Random Misfire P0300 through P0306 - 2000 TJ Wrangler

[carlosa]

2000 Jeep Wrangler TJ 32RH 170k Miles

 

 

I bought this last fall with intentions of making it my forever Jeep since it replaces an identical one I had back when they were new. It had its engine replaced by a sketchy shop. I have been fixing things ever since. Some of the things I have fixed will be listed at the end of this post. In the interest of thoroughly understanding why this is happening and learning as much as possible, i`d like to keep this limited to diagnostics, testing, and troubleshooting rather than buying parts and bolting them on. I am well versed with the FSM, standard tools, and with Jeeps in general having owned everything from a CJ-5,7, XJ, WJ, YJ, TJ (twice).

 

What I currently do not have is a DRB III - if anyone has one they`d like to sell please let me know.

 

The problem: Random misfire flashing check engine light starting just over 3000 RPM and happening any time up to redline. Appears to happen more when cresting a long pass and then engine braking down the other side. Cannot feel any actual misfire. Over a long enough timeline codes will set for all 6 cylinders. No particular cylinder sets it off, but I tend to see p0306 earliest.

 

Data includes fuel trim, TSB 09-003-03, vacuum, fuel pressure, and leak down test results:

Fuel trim at idle is approximately 10% positive. And a snapshot at the time of the code being set is also usually around 10% positive. Photo attached shows the fuel trim at the time of a misfire, as you can see I floored it, light started flashing, and I took a screenshot, very simple.

 

I have not replaced the intake manifold gasket yet, but have thoroughly checked all vacuum lines with propane, and brake cleaner. Have replaced a few that were suspect, but did not actually respond to propane. My understanding is even if this were an issue with the intake gasket, it should not affect under heavy load - but maybe when engine braking? For further proof of this, I capped off all vacuum lines at the manifold including the master cylinder, and was able to reproduce the error in first gear at 3100 rpm driving up my street. A vacuum meter (mityvac silverline) shows the needle bouncing/vibrating wildly at idle - but only within 1 or 2 marks on the gauge. This is indicative of leaking or worn valve guides? Vacuum test does not indicate exhaust restriction at any RPM level.

 

TSB 09-003-03 - removed all rockers and verified a bullseye pattern is present. Did not rotate valves yet as I didn`t have a compressor available, and honestly I fully expected to be pulling the head during this process. Have run seafoam top end cleaner through the throttle body according to directions. Spark plugs are very clean and nice.

 

Leak down test (cold engine):

I have bad hearing, but could not hear or feel anything coming out of the throttle body, or the exhaust. No bubbles in the radiator.

cylinder 6: set at 69, observed 62.5, percent difference: 9.88%
cylinder 5: set at 70, observed 68, percent difference: 2.89% (was holding pressure by hand, cylinder may be lower in the bore than others)
cylinder 4: set at 70, observed 64, percent difference: 8.95%
cylinder 3: set at 69, observed 62, percent difference: 10.6%
cylinder 2: set at 69, observed 60, percent difference: 13.9%
cylinder 1: set at 69, observed 65, percent difference: 5.9%

 

Fuel pressure: 

I had a crappy HF gauge but did not observe any negatives on fuel pressure at any RPM. I can retest if someone would like to see specifics. I did swap in a fuel rail with a shrader valve, this is one of the things that led me to believe the engine is a later model.
 

Things that have been done - most of these are routine or were observed as being broken, missing, or neglected. I may have forgotten some but will update this list.

 

1. APP 985 plugs, removed AP985 plugs - this was done on day one.

2. Removed 180 degree thermostat, replaced with 195 - day one (the first time i drove it at the shop, didn't even have a thermostat in it)

3. Swapped in a known good coil rail from my XJ. No change whatsoever.

4. Built all new battery cables, soldered, marine heat shrink, direct cable from battery to body, and battery to block. ¹

5. Swapped crank position sensor - not because it was suspect, but because it was BROKEN, and eventually did die causing a no start. I assumed the shop broke it when they installed the new motor. ²

6. Crank position sensor died again, 6 months later. ²

7. Correctly rerouted the belt, and installed a new one along with idler just because - this was done in the first month, did not observe any overheating or charging issues.

8. Fully disassembled and cleaned the throttle body. Noticed the IAC casting has been replaced and is not a good fit - so lapped all surfaces on a machinist surface plate, then reinstalled with aircraft form-a-gasket. No improvement, if anything the idle stays a little high at times.

9. Have done the disconnect, ground, reconnect, lights on, lights off thing at times just to force relearn (if that even worked on later models?)

 

¹ When doing the leak down test today I realized there is NO ground strap from the block to the head - will be repaired shortly. Edit: Added the strap, did nothing. See the last screenshot of current results. p0306 the first code to pop up this time.

² What I thought was top end noise all along was one of the four torque converter bolts riding around inside the tone ring of the flywheel. I dropped the trans to replace the flywheel thinking it was cracked, and discovered the bolt. Pic attached. Removed and inspected the flywheel, noted that it does not have the exact same tooth count or tone ring setup as the autozone replacement I bought, so reinstalled. It was in good shape anyway. Used ARP bolts to the torque converter, properly torqued. Also finally got the throttle valve cable perfectly adjusted, it needed an extra click on the tension side in order to shift properly. The bolt riding in there is what broke two crank position sensors - its amazing that they still worked for months with the magnet not even attached.

 

Additional observations:

 

I do not have proof but it almost feels like the misfire code is set at the same time the clutch fan is operating. In other words I can feel the sluggishness of the fan as well as some minor vibration at the time the light flashes. I am considering upgrading to a mishimoto electric fan and not because of current issues.

