Diagnosing severe cylinder 6 misfire

[Retlaw01XJ]

2001 XJ 4.6 stroker with mild Crower 44243 cam and stock valve springs, approx 11,000 miles on rebuild. Daily driver, about 3.3 miles each way, I drive it fairly easy most of the time.

A severe misfire started last week, almost constant but cleared up a few times and ran fine for a few miles.
Started after I added 5 gal of gas from a gas can and left the gas cap off. Drove about 1/2 mile and it missed and stumbled it was barely drivable. It's possible some water got in the open gas neck from melting snow on the roof.
Fault codes showed cyl 6 misfire, cyl 5 misfire, and random misfire.
In addition to the fault codes, I was pulling the injector plug from each cylinder.... all but #6 made the rpm drop..... that's my primary indication the problem is #6. May have had problems with #5 but not sure.

Done so far:
Checked compression. Average about 175 but #6 was 190.
Checked plugs, looked fine. Replaced them with a used set after they got fouled after running with the stumble while testing. no change.
Replaced coil with a Viper style (actually a Caravan V6) and Accel spiral plug wires.
Checked for spark with timing light on plug wires- all firing.
Swapped injectors (Neon) 1-6,2-5,3-4, no change.
Checked fuel pressure with an old air conditioning gauge (all I had!), about 50 psi. (will try again with new gauge)
Vacuum gauge gives fairly steady 18 in of vacuum.
Swapped the bank 2 #1 O2 sensor with a used one.
At some point it started running on all 6 cylinders. Took it for a test drive and it started missing after 1/2 mile.
Poured a trickle of water in the throttle body (maybe 6 oz) at about 2500 rpm..... old trick to loosen carbon deposits.
Disconnected the battery to reset the computer. Ran good. Went for a test beat of a few miles. ran fine.
Drove the 3.3 miles to work today and was fine. Started missing on the way home again.

None of the above permanently fixed the problem. I'm thinking it may be the signal to #6 injector, any other ideas?

My order from Amazon came today.... got a noid light to test injector signal, fuel pressure gauge, and spark tester.

Question.... will the pcm kill a cylinder (turn off injector) if it detects a problem?

I'll report back after checking with the noid light Tuesday.
Anyone have any other ideas???
Thanks...

[Retlaw01XJ]

Update....

Put the noid light on the cylinder 6 injector plug and it was lighting for about a minute then went off.
Shut engine off, and cleared the P0306 (cyl 6 misfire) code off of the pcm with my scan tool. Also had a P0206 code (open #6 injector circuit) which I figure was due to unplugging the injector and plugging in the noid light. First time I've seen that code.

Now.... does the noid light going off definetely ID the problem as being the injector signal?.... or does the PCM turn the injector off after it detects a misfire?

Next, I suppose I trace the injector wires back the to plug at the PCM. Thinking I can put the noid light in parallel with the injector wiring at the PCM plug and see if the light still goes out. Or run a new wire to the injector....

Any other advise?

[beater_renix]

bump...

I'll offer my two cents, but I don't have the same pcm as you. In my (limited) experience things that are changing with time, and in this case the engine warming up, have to do with heat, and thermal expansion. Some ideas thrown about are:

Check the pcm to coil/ coil to wire/ wire to spark plug as well as the injector. I don't know how the PCM works it may be able to sense if the coil fired or not. I'd bet it would kill fuel if it detected a spark related miss. (Someone can chime in and tell me how far off I am.)
a break in a wire, but not it's insulation. Those can be very hard to detect.
Heat soak at the coil, but since your not Coil on Plug anymore, that doesn't sound as likely.

I've having a similar issue on a different vehicle, and I think it's the ECM in my case. On of the transistors in there is transistoring and it's throwing a code. i hope you find it, good luck.

[Retlaw01XJ]

..... I don't know how the PCM works it may be able to sense if the coil fired or not. I'd bet it would kill fuel if it detected a spark related miss. (Someone can chime in and tell me how far off I am.)
....a break in a wire, but not it's insulation. Those can be very hard to detect.

Thanks for the reply.
My most significant find was the noid light showing #6 wasn't firing..... So the big question is if the PCM will turn it off in instances of misfire..... anyone else know??

It's been running okay for the past 1-1/2 weeks, ever since I started unwrapping the tape from the wiring harness near the #6 injector. There was a 180 degree fold in the wire I didn't like, so I straightened it and fussed with the wire a bit, and it's been running pretty good since , but sometimes it feels a bit 'off' during part throttle.... could be the old plugs I put in for testing or the old O2 sensor, I'll have to swap them back out. But no stumbling/misfiring like I initially had.
Also put a tank of premium in it and have been beating on it a bit.... trying to clear out carbon deposits! I did read online about issues with carbon buildup on those back cylinders causing the exhaust valve to not seat properly. I still had good compression numbers, so who knows? I love the 3.3 mile commute to work, but it's hard on the engine.

Weather here is Pa has been horrible, winter just won't go away!

[akadeutsch]

I could be wrong here but I am pretty sure that the PCM will kill any suspect injector as soon as it detects a misfire.

[Retlaw01XJ]

I could be wrong here but I am pretty sure that the PCM will kill any suspect injector as soon as it detects a misfire.

Sheesh..... that makes diagnosing the source of the problem much more difficult!
The misfire has been gone since I fiddled with the wire going to the #6 injector. I'm not convinced that was the problem, but I don't even want to touch the wires since it's running good now!