Jeep Garage  - Jeep Forum banner

'04 WJ, bizarre misfire of sorts

7K views 39 replies 4 participants last post by  Lhutter1 
#1 · (Edited)
Hey guys!

Our 2004 GC Limited has a strange misfire issue of sorts... When I say misfire, it sounds and feels like a misfire, however, there's "no" check engine light, no codes. Exhaust appears louder as I'm driving along and there's a distinct one cylinder that stands out as different. You can feel it as well, especially as RPMs hit above 1500 or so, all gears. With my head under the hood it doesn't sound loud at all. It happens regularly when the ambient is below 67 degrees F, only as the engine gets up around operating temp, it's like someone hits a switch and it happens. Cold engine runs great! If the ambient is "above" 68 degrees or so, it runs excellent cold or hot engine.

I've seen many posts about head gaskets and fuel injectors and such, but those all seem like more permanent issues. I took it to a friend's shop and we put a scope on it. No misfires recorded, but it was obviously missing and the engine stumbling and jumping around. The only thing noted out of spec was the IAT sensor was reading 108 degrees "F". I replaced the sensor and it reads ok, but no change in operation. I also changed the TPS. I've also changed plugs. I thought also it may be coils or fuels injectors as they heat up, so I used freeze spry on each with no difference. Besides, that wouldn't explain why it runs good in warmer air... Anyone else seen this?

Thanks!
Lorne
 
See less See more
#2 ·
Temperature related issues most of the times have to do with a sensor problem.
Maybe a crank or camshaft sensor, which is messing with the injection or ignition timing. Just an idea of me. Probably 2005JGC can help you further without changing all kind of components in the process of troubleshooting.:eek:
 
#3 ·
What TPS did you use? I have read that only OEM Mopar or you will get all kinds of strange behaviors. Which engine? I assume it is a 4.7, not the HO. Is that right?
 
#5 ·
Update: OK, we hooked the WJ up to a diagnostic scope and near as we can tell, the crank and cam sensors are working. Notably, however, we realized that the IAT sensor is still out to lunch (we think). With the sensor connected, the PCM shows the ambient is 108 degrees F steady state, with it disconnected, the PCM reads 84 degrees F (as did the old IAT). Is 84 degrees a default if the IAT shows an open? Anyone else seen this?

Thanks!
Lorne
 
#6 ·
Yep, the 84 F is a default setting when the temp sensor info is lost.
Did you already measure the IAT wiring to ground? From PCM connector C1 point 15 to ground and from the same connector point 4 to ground. The resistance should be infinite, otherwise you could have wire shaving somewhere.
 
#7 ·
The resistance of the IAT sensor at 45 degrees F was around 21k ohms. As I blew warmer air on it, the resistance went down. I have not disconnected the main trunk into the PCM (will later today). I did measure the resistance at the IAT connector with the PCM hooked up. The BLK/RED wire (IAT return) is not infinite (but in the m ohm range) and the BLK/BLU wire is dead short to ground. Question.... Is the sensor ground out of the PCM (pin 4 BLK/BLU) and chassis ground supposed to be the same potential?

Thanks,
Lorne
 
#8 ·
Not completely sure about this, but they talk about a low noise sensor return, so it should not be directly shorted to ground. It can be that inside the PCM it will switch to ground. For that reason you need to remove the C1 connector, so that you can measure the wiring only.
 
#9 ·
Well, at this point I'm thinking either harness at the PCM or the PCM itself. It seems I've heard mention of problems with both. When I disconnect the harness from the PCM all is well, everything looks OK in a resistance test of the IAT harness. When I hook it back up, it still looks OK, no short to ground. When I drive it and it gets to operating temp, it fails again and I have the short back. Soooo, I'm thinking about getting it up to temp on a cool day (when it fails) and hitting the PCM with a hair drier or heat gun for a bit and see if it changes. Not sure what else to do here... Thoughts? :confused:
 
#10 ·
how are you measureing the continuity to ground?

you would be supprised what leaving connectors connected can do to trick you into believing a false reading. if you want to measure continuity to ground with the circuit intact you need to measure voltage drop on the circuit with the circuit operational... never trust an ohm meter on a completed circuit.