 

I have noticed the ever so slight loss of coolant - where after 2 weeks or so it will be down to the top of the tubes when looking through the cap. It does not ever pull fluid from the reservoir, and has not dripped on the ground. Was able to pressurize the system and observe minor dripping from both hoses - going to write this off as solved.

 

This is a 0331 TUPY head.

 

The mileage, make, model of the engine are unknown - I am not entirely sure but it may have come from a WJ. I believe the date code reads 3 12 mx 17. Which is 2003-12-17

 

 

 

 

[Moses Ludel]

carlosa...Really appreciate your attention to detail!  See my red highlighted comments on your findings below:

21 hours ago, carlosa said:

2000 Jeep Wrangler TJ 32RH 170k Miles

 

 

I bought this last fall with intentions of making it my forever Jeep since it replaces an identical one I had back when they were new. It had its engine replaced by a sketchy shop. I have been fixing things ever since. Some of the things I have fixed will be listed at the end of this post. In the interest of thoroughly understanding why this is happening and learning as much as possible, i`d like to keep this limited to diagnostics, testing, and troubleshooting rather than buying parts and bolting them on. I am well versed with the FSM, standard tools, and with Jeeps in general having owned everything from a CJ-5,7, XJ, WJ, YJ, TJ (twice).

Thanks for the details...

What I currently do not have is a DRB III - if anyone has one they`d like to sell please let me know.

Watch the price...These can be expensive since Miller/SPX stopped making the units.  Many need this device.  Let us know what you turn up and be sure all of the adapters and useful programming come with the tool.  Otherwise, the equipment may not adapt to your TJ Wrangler's 4.0L engine.

The problem: Random misfire flashing check engine light starting just over 3000 RPM and happening any time up to redline. Appears to happen more when cresting a long pass and then engine braking down the other side. Cannot feel any actual misfire. Over a long enough timeline codes will set for all 6 cylinders. No particular cylinder sets it off, but I tend to see p0306 earliest.

You hinted about a poor engine replacement.  I would suspect an intake manifold gasket leak that could be causing a cylinder(s) to lose vacuum.  This low manifold vacuum at an individual cylinder(s) could throw a code.  Of course this could be an injector, wiring or poor grounds to the engine as well, but I'm guessing you've already addressed those possibilities?  The uphill driving trigger hints about a vacuum leak, which could be the intake manifold, the cruise control circuit or the cruise vacuum reservoir.  Check the vacuum circuits and intake manifold gasket for leaks.

Data includes fuel trim, TSB 09-003-03, vacuum, fuel pressure, and leak down test results:

Fuel trim at idle is approximately 10% positive. And a snapshot at the time of the code being set is also usually around 10% positive. Photo attached shows the fuel trim at the time of a misfire, as you can see I floored it, light started flashing, and I took a screenshot, very simple.

Helpful...

I have not replaced the intake manifold gasket yet, but have thoroughly checked all vacuum lines with propane, and brake cleaner. Have replaced a few that were suspect, but did not actually respond to propane. My understanding is even if this were an issue with the intake gasket, it should not affect under heavy load - but maybe when engine braking? For further proof of this, I capped off all vacuum lines at the manifold including the master cylinder, and was able to reproduce the error in first gear at 3100 rpm driving up my street. A vacuum meter (mityvac silverline) shows the needle bouncing/vibrating wildly at idle - but only within 1 or 2 marks on the gauge. This is indicative of leaking or worn valve guides? Vacuum test does not indicate exhaust restriction at any RPM level.

A manifold vacuum leak will drop available manifold vacuum, even more noticeably under load.  I had a vacuum line leaking at the cruise control circuit reservoir on our XJ Cherokee 4.0L and did not notice a symptom unless going up a grade:  The cruise would kick out.  This was due to low manifold vacuum under load on a grade.  Since the vacuum leak affected the engine uniformly and to a milder extent (the O2 sensor could compensate), no code appeared.

TSB 09-003-03 - removed all rockers and verified a bullseye pattern is present. Did not rotate valves yet as I didn`t have a compressor available, and honestly I fully expected to be pulling the head during this process. Have run seafoam top end cleaner through the throttle body according to directions. Spark plugs are very clean and nice.

 

Leak down test (cold engine):

I have bad hearing, but could not hear or feel anything coming out of the throttle body, or the exhaust. No bubbles in the radiator.

cylinder 6: set at 69, observed 62.5, percent difference: 9.88%
cylinder 5: set at 70, observed 68, percent difference: 2.89% (was holding pressure by hand, cylinder may be lower in the bore than others)
cylinder 4: set at 70, observed 64, percent difference: 8.95%
cylinder 3: set at 69, observed 62, percent difference: 10.6%
cylinder 2: set at 69, observed 60, percent difference: 13.9%
cylinder 1: set at 69, observed 65, percent difference: 5.9%

These are each very acceptable readings, some actually sealing better than expected (8-10% would be ideal on a stock engine with normally gapped piston rings).  You could compare this to a cranking compression test with a compression gauge just to eliminate any reliability issues around these leak down readings.  Interesting how this tester shows two pressure readings, and you're required to subtract one from the other to find the difference, i.e., the volume of leak or the percentage of leak.  At left is presumed as input or compressor line pressure; right is the actual reading from the cylinder with piston at TDC on its compression stroke with both valves closed?

