.e. in the shade so it's not a billion degrees in the car while they're programming it).

Doing a compression test first is definitely a good idea just to make sure you're not dealing with a dead hole, I sort of doubt it though since the ECM reset magically makes it work again for a while.

There's also an ignition condenser that might be failing but I've never personally seen it cause anything like what you're experiencing. Usually it's just spotty coil voltage and burning up coil packs regularly when those things go out.

Have you THOROUGHLY wiggled all the wire harness just to rule out a broken wire that's intermittently making contact?

 

Alternately you could just solder / splice in a new GPT wire from F108 to that coil pack to verify it's not an intermittently bad wire.

 

Did you take that F108 ECM harness off?
Yes quite a few times when checking continuity and when I sprayed cleaner on the ECM pins.

Take that harness off and inspect those pins thoroughly, including inside ECM.
I haven't fully removed the ECM just yet but it is pretty darn clean in there let me say.

There's also an ignition condenser that might be failing but I've never personally seen it cause anything like what you're experiencing.
If there is an issue with the condenser the car would likely not start. Also it would be pretty random for it to only affect cyl 3.

Have you THOROUGHLY wiggled all the wire harness just to rule out a broken wire that's intermittently making contact?
I mean, not super thoroughly, but I have checked continuity a few times each after misfires were present and found it to be good each time without any hesitation from the multi-meter. It would be pretty darn unlikely for it to start intermittently contacting only as I am checking it if that makes sense.

 

Yes I have thought about soldering a new wire but it's a bit of a pain. I have doubts that the wire is intermittently going bad, because once the engine misses, it stays missing. And once the ECM is reset it stays good until it misses again, which causes it to stay missing.

 

I thibk the plan is to drive it until the code shows up again (if it does), then remove the ECM, send it to be tested somewhere just to verify whether it may have issues or not and go from there. If it is supposedly ok, then I will dive into more extreme harness testing (even though I think the harness is good). If it is bad, then new ECM final solution and be done hopefully.

Thanks for all your comments cleric670, you've been very helpful!

 

Once it starts the misfire, can you just park it and try the next morning to see if it STILL misfires? Might be something on the ECM getting hot and malfunctioning.

Definitely looking like it's probably an ECM issue.

 

I can try that yes.

So far I have had an experience similar to what you are describing.

Typically on first trip, I get the flashing SES and then it seems to go away and come back with increasing rpms until I hit the second trip which keeps the SES solid.

After the light is fully on, letting the car cool down has not seemed to remove the issue. When the light was fully on last week, I left my car parked in a shaded garage for a few hours and came back to see the light still on with the car having a rough idle. The next day I verified the misfire and the missing signal to the coil pack, which seems to resolve every time I reset the ecm

 

Misfire started again today. Been about ~30mi no issues, but it came back immediately on starting my car at noon. It is very hot right now so heat is a potential factor. Going to wait til it cools down in the evening and start her up again to compare.

Do you know any reputable companies that specialize in ecm diagnostics/repair @cleric670 ? I would prefer to have it looked at since this seems like a random bug more so than detrimental ecm issues and buying a new one straight away will not be cheap.

 

Just wanted to update that it seems like heat is making a difference.

After the misfire came back I waited til the evening when it was cool again and started the car up again. I cleared the previous dtc and it didn't come back. (and cylinder 3 was firing okay)

So looks like something with heat is causing the ecm to malfunction.... but only on the output signal to coilpack 3 which is very strange

 

Another update. (I really want to verify everything possible before trading in an ECM)

Was curious to see if bad grounds were playing any role in the issue. Battery ground terminal had some corrosion which I cleaned off to the best of my ability.
When checking voltage between battery and frame it gets identical voltage to battery:

• 12.6V (when off)
• 14.3V (when on)

When checking voltage between battery and engine it had:

• 12.6V (when off)
• 14.4V (when on)

these values are not rounded, they are exact and were consistent. it was strange to me that the engine continually had a higher voltage differential. My understanding is that usually bad grounds displayed themselves in decreased voltages, but otherwise you would only see up to battery voltage, not usually above. Not saying this means anything but it was interesting. Since the main grounds seem pretty good I don't think it is worthwhile to check all individual ground connections, but if someone disagrees please advise me why they could still be an issue.

The big thing I discovered is that the coilpack harness connector when "not" misfiring is getting a weak signal that fluctuates somewhere between 0.05V - 0.17V at idle where other cylinders are getting the steady 0.2V. When it starts misfiring badly and is detected by the computer the signal drops to 0V for cylinder 3. Just want to be double/triple sure, does this in any way shape or form sound like bad connections/ground? (Reminder: This connector has proper power, proper ground, and proper continuity on the signal wire, so it seems as those everything connectivity wise is in check, and the ecm is just strangely dishing out a bad signal for some unknown reason).