Rough Idle And P0340 / P1391

[Akula69]

Ok folks, got a real head scratcher here concerning idle and ignition.

 

My son's 89 MJ now has a 99 XJ driveline and axles. The only thing that is 89 is the sheet metal, so we are basically looking at a OBDII system. The problem is the engine idle becomes erratic at random times - there is no rhyme or reason to when it happens: at cold start, after running for a couple of minutes, after running for 1/2 an hour, after 5 times around the block - you get the idea. It happens when both at idle or in gear...driving or not. The engine always starts immediately upon turning the key, and when it runs correctly it is dead nuts on 700 RPM and smooth as silk. When it starts the crap, it hesitates several times, then the MIL lights with the two codes, and the idle varies between 350 and 1000 RPM, jumping up and down irregularly.

 

The codes are P0340 - no cam sensor signal and P1391 - erratic cam sensor signal.

 

We have exhaustively worked this problem (for about 4 weeks), and it is truly spoiling his conversion experience. What compounds the issue is the engine ran without this problem in the donor before being rebuilt, so whatever it is, we did it. So far we have:

 

Checked all sensors.

Replaced the cam sensor x 2, with NAPA (Echlin).

Replaced the crank sensor x 2, with NAPA (Wells).

Completely removed the front of the engine and confirmed the timing chain is correct and sprocket marks are in time.

Replaced the spark plugs x 2 with champion plugs gapped at .035.

Replaced the rotor and cap (Mopar).

Replaced the plug wires (Mopar).

Replaced the Distributor (NAPA).

Removed he valve cover and watched while turning harmonic balancer by hand to ensure the valves are in time with TDC (as piston # 1 is coming up and the rotor is approaching 5:00 O'clock the intake valve is closing; then intake valve closes and as rotor passes 5:00 O'clock the exhaust valve begins to open).

Replaced both O2 sensors.

Replaced the IAC.

Replaced the TPS.

Ohmed out the wires from the sensors to the PCM.

Pulled the wiring harness and removed all the loom, checked the wires for melted or torn insulation. Did find two of the cam wires had damage, soldered and shrink - tubed the repairs - re-ohmed wires.

Replaced PCM with rebuilt one (but removed and sent rebuilt one back when it did not solve problem).

Checked the running voltages at both the cam and the CPS - dead nuts 5V.

"Cleared" the PCM capacitors by touching the (disconnected) battery cables together for 30 seconds.

Used smoke machine to check for vacuum leaks and found none.

Hooked up oscilloscope to check square wave pattern for faults on both the cam and the crank sensors. When checked, they are perfectly in sync while the engine is running fine but when the engine starts it's crap they go out of sync. (for those who wonder, the in-sync distributor is 2 degrees out at idle...when the behavior I describe starts the distributor goes instantly to 30 degrees out).

 

I have read that the distributor hold-down fork should not be modified in any way as the ignition timing is controlled by the PCM,  but when taken to a local garage (that I trust), the owner states he has heard that modifying the fork (to be able to turn the distributor) has resolved issues such as this....but before I do this I'd like to get input from you guys about the problem. 

 

Any and all ideas are welcome....because we are out of options.

 

Thanks

[HOrnbrod]

A weak battery can cause the P0340 code - how old is the MJ battery or did you use the 99XJ battery? I'd have it checked anyhow, not just the crappy load test that Autozone does, but using the battery analysis machine most dealers have. My daughter's Suzuki was throwing this code; turned out to be the battery. Just a thought...........

[RyanL]

It is a brand spanking new battery bought one month ago.

A weak battery can cause the P0340 code - how old is the MJ battery or did you use the 99XJ battery? I'd have it checked anyhow, not just the crappy load test that Autozone does, but using the battery analysis machine most dealers have. My daughter's Suzuki was throwing this code; turned out to be the battery. Just a thought...........

[Akula69]

Thanks Don - unfortunately I left that one out....we replaced the battery with a brand new one.