Fuel pressure: 

I had a crappy HF gauge but did not observe any negatives on fuel pressure at any RPM. I can retest if someone would like to see specifics. I did swap in a fuel rail with a shrader valve, this is one of the things that led me to believe the engine is a later model.

Any '97-up TJ engine would be single-rail EFI with no return line to the tank.  As you know, 2000 should be a coil-on-plug engine and distributor-less.  Is this the case?  The "fuel regulator" looking device at the end of a single rail system is actually a fuel damper (like an accumulator), not a regulator.  The regulator is atop the fuel tank.  I'm guessing that you've tested fuel pressure at the Schrader valve on the rail?  Have you also tested fuel volume to be sure the filter at the tank/regulator is flowing fuel?  The fuel sock at the bottom of the fuel module (in tank) is unrestricted?  Lack of fuel flow volume and pressure could create a trouble issue.
 

Things that have been done - most of these are routine or were observed as being broken, missing, or neglected. I may have forgotten some but will update this list.

 

1. APP 985 plugs, removed AP985 plugs - this was done on day one.

2. Removed 180 degree thermostat, replaced with 195 - day one (the first time i drove it at the shop, didn't even have a thermostat in it)

3. Swapped in a known good coil rail from my XJ. No change whatsoever.

4. Built all new battery cables, soldered, marine heat shrink, direct cable from battery to body, and battery to block. ¹

5. Swapped crank position sensor - not because it was suspect, but because it was BROKEN, and eventually did die causing a no start. I assumed the shop broke it when they installed the new motor. ²

6. Crank position sensor died again, 6 months later. ²

7. Correctly rerouted the belt, and installed a new one along with idler just because - this was done in the first month, did not observe any overheating or charging issues.

8. Fully disassembled and cleaned the throttle body. Noticed the IAC casting has been replaced and is not a good fit - so lapped all surfaces on a machinist surface plate, then reinstalled with aircraft form-a-gasket. No improvement, if anything the idle stays a little high at times.

Alignment right and no IAC orifices plugged with Form-a-Gasket?

9. Have done the disconnect, ground, reconnect, lights on, lights off thing at times just to force relearn (if that even worked on later models?)

¹ When doing the leak down test today I realized there is NO ground strap from the block to the head - will be repaired shortly. Edit: Added the strap, did nothing. See the last screenshot of current results. p0306 the first code to pop up this time.

Block to head is not an essential ground; the head and block are iron and tied together with head bolts.  Alternator ground to block and block or head to battery is the key.

² What I thought was top end noise all along was one of the four torque converter bolts riding around inside the tone ring of the flywheel. I dropped the trans to replace the flywheel thinking it was cracked, and discovered the bolt. Pic attached. Removed and inspected the flywheel, noted that it does not have the exact same tooth count or tone ring setup as the autozone replacement I bought, so reinstalled. It was in good shape anyway. Used ARP bolts to the torque converter, properly torqued. Also finally got the throttle valve cable perfectly adjusted, it needed an extra click on the tension side in order to shift properly. The bolt riding in there is what broke two crank position sensors - its amazing that they still worked for months with the magnet not even attached.

You've really been there!  A lot of installer error findings for this engine...I like your attempt to identify year/application of the various components.  The vehicle and powertrain could be a mix as you suspect, you need parts confirmation each step of the way.

Additional observations:

 

I do not have proof but it almost feels like the misfire code is set at the same time the clutch fan is operating. In other words I can feel the sluggishness of the fan as well as some minor vibration at the time the light flashes. I am considering upgrading to a mishimoto electric fan and not because of current issues.

Thermostatically or electrically controlled fan?  If mechanical/thermostatic, the vibration might be significant but shouldn't throw an engine code unless there is an EFI connection that is loose/poor, and it gets an open by the vibration.

I have noticed the ever so slight loss of coolant - where after 2 weeks or so it will be down to the top of the tubes when looking through the cap. It does not ever pull fluid from the reservoir, and has not dripped on the ground. Was able to pressurize the system and observe minor dripping from both hoses - going to write this off as solved.

Good...A casting crack near an exhaust valve seat needs to be ruled out.  These can be troubling and often turn up only with the engine running under compression and heat.  A good test for this is a tailpipe emissions scope wand waved over the radiator fill neck with the engine idling and warm.  (Make sure not to get scalded by searing coolant when opening the system!!!  Use a lower test point temperature, just enough heat to open a casting crack.)  Combustion gases flowing through the crack will show up in the cooling system and are readable with an exhaust analyzer.  Head gasket leakage into the coolant passages is also readable.  Obviously, if this leakage is substantial, coolant will be blowing out the reservoir and overflow from extreme cylinder pressure and coolant volume.  Rule on this...

This is a 0331 TUPY head.

 

The mileage, make, model of the engine are unknown - I am not entirely sure but it may have come from a WJ. I believe the date code reads 3 12 mx 17. Which is 2003-12-17

Well, it's either a TJ or a WJ of that vintage,the tail end of the WJ era or later TJ.  The head casting would be the more significant clue.

 

 

 

You've done homework and taken a professional approach.  To quickly test the intake gasket for leakage, try a lower volatility spray like WD-40.  With the engine idling, avoid high exhaust heat areas and spray a light mist along the head to intake manifold seam.  You're listening for a change in rpm.  If your hearing is a concern, attach a tachometer while performing this test and watch for a fluctuation in engine rpm.  Isolate the IAC if necessary during the test.

Let us know what you find...

Moses

[carlosa]

Thank you for the detailed response.

 

I have to say i`m very happy to hear from you. I have your Jeep book, and also have read many articles written by you for various magazines since the mid 90's.