I was taught a quick check of a simple circuit like this is to unplug it which should give you an open or "voltage high" code, then short it to ground which should give it a voltage low or shorted to ground code. If you get both codes you know your circuit is A) intact and B) able to see when it is wrong so one could assume fully operational... if you see both codes it needs an IAT sensor.
 
#11 ·
I guess my question then would be, is the chassis ground and sensor signal ground (pin 4 on PCM) supposed to be the same potential or should there be a difference? I'm under the impression that they should be different, with the sensor ground being somewhat isolated from the chassis ground. Please correct me if I'm wrong though, I hate chasing up the wrong tree. :) Thanks!
 
#12 ·
Sorry for the slow responses. Been a lot going on making it difficult to get into the garage to do any testing. I appreciate everyone's input...

OK, the pin to pin voltage at the IAT sensor with the sensor unplugged is 5.3 volts, same result with the blk/red wire on the connector to chassis (alternator case). The voltage from chassis ground to sensor ground (blk/blu wire on the connector) is 0 volts. I'll have to borrow my friend's auto scope and see what actually happens when I short the connector out and maybe even try a pot in the circuit to see if there's a difference. Thanks!
 
#13 ·
Can anyone tell me what the resistance range should be of the IAT NTC? I remember checking it out of the vehicle a while back at around 45 degrees F ambient and the resistance was around 21k. I'm planning on replacing it with a pot to see if it makes a difference. Does that value seem within the range? If so I'll try a 50k pot...

Thanks!
L.
 
#14 ·
I put a 50k pot in place of the IAT sensor and did maybe notice a very slight difference. Seemed to be a bigger difference in idle stability. Not knowing what the resistance should be at 30 degrees made it a challenge and I really don't know if I was actually within range. I was able to see the voltage go up and down with the turning of the pot. At short circuit I do get a fault, as was pointed out I should. This evening I'm taking it to a friend's garage for a few other things and am going to hook it up to his scope to see what the PCM thinks while I'm there.

Thanks,
Lorne
 
#15 ·
Well, I never made it to my friend's shop over the weekend, however, I may have stumbled on to something else... As a side thought, I figured maybe a secondary on a coil was failing, so I picked up a coil and started playing leap frog with the coils, figuring maybe I would see different result if it was one of the coils. The problem never changed (so I thought). Later that evening we took it for a drive and I noticed that, though the loudish exhaust and vibration were still there, it was diminished by a considerable amount. Same thing the past couple of days. I'm wondering if I just moved the problem to a different cylinder or could I have had 2 weak coils and one is still weak? The IAT problem hasn't gone away, maybe a combinational problem? I think I need to make a coil tester to verify. Any thoughts?

Thanks!
Lorne
 
#16 ·
Hey guys!

Update: Ok, I made a harness and hardwired to the IAT sensor eliminating that connector. I used twisted pair wire to cut down on potential noise issues. I then cut the blk/red and blk/blu wires leading into the PCM about 2" back and spliced the new harness into it. Same results. It looks like the PCM is providing the voltage, but not responding to the changes in resistance. Anyone else thinking I need a new PCM? Other ideas?

Thanks!
Lorne
 
#19 ·
When i don´t know, i prefer to keep silent then saying something that doesn´t help.
But if you insist: yes, buy a new PCM, it could be the problem. And if not, you just try something else.......:D
 
#20 ·
Thanks Frank! I wasn't sure the posts were getting through. :) Thing's driving me batty! lol I do find though, that the colder it is out, the worse it runs, especially cold start now. Runs like the choke's not on enough and you have to play with the gas pedal a bit to accelerate til it warms up a bit. Never seems right though. Not sure if I have two problems at hand or just one that's being a PITA! :D I'll order a PCM Friday and see what happens!

