Mystery P0300 root cause?

[jeepwarrior]

2007 jeep liberty v6 3.7L automatic
190k miles


Hi
I'm chasing down a difficult P0300 code with CEL flashing but it only happens when coasting downhill over 50+ mph. I've checked and tested and replaced dozens of things in the process. I also am now getting random long recently might be related starts. P0300 just started happening on way home on long trip.

Freeze frame data , nothing looks odd other then 400rpm which is probably normal in that situation

Idle isn't as smooth as it should but not that bad, usually within hovers at 600rpm +- 30

Misfire counts
I notice a pattern where misfires are spread across all cylinders but bank 1 cylinders 1,3,5 are the worst, sometimes 2-10x worse on that side. And #6 cylinder always has least amount of misfires maybe a handful where #1, 5 can have 100+


Tested and replaced

Fuel system
Fuel pressure 56psi

Fuel pump waveform, looks a bit choppy w 1.5v different commutator humps

Fuel pressure leak down * I do sometimes get different leak down results, sometimes it drops to 38psi over 5min.

***im replacing fuel pump tommorow since between radom bad looking leak down tests and uneven choppy waveform, could possibly solve both the misfires and or random long starts. Even though psi and flow rate check out I'm not confident it's not the fuel pump.

Fuel flow, seems fine, 1gpm unrestricted flow

Replaced injectors w new
Only use gas from top tier station 87octane

---
Ignition
Tested spark strength
Tested waveforms for primary and secondary
Replaced plugs w oem
New coil packs
---
Timing
Replaced cam and crank sensor
Tested cam crank correlation waveform
---
Sensors Tested and replaced
All 02 sensors Replaced
Map sensor is only year old and live data looks correct, checked w vaccum guage on intake too No issue
---
Engine breathing
Smoke leak Tested intake and exhaust system
Checked catalytic converter is good
Zero exhaust back pressure
New air filter
Cleaned throttle body and plate

*Tested cleaned egr valve, I did find leak coming out pintle vent hole and plugged that, no change, but mine and salvage yard one both leak outside air from pintle shaft seal through vent hole, I think it's designed to leak there a little?
---
System voltage good 14.2v
Battery load test good
Block and body Grounds all good no voltage drop
---
Mechanical
Compression test looks all within spec

Cylinder Leakdown test, I've checked some with bad leaking harbor freight tool bit seemed ok, retesting today w better tool

Evap purge valve seems fine


---
PCM
Only odd thing I'm noticing is that it doesn't go into open loop drive deceleration mode ever unless I turn off overdrive by hitting shifter button. I'm assuming deceleration mode in pcm only works with overdrive off by design, but there's no detailed pcm operation information anywhere I can find. Service manual is vague, and I don't know anyone with a Jeep liberty I can compare to.

 

[lfhoward]

Wow you have been thorough! I wish I could think of something that you didn’t do yet.

With your 07 you have drive by wire, like the 08+ KK. There is no throttle cable or idle air controller like on previous years. It could be something computer related maybe.

Since it is not confined to one cylinder I was thinking the fuel:air mixture must be off somehow. Check your MAP sensor to see if it is an aftermarket or Mopar unit. These Jeeps seem to run right using OEM sensors and aftermarket ones can cause issues even when new. Also an intermittently weak fuel pump came to mind, so I’m glad you are ruling that out.

Otherwise, not sure. I will think about it and if I come up with anything else I will let you know.

No other codes or pending codes after it misfires coasting downhill?

 

[jeepwarrior]

Wow you have been thorough! I wish I could think of something that you didn’t do yet.

With your 07 you have drive by wire, like the 08+ KK. There is no throttle cable or idle air controller like on previous years. It could be something computer related maybe.

Since it is not confined to one cylinder I was thinking the fuel:air mixture must be off somehow. Check your MAP sensor to see if it is an aftermarket or Mopar unit. These Jeeps seem to run right using OEM sensors and aftermarket ones can cause issues even when new. Also an intermittently weak fuel pump came to mind, so I’m glad you are ruling that out.

Otherwise, not sure. I will think about it and if I come up with anything else I will let you know.

No other codes or pending codes after it misfires coasting downhill?

Thanks haha, I made a big check list as just about anything can cause misfires. I actually still have more tests to do, like borescope cylinders abd injectors ports. And Leakdown test again of all cylinders, just for the heck of it I'm going to swap everything from cylinders 1, to 6 and see if misses change pattern.

Also need to check knock sensors are not the issue

I've done no tests on transmission, maybe the torque converter or lock up issue? Since it's misfires only when engine braking and zero throttle


its a aftermarket sensors, but I check their values and function and it didn't make things worse or better either.