 

Rather than reply again inline, i`ll just address some of the points that are most pertinent to the current state of affairs. That may also make it easy to follow for others encountering this issue.

 

DRB III Yes, quite expensive used on ebay - anywhere from $1200-2000. I consider it a necessary investment if I am going to have this Jeep for many years to come. There are chinese knock-offs that I have yet to test, but may depending if I can win an auction.

 

Intake Manifold Leaks - I agree that the highest explanation for the positive fuel trim would be a vacuum leak. I have capped off all connectors and promptly reproduced the issue. I was careful to apply a minute layer of the aircraft sealant to ensure that it would not run or otherwise gum up the operation like RTV or silicone type sealants might.

One thing that leads me to question the vacuum leak hypothesis is that the fuel trim stays almost exactly the same at higher RPM's. I have attached two screenshots demonstrating this - one is at idle, one is at 2500 RPM. I did test out WD-40 this morning with no noticeable change in idle, while monitoring the idle on my scanner.

 

Exhaust causing fuel trim issues? This still leaves a potential exhaust leak that could be fooling the o2 sensors. I think it would be safe to go ahead and remove the intake and exhaust and reseal everything down to the collectors, while visually checking that the catalytic converters are in good shape. I just wanted to rule out pulling the head before doing this.

 

Fuel pump/injectors causing fuel trim issues? I honestly have not looked at the injectors in any way during this adventure, simply because it runs so well at all times. I can never feel the misfire happening. I will swap in a spare set of injectors I have (that came with the fuel rail), after I clean them thoroughly. Will use all new o-rings. I obtained another fuel pump pressure gauge from Harbor Freight. The quality is lacking - but after opening up the guage and resoldering the fitting, it does not leak! So will test out the fuel pressure. I am aware of the fuel pump capacity test needing to reach 1/4 liter in 7 seconds, will do this as well.

 

Yes, this is a coil on plug - distributorless engine. My original intent to obtain a DRB III was to test the cam position sensor alignment as part of my skepticism of the previous shops work. I don`t think this can cause a misfire anyway, so it has been moved further down the list.

 

The fan is a thermostatically controlled system. Using the factory style clutch on what appears to be a brand new water pump. 

 

Head cracking - I haven`t seen any indication of a cracked head in the form of bubbles, residue, or anything else in the coolant - but I will send off a sample or engine oil and coolant to Blackstone to have it checked. I am hoping that this head is one of the later ones where the problems were solved. But, even if it is cracked, would this cause misfire readings on all cylinders as well as random misfire?

 

Proper compression test - I can do this asap and report back.

 

Again, thank you for accepting the challenges of this topic. I will absolutely report back any findings or eventual fixes that solve this problem. From my readings on many forums, either the problem is never solved - or the person never returns to post additional details. This does seem to be a common issue, and those who have reported solving it have done it in the most round about ways. Spending thousands of dollars on dealer labor replacing cats, head, etc. Then discovering that it was a faulty sensor.

[Moses Ludel]

18 hours ago, carlosa said:

Thank you for the detailed response.

I have to say i`m very happy to hear from you. I have your Jeep book, and also have read many articles written by you for various magazines since the mid 90's.

You're welcome, carlosa...I appreciate the many folks who have supported my work from the early 'eighties forward and trust that our exchanges here at the forums will benefit others.  We're in a DIY era with endless online information flying around.  On that note, researched and thoughtful information serves far better than shooting from the hip.  I value your thoroughness and effort to take the right, professional approach.

Rather than reply again inline, i`ll just address some of the points that are most pertinent to the current state of affairs. That may also make it easy to follow for others encountering this issue.

DRB III Yes, quite expensive used on ebay - anywhere from $1200-2000. I consider it a necessary investment if I am going to have this Jeep for many years to come. There are chinese knock-offs that I have yet to test, but may depending if I can win an auction.

The tool is very useful.  There are other aftermarket "professional" grade scanners (OTC, Snap-On, etc.) that also work, but any of these packages require adapter cabling and software that is model specific and current.  For late model vehicles, Chrysler has abandoned scan tools completely at the dealership level and uses a PC based interactive software and Wi-Fi interface that eliminates the need for a standalone scan tool.  

StarScan (sequel to DRB-III) was supposed to be the end all tool, it only works on select vehicles and model years that predate the current WiTech and postdate DRB-III/OBD-II.  I have a virtually new StarScan that friends at Miller/SPX (now Bosch) say has similar architecture to the DRB-III but has no software to interface with OBD-II.  This expensive CAN-bus application tool/scanner sets in my toolbox, waiting for us to acquire the right Chrysler vehicle for our fleet!  (A JK Wrangler or later Ram 3500 Cummins?  Not any time soon, we have well-functioning Jeep and Ram/Cummins rolling stock now!  They're OBD-II.)

Intake Manifold Leaks - I agree that the highest explanation for the positive fuel trim would be a vacuum leak. I have capped off all connectors and promptly reproduced the issue. I was careful to apply a minute layer of the aircraft sealant to ensure that it would not run or otherwise gum up the operation like RTV or silicone type sealants might.

One thing that leads me to question the vacuum leak hypothesis is that the fuel trim stays almost exactly the same at higher RPM's. I have attached two screenshots demonstrating this - one is at idle, one is at 2500 RPM. I did test out WD-40 this morning with no noticeable change in idle, while monitoring the idle on my scanner.