 

Also left out that we had extensively checked the fuel pump and regulator. The gauge reads rock steady at 47 PSI both parked and moving. The injectors are Mopar, but rebuilt with new O rings and strainers.

 

After the first post we took it for a ride and put about 5 miles on it before it threw the two codes again. Funny thing is, even though the idle is rough the throttle response is smooth all through the range.

[HOrnbrod]

How did you ckeck the crank and cam position signals? Use a dual trace o'scope and look at both square waves simultaneously to see if they are both synced up? Correct me if I'm wrong, but the PCM looks for the crank sensor for primary timing, then the cam signal only if the crank signal fails. Can you tell on the scope which signal goes wacky when it happens? Might provide a clue as to where to look.

 

Also, does the 99 engine still have the timing marks on the timing cover and harmonic balancer? If so, it might not be a bad idea to throw a timing light on it and eyeball the timing marks. Mine is usually at around 14* BTDC, and is steady w/o jumping around while changing RPMs.

 

[Akula69]

Don, we used a dual trace scope and did exactly what you described. The square wave from the CPS starts slightly at the end of the CAM's wave when they are in sync. I am unsure but tend to agree with you on the CPS being primary.

 

The square waves are normal until it happens, then the CAM throws an artifact (it is very small spike wave), then after several minutes the CPS throws an artifact, which is a very similar (in shape) spike. The occurrences are totally at random, no discernible pattern. Then, both of them just chill out for random periods of time, then it happens again.

 

the 99 still has the timing marks. Never thought of the timing light - haven't owned one in years....but on the way to get one now.

 

 

How did you ckeck the crank and cam position signals? Use a dual trace o'scope and look at both square waves simultaneously to see if they are both synced up? Correct me if I'm wrong, but the PCM looks for the crank sensor for primary timing, then the cam signal only if the crank signal fails. Can you tell on the scope which signal goes wacky when it happens? Might provide a clue as to where to look.

 

Also, does the 99 engine still have the timing marks on the timing cover and harmonic balancer? If so, it might not be a bad idea to throw a timing light on it and eyeball the timing marks. Mine is usually at around 14* BTDC, and is steady w/o jumping around while changing RPMs.

 

[Akula69]

Forgot to add - we attached the o'scope at both the plug side of the harness and the PCM side of the harness, and the signals were the same...

[RyanL]

We used a timing light on it today and this is what we got.

 

 

The white mark is TDC on the harmonic balancer. Is it supposed to move around that much? Any thoughts? The engine ran great for as long as we were out there (30 plus minutes) and only threw the code P0138 (High O2 sensor circuit).

 

This problem is a tough one...

[HOrnbrod]

That's dayem solid and normal  - but you never left idle it sounds like. I can't see the graduated timing marks on the timing cover (the water pump pulley was blocking them), and the RPMs seemed to never change. To see the timing marks on the cover the aux fan needs to pulled. It's hard to see. At about 1200-1500 RPM the mark will jump up from 0* and sync at about 12*-14* BTDC, and should remain fairly steady up to about 3000 RPM w/o flying around. That's what you are looking for. 

 

You can't adjust the timing, but the timing light verifies that the ECU is doing it's job to advance the timing when increasing the RPMs, the distributor gear is engaged properly, and the distributor sync sensor is passing traffic okay.

[RyanL]

Ahh I understand now. Tomorrow, I will pull off the aux fan and repeat the test through the RPM range and post the findings.

 

If our results don't match up does that indicate a faulty ECU?

That's dayem solid and normal  - but you never left idle it sounds like. I can't see the graduated timing marks on the timing cover (the water pump pulley was blocking them), and the RPMs seemed to never change. To see the timing marks on the cover the aux fan needs to pulled. It's hard to see. At about 1200-1500 RPM the mark will jump up from 0* and sync at about 12*-14* BTDC, and should remain fairly steady up to about 3000 RPM w/o flying around. That's what you are looking for. 