The MAP sensor reports barometric pressure correct, and drops down at idle as it should,

Long term Fuel trims currently at 5%


Jeep runs fine other then throwing code and flashing downhill cel

 

[KJowner]

Inlet manifold air leak maybe?? Lean mixture when the throttle butterfly is causing a vacuum.

 

[jeepwarrior]

Inlet manifold air leak maybe?? Lean mixture when the throttle butterfly is causing a vacuum.

Hi, as I wrote the intake and exhaust was smoke tested for leaks, and long term fuel trims around 6% or so no indication of vaccum leak

 

[Jeremy-WI]

I think the deceleration mode only works going downhill using cruise control on my 2006 and then it happens when the speed is about 5 mph over the cruise setting

 

[jeepwarrior]

I think the deceleration mode only works going downhill using cruise control on my 2006 and then it happens when the speed is about 5 mph over the cruise setting

Hi, So your scan tool doesn't show pcm entering open loop drive deceleration mode normally driving downhill unless Cruise control is active?

You can feel it to even without scan tool, it feels like engine braking vs no engine braking. Feels like if it were stick shift and went downhill without the clutch with it in gear not neutral, you can feel the slight compression braking.

Mine it feels like transmission is neutral the whole way downhill and there's no engine braking, I have to constantly use brake pedal to slow car down. Usually after about 10-20 seconds my liberty starts flashing the CEL with p0300

 

[Jeremy-WI]

I haven't hooked up a scantool to see what it may report but I am sure that it just shifts the trans into third gear with the TCC locked

My 96 Chevy would shut off the injectors going downhill long enough

 

[jeepwarrior]

I haven't hooked up a scantool to see what it may report but I am sure that it just shifts the trans into third gear with the TCC locked

My 96 Chevy would shut off the injectors going downhill long enough

So you're saying w Cruise off and overdrive on you can feel it compression braking vs feeling like you're in neutral coasting?

 

[Jeremy-WI]

No braking then, the automatic braking only happens while using cruise

 

[jeepwarrior]

No braking then, the automatic braking only happens while using cruise

It's stupid the factory service manual doesn't explain the pcm inner workings or give a logic tree so a person can know exactly what inputs and outputs and decisions the pcm is suppose to make. The factory service manual just says under hard deceleration the pcm switches to deceleration mode, which I can see it switch to "open loop drive" instead of closed loop, BUT it only switches if OD overdrive button is selected. I just wanted to know if that's how it's supposed to work or something is wrong with my pcm or transmission or motor etc as to why it's not working as it should.

 

[jeepwarrior]

I'm going to hack the pcm and know it's hardware and firmware as soon as I get a chance to setup a breakout box and J box and whatever software I need as it's ridiculous that I or any mechanic has no idea what is going on inside a critical component. I have no understanding of its inner workings or how to confirm something is wrong w a module other then smoke coming out of it, or it's totally dead and not outputting ignition and injector signals even with confirmed good inputs. And hook up benchtop simulator so I can test and fix and diagnose models. And code and program modules as well if I have install a new or salvage model on any car

I know its a chrysler NGC computer that uses ground side switching, I know the inputs and outputs

 

[jeepwarrior]

I know the module sometimes will set a dtc code, but sometimes they fail without a code, like a loose component inside or cold joint and then the car does different things randomly depending on temperature or vibration. Or capacitor or driver, etc is burnt out

 

[ikuo78]

I guess the remaining mechanical possibilities are below.

・pcv
・Throttle butterfly play
・Canister deterioration
・Cam chain tensioner

 

[jeepwarrior]

Hi
Pcv was replaced

It's got electronic throttle body plate and feels tight no play in it, was cleaned also

You mean evap canister? No evap codes, how would I test canister? How would that cause p0300?

I checked the cam crank correlation waveform with known good, it's right on, without any changes watching it, and no chain slap noises?

 

[ikuo78]

If the canister cracks, it will breathe in the atmosphere directly.
Only when the valve is open.
Listen and visually check.

If there is a blockage in the evaporation system, the pressure in the fuel tank will be disturbed.
As a result, fuel pressure becomes unstable.
In Liberty's case that includes a vapor vent filter.
This is suspected if there is a large amount of air coming in and going out when you open the fuel cap after driving.

 

[dtennes]

Maybe an Intake Air Temp sensor and/or IAT connector problem.

This may seem "left field"... however, I have been chasing a P03xx ghost-in-the-machine problem for over a year now and I stumbled into a piece of trivia that may be the ultimate root cause. Hear me out.

If you search for stuff I've posted and pay attention to timeline, you will find that I have done everything you are doing and much more. Including complete custom wiring harness for coils and injectors. After all of that the Liberty seemed to be running fine for a couple of months then about a month ago, out of nowhere, I got a rogue P0300 (indeterminate misfire) *plus* random P0301 and P0302 codes. So I cleared the codes and waited. 30-50 miles later the codes pop up again. WTF.