The trim could be relatively unaffected, as the O2 sensor can still compensate A/F ratios within its parameters.  This is with an unloaded engine, your symptom occurs under load and during recovery from a load.  Each of these modes is vacuum sensitive.  I'm back to the quick WD-40 check for vacuum leaks around the intake manifold.

Exhaust causing fuel trim issues? This still leaves a potential exhaust leak that could be fooling the o2 sensors. I think it would be safe to go ahead and remove the intake and exhaust and reseal everything down to the collectors, while visually checking that the catalytic converters are in good shape. I just wanted to rule out pulling the head before doing this.

This is very sensible since there is an upstream and downstream sensor, too.  These manifolds are notorious for cracking, as you know, and exhaust dilution could occur with minor or negligible noise.  You share that noise could be hard to detect anyway, I know the feeling as an ex-heavy equipment operator, you might want to whiff the exhaust system with a tailpipe exhaust probe.  On stainless like this exhaust manifold, cracks are usually visible and not difficult to find under close inspection off the engine.

Fuel pump/injectors causing fuel trim issues? I honestly have not looked at the injectors in any way during this adventure, simply because it runs so well at all times. I can never feel the misfire happening. I will swap in a spare set of injectors I have (that came with the fuel rail), after I clean them thoroughly. Will use all new o-rings. I obtained another fuel pump pressure gauge from Harbor Freight. The quality is lacking - but after opening up the guage and resoldering the fitting, it does not leak! So will test out the fuel pressure. I am aware of the fuel pump capacity test needing to reach 1/4 liter in 7 seconds, will do this as well.

I like the fuel rail pressure test that you can perform now.  Fuel pressure does have an effect on flow volume.  One performance trick involves a boost in fuel pressure with stock injectors.  Similarly, you could be running more flow if the regulator (again, it's at the tank) is off spec or malfunctioning.  Pump pressure has more to do with creating the available volume/flow; the regulator is constantly dumping overage of fuel directly back into the tank.  Since this occurs on grades under load and decelerating, I'd be less suspect of the pump and regulator but still concerned about the pump's pickup sock.  A quick test of both fuel flow volume and pressure, each read at the Schrader valve, would be insightful.

Yes, this is a coil on plug - distributorless engine. My original intent to obtain a DRB III was to test the cam position sensor alignment as part of my skepticism of the previous shops work. I don`t think this can cause a misfire anyway, so it has been moved further down the list.

I'd still want to be sure that the shop indexed the cam position sensor properly.  You're also raising the question of valve timing.  How many miles does this engine have?  Is there a possibility of a loose timing chain and valve timing error?  This would affect vacuum and timing.  No camshaft position fault codes?

The fan is a thermostatically controlled system. Using the factory style clutch on what appears to be a brand new water pump. 

This would make engine faulting less likely with the exception of load.  I'm surmising that the fan coincidence is load/vacuum related.  The fan "coming on" would create more engine load, which could be a trigger for the faulting.  Is the MAP sensor testing on spec?

Head cracking - I haven`t seen any indication of a cracked head in the form of bubbles, residue, or anything else in the coolant - but I will send off a sample or engine oil and coolant to Blackstone to have it checked. I am hoping that this head is one of the later ones where the problems were solved. But, even if it is cracked, would this cause misfire readings on all cylinders as well as random misfire?

Not likely unless spent coolant affects the oxygen sensor readings and function. I was noting this strictly as a possibility, but if the coolant loss was leaky hoses, you do not have a casting crack issue here.

Proper compression test - I can do this asap and report back.

This is simply a check-and-balance.  Though you know I advocate for leak down tests to pinpoint trouble and get a better sense for actual cylinder, valve and ring wear, the leak down tests are not fail-safe.  I've seen carbon beneath a valve face throw off a leak down test reading.  Still, the leak down tester is the best diagnostic tool short of an engine teardown.  Of course, leak down tests do not indicate bearing and lubricated surfaces wear.  But your Blackstone testing of the drain oil does!

Again, thank you for accepting the challenges of this topic. I will absolutely report back any findings or eventual fixes that solve this problem. From my readings on many forums, either the problem is never solved - or the person never returns to post additional details. This does seem to be a common issue, and those who have reported solving it have done it in the most round about ways. Spending thousands of dollars on dealer labor replacing cats, head, etc. Then discovering that it was a faulty sensor.

I agree.  Sensor faulting at the MAP, crankshaft position sensor or CTS is common.  The CPS could have been knocked around according to your photos...Often, a bench or on-vehicle test of these sensors is inconclusive.  I've had aftermarket Brand-X TPS and and O2 sensors cause trouble because they are inconsistent with OEM voltage parameters.  I stick with Mopar or its industry supplier (like NTK by part number for an O2 sensor).  When the going gets touchy, OEM parts can often be the cure.

Moses

 

[carlosa]

EDIT: Moses, you are going to end up being totally right about this being an air or vacuum leak. I tried again with the wd-40 today on a warm engine and could not manipulate the idle. So a tear down and inspection of the gasket and manifold will be in order. Remember that I also drove with all vacuum lines capped off at the intake. So i`ll pull the intake and exhaust manifold, check for cracks, and replace all gaskets next. The reason for this is .. I just hooked up the fuel gauge and see no real issues either at idle or at 2500 RPM. Pics attached.

 

Thanks for all of this info. Its going to take me a few days to test everything and then I will report back. I will be doing a full series of testing on the fuel pump - from the flow capacity, pressure, voltage drop, etc. My gauge leaked again yesterday so I threaded the fitting back in with a thin layer of JB weld, i`m expecting it to be well cured and will test the pressure tomorrow both with key on, and with a direct 12v to the pump.