 

You can't adjust the timing, but the timing light verifies that the ECU is doing it's job to advance the timing when increasing the RPMs, the distributor gear is engaged properly, and the distributor sync sensor is passing traffic okay.

[Eagle]

Random thought (perhaps fitting for a random problem):

 

This is a complete transplant. Did you move a complete 1999 wiring harness over to the MJ, or were portions (or all) of the wiring harness home-built?

 

Either way, is everything routed EXACTLY the way it was in the donor vehicle? Is there a possibility that you routed some wire associated with the cam sensor or CPS in such a way that it's picking up random inductive interference?

[RyanL]

Everything on the MJ is standard for the 1999 year XJ except the gas tank which uses the existing wiring, but just a different sender (1998 Dodge Dakota).

 

As for the routing, we followed the wiring as it went. Ex. The injector wiring rail (plastic rail next to valve cover) has three mounting pins that fit over the head bolts by the intake, and the sensors fall from there. Some wiring may not be in the exact position as the donor vehicle, but wouldn't that lead to the problem occurring ALL the time, instead of at random? The CPS routes right off the end of the aforementioned injector wiring rail, and goes down the edge of the intake to the CPS. The cam sensor is part of a cluster of connectors and can only be routed one way.

Random thought (perhaps fitting for a random problem):

 

This is a complete transplant. Did you move a complete 1999 wiring harness over to the MJ, or were portions (or all) of the wiring harness home-built?

 

Either way, is everything routed EXACTLY the way it was in the donor vehicle? Is there a possibility that you routed some wire associated with the cam sensor or CPS in such a way that it's picking up random inductive interference?

[HOrnbrod]

Ahh I understand now. Tomorrow, I will pull off the aux fan and repeat the test through the RPM range and post the findings.

 

If our results don't match up does that indicate a faulty ECU?

Not necessarily. What you're looking for is the timing advance. If it advances from idle to approx 12*-14* BTDC and hangs there up through the RPM, you can eliminate many things. But it won't be rock solid like an old school points and vacuum advance system because of all the other inputs to the ECU, so expect some jumping. Just check it out if you can and see what happens. Since you can run fine for awhile until you throw the code, it might not pinpoint anything because your problem is intermittent. But it can eliminate some possibilities causing your problem.

[Akula69]

The transplant is pretty much identical down to the placement of the engine bay components, as we wanted to ensure the harnesses would 'reach' their components. As Ryan said, the only variation I have seen up front is the Wells CPS has a slightly shorter whip (where it connects to the harness) that causes it to run in a straight line past the exhaust manifold...but I would consider that a good aspect (not as prone to melting) rather then an electrical interference one. The first CPS replacement was the exact length, but it was from Autozone and I wanted to eliminate the possibility of a "broke in the box" sensor.

 

We did make custom harnesses from the original XJ body harnesses in the cab back to the XJ tail light harnesses, but it was mainly just lengthening the existing wires. We did remove and cap the plugs for the rear door harnesses on both sides of the body harness. We did connect the Dakota sender plug to the XJ fuel sender wires. Although the wire colors were different the diagrams clearly showed the correct connections.

 

The harnesses were originally cleaned and re-wrapped where necessary, but after the onset of this problem the top harness (all sensors) was completely stripped and checked before being re-wrapped again with newer loom.

 

Last night I noticed the Temp sensor on the thermostat housing was fractured where the plastic connector meets the nut. We replaced it and ran the truck for another 5 miles. It behaved for the most part - except for the P0138 - 2nd O2 sensor high voltage, but it was much cooler here at that time. Although the problem has occurred at random times I am starting to suspect excessive heat may be a factor (although it has never overheated and hangs at 210 degrees/aux fan works as advertised). We will test run it today after we get the new video of the timing and post the results.

 

Thanks guys

 

Random thought (perhaps fitting for a random problem):

 

This is a complete transplant. Did you move a complete 1999 wiring harness over to the MJ, or were portions (or all) of the wiring harness home-built?