Thanks again,
Lorne
 
#21 · (Edited)
OK, so I ended up borrowing my friend's scope/diagnostic tool to use while driving and learned a great many things. I don't think I need a PCM, here's why... The odd readings I've been getting on the IAT sensor are bogus numbers, actually from poor placement of the sensor. At cold start-up (which I had never tried the scope at) the sensor is within a couple degrees of ambient. As I would drive along, the temp would rise a bit. Sitting at a traffic light, the temperature where the sensor is located ran up to around 94*F. After my 20 minute drive on the highway it managed to get down to 20* above ambient, which was 52*F It never got all the way back down to ambient. If I let it sit and idle for any period, the temp would hover around 108*F (hence my high reading), even after shutting it off for a few minutes had the same results. So that knocks that out of running for real problems (other than design flaws lol).

I'm starting to wonder if it might be a flaky fuel injector? With a stethoscope of sorts, each one is clicking away just fine. Wondering though if one might be dripping or squirting instead of spraying nice getting worse as it heats up? I've run Seafoam through a couple times. I don't think it would be totally dead, because if I disconnect any of them one at a time it makes it run worse, same with the coils. Anyone got ideas on how to test that portion of the injectors? Other ideas or directions to check?

Thanks!
Lorne
 
#22 · (Edited)
It could be a clogged injector that has a poor spray pattern. I would suggest some BG 44K in the tank to clean the injectors. Speaking of BG 44K, I should put some in the WJ...

That being said... I would expect poor spray pattern to create a lean condition and the PCM to add fuel. Any idea how old the upstream O2 sensors are? Ever replaced the TPS? Both of those are problem areas on WJ's with the 4.7. I replaced both and noticed a considerable difference with mine.
 
#23 ·
Hmmm... Never heard of BG 44K, but I'll look into it today. I changed the TPS a while back. Didn't change anything and looking at the scope it followed well as I played with the pedal, so I'm thinking it's probably ok. Hadn't thought about O2 sensors much. They were working and looked ok through my eyes on the scope, but in reality they may not be sending accurate data.

Thanks!
Lorne
 
#25 ·
Hmmm... Never heard of BG 44K, but I'll look into it today. I changed the TPS a while back. Didn't change anything and looking at the scope it followed well as I played with the pedal, so I'm thinking it's probably ok. Hadn't thought about O2 sensors much. They were working and looked ok through my eyes on the scope, but in reality they may not be sending accurate data.

Thanks!
Lorne
The before cat O2 sensors should give a quite unstable voltage reading (with engine on), the more going up and down, the better the sensor is. If it shows a steady voltage or small differences, its time to change it.
 
#24 ·
BG 44k = BG Company

Recommended by Nissan/Infiniti and BMW for fuel systems.
 
#26 ·
The readings weren't steady, they jumped around a bit, but were tracking pretty close to each other. There were two sets of numbers per O2 sensor. The first number on each would jump all around, which, from reading looks like they're supposed to, and the second number was a little more stable , but changing some. My original thinking was that since they're tracking reasonably close together, they should be ok. Please correct me if I not thinking that right.

After talking to my friend with the shop, he suggested an induction cleaning, which includes a bottle of BG 44K in the tank. :D He claims it's like giving the engine really good an enema. Something we all could use now and then! lol He claims that if O2 sensors are just dirty it will clean them as well, so I figure it's worth a shot.

Thanks!
Lorne
 
#27 ·
The readings weren't steady, they jumped around a bit, but were tracking pretty close to each other. There were two sets of numbers per O2 sensor. The first number on each would jump all around, which, from reading looks like they're supposed to, and the second number was a little more stable , but changing some. My original thinking was that since they're tracking reasonably close together, they should be ok. Please correct me if I not thinking that right.
2 Sets of numbers per sensor? Was one set not for the upstream and the other one for the downstream sensor? The downstream sensors have normally more stable readings
 
#28 ·
That's interesting and I hadn't thought of it. In the analyzer they're just labeled O2 sensor 1/1, 1/2, 2/1, 2/2. Hadn't considered 2 downstream as well. I'm taking it in to my friend's for the induction cleaning later today and I'll mention that to him and get his thoughts on it as well. I did put in BG 44k last evening. Yesterday we also did an injector balance test and there was a notable difference in shutting down each injector. If none of the above works, next step may be pull the fuel rails and activate each injector to check spray pattern. Running out of stuff to try...

Thanks,
Lorne
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top