 

I found that the replacement non-oem throttle valve cable is binding a bit, so just went and grabbed one from a 1997 ZJ that is silky smooth.

 

Going to go ahead and attach a freeze frame from todays drive. This is after about an hour of varying speeds and I really had to push it to get the MIL to trigger. It went like this: Floored acceleration up to 80 mph, let off gas, coast down to about 60 mph, check engine light flashing during the entire coast down. Tapped the gas fairly hard and the light went solid.

 

I am trying to find a copy of my scan software that will let me monitor misfires. I know the vast majority of them do not, but autotap and autoenginuity do.

 

[Moses Ludel]

Looking forward to your findings, carlosa...You were uncomfortable with the shop's engine installation work, and the symptoms suggest a vacuum leak.  If the shop installed the intake and exhaust manifolds, your close inspection and gasket replacement could prove fruitful!

Your solution will be valuable to others who wrestle with this kind of MIL/code issue...Looking forward.

Moses

[carlosa]

So i`m not going to call this a resolution, but you got me thinking about that IAC. That is a hard device to bench test. I swapped in my throttle body from a 2001 XJ wholesale, with all sensors and everything and got this at idle (see attached image). I also beat on it a bit on the free way at all rpm ranges and could not get a flashing light. Idle is also very smooth. 

 

I have two concerns with calling this done:

 

#1 - I reset the computer again for relearn. I may need to get several more trips in before I can really write it off as solved. The Jeep did run a little bit hotter than usual, hitting 200 degrees .. it has rarely if ever even hit 195. I would say this is a result of a much less than positive fuel trim aka richer condition due to a better stoichiometric ratio. There is also still a tiny bit more positive fuel trim than i`d like to see, but it is well within spec and has no wild swings into the 20's and 30's %.

#2 - I can`t call it done until I devise a way to test the map, tps, and iac and see exactly what is going on with them that is out of spec. Obviously the IAC is not regulating air flow, as Moses mentioned several times in above posts. I did inspect the removed part and it is not gummed up with form a gasket, and besides .. this even happened long before I took it apart to clean it and reseal it. I will be posting again with photos and exactly what I find.

 

 

[Moses Ludel]

A near wrap, carlosa...Comments below:

On 5/17/2017 at 6:36 PM, carlosa said:

I have two concerns with calling this done:

 

#1 - I reset the computer again for relearn. I may need to get several more trips in before I can really write it off as solved. The Jeep did run a little bit hotter than usual, hitting 200 degrees .. it has rarely if ever even hit 195. I would say this is a result of a much less than positive fuel trim aka richer condition due to a better stoichiometric ratio. There is also still a tiny bit more positive fuel trim than i`d like to see, but it is well within spec and has no wild swings into the 20's and 30's %.

Likely "normal" for the engine and cooling system at these A/Fs.  Trade-offs as you hint.  Good indicator of the relationship between A/F and combustion temperatures.  Check CTS voltages to see whether the temp sender is accurate.  This is another area where aftermarket (non-OEM) coolant temp sensors can be off on voltage.  I installed a Brand-X sensor in the '99 XJ Cherokee 4.0L, no other changes, and the temp gauge reads 200 instead of 195 when it clearly should read 195.  Just sayin'! 

#2 - I can`t call it done until I devise a way to test the map, tps, and iac and see exactly what is going on with them that is out of spec. Obviously the IAC is not regulating air flow, as Moses mentioned several times in above posts. I did inspect the removed part and it is not gummed up with form a gasket, and besides .. this even happened long before I took it apart to clean it and reseal it. I will be posting again with photos and exactly what I find.

You're on it...This is a process of elimination while pinpointing sources of trouble.  I'm confident that system problems like this do have solutions.  You've gone a long way toward solving the problem for yourself and others!

Moses

 

[carlosa]

Alright here is another little update with photos. I have tested various things over the past week, but haven`t had as much time to really work on it. That and i`m waiting on more tools to arrive to be able to test other things.

 

1. I did a traditional compression test. Cylinders were from 148 to 160 PSI. Most were basically right in the middle at around 155.
2. I pulled the injectors. Observed small amounts of rust inside the fuel rail, found one cracked injector - nothing leaking on the outside. Ordered new baskets. I pulled the one basket from the cracked injector and it was quite rusty and dirty but not sure its enough to affect the flow. By rusty I mean the nylon screen is orange rather than some shade of nylon. Replaced the one cracked injector with a spare I had
3. Going to do an injector balance test before & after I change the screens. I just want to see if there is any flow issues so that I can see the difference. Likely will replace all with a rebuilt set, anyway.
4. Did a fuel pump volume test. Spec is 1/4 liter per 7 seconds. I received quite a bit more than 1/4 liter as shown in a photo below.
5. The manual specifies that the coil rail should reach 25 KV. I tried testing this by removing the rail, and hooking up a cheap spark test tool. It worked, but guessing where the spark fell on the gauge was kind of a joke. If I had to guess I would say it was right at the threshold or slightly below 25 KV on all cylinders. But, I tested two different coil rails with the same results. So for now i`ll rule that one out. Gauge is pictured.

Have driven about 200 miles mixed city and freeway and have not had a light .. but I know if I push it and drive harder it will trigger.

 

Still to come: Exhaust pressure test before/after cats, fuel injector balance test, locate a DRB III !!! I really want to see live misfire data, if any - and so far have not been able to recover that from any scanner I have. Auto Enginuity, Torque pro, OBD Link, etc.