 

Either way, is everything routed EXACTLY the way it was in the donor vehicle? Is there a possibility that you routed some wire associated with the cam sensor or CPS in such a way that it's picking up random inductive interference?

[Akula69]

Broke down and took the truck to the stealership yesterday. Got there at around 0700 (before the service desk folks)  and had the chance to walk through the service bays - it was really disturbing to see all the mid- and full size Ram pickups in there with the transfer cases / transmissions down. The first tech that showed up said they were 4 days behind on service :ack: .

 

Anyway, made friends with the service manager who hooked the DRBIII to the truck and put it through it paces. Among the observations he / I made:

 

1) All Chrysler vehicles will occasionally go 'out of sync' occasionally. seeing that on a scanner is not necessarily an indication of problems.

2) $6000.00 scanners really are better and can make the truck do really cool stuff!

3) Small vacuum leaks inside the master booster , while not visible, will make very strange things happen to the way an engine runs.

 

He found the leak by watching the MAP pressure reading on the scanner while pressing the brake pedal. when the pedal went down the PSI dropped from around 17 to 10.

 

and, best of all:

 

Sometimes folks don't charge you because they are just plain nice! :clapping:

 

No running problems since booster was replaced, but still keeping fingers crossed. However, the good pressure of the new booster has uncovered brake line leaks.... :wrench:

[Akula69]

Well, the new master booster did not solve the problem, but it did eliminate the P1391 code. Drove it for about 30 miles today without problems, then when sitting at idle I slowly revved the engine from 700 up to 4500 and then let if fall off on its own. It threw the P0340 code directly after.

 

Next step is direct wiring from the PCM to the camshaft position sensor. I have an extra (salvage) harness that I will cut the plug from. The service mamager said that breaks internal to the wire (inside the insulation) can reduce the strands to 1 that will still show continuity but cause intermittant signal.

 

After that, who the he!! knows :dunno: .

[Akula69]

As of tonight we have:

 

Purchased new harness plugs for the CMP and the CKP, and ran seperate wiring from the plugs to the PCM, bypassing the harness entirely.

Purchased a new front O2 sensor to match the back one, since we were told a mis-match can cause problems we have been experiencing.

Ordered a THIRD OEM CKP from the stealership - after being told that it it the only "reliable' way to ensure the CKP is not "throwing stray signals into the circuit".

 

The problem has now solidified itself to only happening when the engine is warmed up (closed loop) and at idle. I can drive this truck around for 30-60 minutes and not experience a problem, but park it at idle for 5 minutes and it throws the P0340 and the P1391. I am thinking if the OEM CKP does not resolve the issue the only thing left is the PCM....

 

Any thoughts are welcome!

 

Thanks!

 

Jake

[oldseddie]

I have a stack of obd2 ecm's sitting around. give me the #'s and maybe i'll have the right one. I got a 97 w/ax15 doner so I don't need them for my project.

[Akula69]

Alrighty - some actual progress! I installed the stealership CKP this evening and then ran the truck for about 30 miles without problems. After that I let it idle in the driveway for an hour while I ate dinner, then I drove it for another 10 miles.

 

NO CODES! :banana:

 

I will slowly start reversing the harness changes (that I made to bypass it) tomorrow and monitor the progess.

 

After the successgul run, I compared the Wells CKP to the stealership one, and noted the Wells model did not have any magnetism at all...in fact, when I researched it on the internet the infor sheet on it (from NAPA) says it is an "advanced electronic model that does not need to be magnetised". Hmmmm.

 

Thanks for everyones help. :cheers:

[HOrnbrod]

Good to hear Jake. I'm a believer in using OEM factory parts whenever possible if available, especially electrical parts. Costs more up front, but usually saves big in the long run. I have a stockpile of most all the Mopar electrical OBD1 stuff I'll ever need. Unfortunately it's drying up quickly, especially the Renix stuff.