[carlosa]

[Moses Ludel]

carlosa...Great work!...This is the scientific way to approach an automotive problem.  All things mechanical have solutions, and you're after it!  Thanks for sharing the details and photos, you are demystifying the issue and approaching a lasting solution: it's a process of elimination...

I like your hands-on checks.  The MIL and trouble code throws can be useful, but much of what you're finding would not show up on a code reader.  Your scan tools are good, but as you note, the interactive DRB-III gave dealership techs a huge advantage.  These tools also put a big dent in the dealer's tool room budget.  Current use of WiTech is clearly more cost effective, the system uses the same level of sophistication as DRB-III yet works with a common PC base and an upload/download interface directly to Chrysler.  There were rumors of WiTech being backward compatible at some point to OBD-II (the DRB-III scan tool era).  I'm unclear whether that materialized.  Getting a DRB-III tool and accessories is pretty much for OBD-II vehicles and earlier.

Your troubleshooting methodology is very helpful to others.  Feedback on various scan tools would be equally valuable, as everyone wants a tool(s) that solves problems.  You're very busy now with the current troubleshooting, but if/when you do have time, a new topic on the assets and limitations of various scan tools, from your experience, would be hugely helpful to others.

One question:  How are you ruling out the PCM as a trouble source?  When our 2005 Ram Cummins had only 800 miles on the odometer, the engine developed a random and severe misfire.  The ECM threw injector misfire codes and nothing more.  This ECM/PCM could not interrogate itself in that state, and Chrysler would not accept that the ECM could fail without throwing a code.  We did have a genuine misfire code (classic #1 cylinder), and the dealer tech, Eric Benson, was instructed to swap injectors around and even change an injector.  Eric performed lamp load tests on the engine's ground circuits, he basically eliminated any issue related to the misfire codes thrown.  

After pleading with Chrysler to allow installation of a Cummins Recon ECM, the engineers reluctantly authorized an ECM replacement.  The truck now has 159,000 miles of trouble free engine operation.  There were 25 other new Ram/Cummins trucks in the Zone that were stalled with the same problem.  Eric Benson at Carson City Dodge was the solution...A couple of years ago, I received a letter from Chrysler stating that the ECM on all Cummins/Ram trucks within our VIN cohort would be granted lifetime warranty on the ECM.  Chrysler was only a decade behind Eric.

Have a good holiday weekend...

Moses

P.S.:  I like the "cheap" Lisle spark tester.  Sinisterwillys1940 could have used this tool in his recent spark wire testing.

[carlosa]

I have not eliminated the PCM at all. My approach is to test and eliminate the least invasive or easy access items, before moving on to more invasive or less easy to test items. So for example - the vacuum leak that you mentioned a few times is still high on my list, but I want to avoid tearing into that engine that far until I have eliminated everything else. The computer very well could be suspect. My auto enginuity won`t even connect to it. I have a solution for that as well that I will work on eventually.

The situations where a PCM fails and needs to be replaced are very interesting to me. Interesting how it can sort of fail and still allow driveability. 

At the moment its a waiting game either for ordering tools, or for doing a few things then working, then doing a few more. I currently don`t have any budget restrictions on getting this fixed, so I will buy and test anything that I can get my hands on -- however without just "changing parts".

[Moses Ludel]

I guess you could call it "driveability".  Our Ram limped home from Southern California's Off-Road Expo to the Reno Area with a slight misfire.  Cold startup was rough with an unstable idle.  This evolved into a motor mount shaking miss and stumble that sent us immediately to the dealership...Yes, this was nearly 1000 miles of driving with a defective ECM. 

[carlosa]

1 hour ago, Moses Ludel said:

I guess you could call it "driveability".  Our Ram limped home from Southern California's Off-Road Expo to the Reno Area with a slight misfire.  Cold startup was rough with an unstable idle.  This evolved into a motor mount shaking miss and stumble that sent us immediately to the dealership...Yes, this was nearly 1000 miles of driving with a defective ECM. 

That is amazing, i`d love to find out exactly what it is inside the computer that fails. There arent any moving parts, so maybe a resister or continuity issue? Or a corrupted database table?

 

 

[carlosa]

Injector balance test results are in. 

 

For those not familiar: An injector balance test basically fires an injector a timed, specific number of times. You can then fire each injector and compare the results. Results are driven by the "leak down" PSI on the fuel rail. Each time you fire one, you re-prime the fuel rail and blow out unburned fuel, by cranking the engine. Ignition coil is unplugged during this scenario. The fuel rail was pressurized to basically 45 PSI by cranking a few revolutions using a hand crank button that links the battery to the starter solenoid. The pulser was set on 50 pulses, each 7 MS in length.

 

#1: 33 PSI

#2: 32 PSI

#3: 32 PSI

#4: 33 PSI

#5: 33 PSI

#6: 33 PSI

 

So, these are not terrible. But it does indicate that 4 injectors are flowing less fuel in the same amount of time, as the other two. One of the 2 was replaced with a spare junk drawer injector, due to finding a crack.

Next step: Replace the baskets on all 6, and repeat. If I find that they balance out better OR perhaps flow more fuel then that is good news! I don`t have any sort of specification on how much fuel they should flow under these conditions.

If I see an improvement after changing the baskets, I may order a set of guaranteed quality rebuilt 4 hole Bosch injectors. I would not be doing this for any of the "perceived" benefits of upgrading to 4 hole, other than the fact that the price is similar, and they are not susceptible to cracking.

[Moses Ludel]

Very thorough and sensible test, carlosa!  For owners unfamiliar with EFI, the duration (time interval) of the fuel pulse determines the volume of fuel flowing.  The nozzles are a constant size, the injector pintle's pulses (50) are controlled, and in this test, each injector opening is timed for 7 ms.  The variable factor is the fuel rail pressure, which in this case carlosa has monitored closely.  The pressure varies due to fuel flow volume and function of the PCM, wiring circuits and battery voltage, the grounds, injector pintle, the injector nozzle flow (or restriction) and the accuracy of the timed pulses.

Injectors '101':  The fuel flow volume with EFI injectors is determined by 1) the period the pintle stays open and whether it seals and seats properly, 2) the injector nozzle sizing (or restrictions like gum) and 3) the pressure of the fuel at the rail.  For this reason, if the injectors are in good condition, unrestricted and matched for size, the fuel flow volume can change due to either the pintle opening period or the fuel rail pressure when the pintle opens.  Fuel flow will vary due to the injector nozzle sizing, the pintle opening interval (controlled by the PCM) or the fuel pressure at the rail (controlled by the pressure regulator setting and an unrestricted return flow).

What I like is your balanced approach to troubleshooting.  You follow up on the data readouts with mechanical tests.  You're not wasting money or simply "parts changing" to get ahead of the problem.  You want a definitive answer for "why" the problem is occurring.  This leads to a permanent fix and peace of mind, knowing that you drilled down and found the trouble source and corrected the problem(s).

This is a far cry from relying upon the MIL codes from a code reader (something we all do for starters, let's not knock it!) and a race to the parts house with a list of components associated with each trouble code.  The particular trouble codes addressed here have been elusive, often misleading and costly for many owners.  You're working toward a clear solution to the problem.

In fairness to many owners, the aim is to stay out of the dealership service bay and $100 plus per hour troubleshooting.  As we have discussed, there is a time for a DRB-III scan, and that can be less costly than needless parts replacing.  OBD-II was breakthrough technology intended to benefit consumers and independent shops. It established a standard diagnostic plug design for retrieving "universal" trouble codes to simplify fuel-and-spark management diagnostics. Unfortunately, the vehicle manufacturers still held the cards.  They upstaged the aftermarket with proprietary high-end diagnostics and reflashing tools like Chrysler's DRB-III scanner and accessories.  

Dealership diagnostics have been more sophisticated and model-specific than most of the aftermarket tools available to independent shops. The DIY techs found it cheaper to sublet diagnostics to the dealership. While the law insisted that the aftermarket have access to diagnostics equipment, the dealerships were clearly ahead of the aftermarket.  

Today, we're finding higher end scan tools from Snap-On, OTC and others that come much closer to the dealerships' assets.  Like you're discovering, many of these tools pop up in the used market and can serve well.  Your diagnostic tool experience and more importantly your troubleshooting methods prove valuable to others.

Moses

[carlosa]

On 5/16/2017 at 10:11 AM, Moses Ludel said:

I agree.  Sensor faulting at the MAP, crankshaft position sensor or CTS is common.  The CPS could have been knocked around according to your photos...Often, a bench or on-vehicle test of these sensors is inconclusive.  I've had aftermarket Brand-X TPS and and O2 sensors cause trouble because they are inconsistent with OEM voltage parameters.  I stick with Mopar or its industry supplier (like NTK by part number for an O2 sensor).  When the going gets touchy, OEM parts can often be the cure.

 

Hi Moses,

Random misfire issue may be solved

Sorry for my long absence, i`ve been very busy with work and parked the Jeep for the summer - I ran out of time to tinker with it and was afraid I might be hurting it even more.

 

Your quoted post above had me thinking for a long time about the quality of the CPS. I had a premium Napa  sensor in the Jeep after I pulled the transmission and fixed the flywheel bolts. This video came up recently, so I added it to my offline viewing list and watched it during an international flight. 

 

 

When I landed I decided to just take a risk and change a part... I ordered a new Mopar OEM sensor off of ebay. As far as I can tell it is identical to the Napa premium sensor, both are made in Mexico and both have the same general feel and quality. I put the new sensor in, took my time getting it snugged down tight against the spacer, and took it for a drive. Its now been 5 days of hard driving in first and second gear to keep the RPM's up, and I have had no re-occurrence of the random misfire codes. I have tried duplicating every scenario that previously would be a flashing light within seconds, and cannot. Engine braking down the hills here in the rain is much safer than riding the brakes on this Jeep and that alone is a bonus.

 

Now I have a transmission issue, and some free time .. so i`d like to get back into some of the other posts we talked about. I may also try to build some sort of test rig to test and understand what goes wrong in these sensors, since it is a widespread problem. Otherwise, maybe I should invest in (hoard) Mopar sensors for the future.

[Moses Ludel]

Carlosa...Glad you got results.  The aftermarket sensor issue is epidemic.  Apparently, the voltage and ohms parameters are wrong.  It's likely that suppliers are cataloging parts to meet a variety of vehicle/engine applications, creating this kind of trouble.  I'm especially leery of X-brands from offshore, though its not necessarily the foreign manufacturers' fault as much as the attempt to make these parts fit a wider range of applications.

Summing it up as shared earlier, I cross the OEM part to the OEM suppliers' part number and get the OE equivalent part:  I buy Bosch, NTK and other major brands if pricing is better than the Mopar labeled part.  During the Great Recession, even classically valued brands like AC/Delco went down in both quality and fitment accuracy.  When otherwise not available, buy OEM Jeep/Mopar parts.

Moses