I have a local mechanic who really knows electronics, but this is a weird one. He mentioned that it might have something to do with the IAT sensor. No explanation, no rationale, just a suggestion that popped into his head out of the blue like some sort of Oracle of Delphi thing.

It turns out that the connector for my IAT doesn't quite lock onto the IAT itself, so it can wiggle loose and still be in place. I cram the thing down as far as I can, but it still doesn't lock (POS aftermarket pigtail). It seems to do the trick but after a couple of weeks of bouncing over parking lot speed bumps the P03xx and random cyl misfires pops again, so I check that connector... which slides right off with a slight tug. Cram it down again... a couple of weeks later the codes pops again.

I had a spare segment of wiring harness I got from a salvage yard that included an original Mopar IAT connector. I cut that off into a pigtail and spliced it in place of the POS aftermarket pigtail and it locks on. Going on 2 weeks with no codes so far, still locked on tight.

If you cannot track down an original connector through ebay or a trip to a salvage yard it can be fabricated. The key thing is that the housing must lock firmly in place.

There is no exact Mopar part number for the pigtail connector housing because it is part of the complete wiring harness, but the plug that actually locks has the following markings: AMP F4 PBT GF15.

This was originally an AMP housing, and AMP is a division of TE Connectivity. When I track the lineage through TE I found this:

HOUSING ONLY (https://www.te.com/en/product-2822363-1.html
Purchase: https://www.digikey.com/en/products...3-1/5762219?s=N4IgTCBcDa4BxjAZgGxILQEYQF0C+QA

FEMALE RECEPTICLE (https://www.te.com/en/product-171662-5.html?compatible=2822363-1)
Purchase: https://www.digikey.com/en/products/detail/te-connectivity-amp-connectors/171662-5/1891347

WATERPROOF WIRE CAVITY SEAL (https://www.te.com/en/product-2822357-1.html)
Purchase: https://www.digikey.com/en/products...-1/5357192?s=N4IgTCBcDa4BxjAZgKwHYC0BGEBdAvkA

...and you will of course need some TXL wire (check confirm the correct gauge), buy some at AutoZone.

***You need all 3 to complete fabrication***, plus some 18awg TXL (or silicon insulated) wire from AutoZone, and a decent crimping tool. Don't buy parts directly from TE (commercial quantities only), use the purchase links or look up the parts elsewhere (Mouser.com).

I will continue to monitor the results of installing a proper locking connector for the IAT, and if this turns out to be yet another goose chase I will update this posting.

 

[jeepwarrior]

Maybe an Intake Air Temp sensor and/or IAT connector problem.

This may seem "left field"... however, I have been chasing a P03xx ghost-in-the-machine problem for over a year now and I stumbled into a piece of trivia that may be the ultimate root cause. Hear me out.

If you search for stuff I've posted and pay attention to timeline, you will find that I have done everything you are doing and much more. Including complete custom wiring harness for coils and injectors. After all of that the Liberty seemed to be running fine for a couple of months then about a month ago, out of nowhere, I got a rogue P0300 (indeterminate misfire) *plus* random P0301 and P0302 codes. So I cleared the codes and waited. 30-50 miles later the codes pop up again. WTF.

I have a local mechanic who really knows electronics, but this is a weird one. He mentioned that it might have something to do with the IAT sensor. No explanation, no rationale, just a suggestion that popped into his head out of the blue like some sort of Oracle of Delphi thing.

It turns out that the connector for my IAT doesn't quite lock onto the IAT itself, so it can wiggle loose and still be in place. I cram the thing down as far as I can, but it still doesn't lock (POS aftermarket pigtail). It seems to do the trick but after a couple of weeks of bouncing over parking lot speed bumps the P03xx and random cyl misfires pops again, so I check that connector... which slides right off with a slight tug. Cram it down again... a couple of weeks later the codes pops again.

I had a spare segment of wiring harness I got from a salvage yard that included an original Mopar IAT connector. I cut that off into a pigtail and spliced it in place of the POS aftermarket pigtail and it locks on. Going on 2 weeks with no codes so far, still locked on tight.

If you cannot track down an original connector through ebay or a trip to a salvage yard it can be fabricated. The key thing is that the housing must lock firmly in place.

There is no exact Mopar part number for the pigtail connector housing because it is part of the complete wiring harness, but the plug that actually locks has the following markings: AMP F4 PBT GF15.

This was originally an AMP housing, and AMP is a division of TE Connectivity. When I track the lineage through TE I found this:

HOUSING ONLY (https://www.te.com/en/product-2822363-1.html
Purchase: https://www.digikey.com/en/products/detail/te-connectivity-amp-connectors/2822363-1/5762219?s=N4IgTCBcDa4BxjAZgGxILQEYQF0C+QA

FEMALE RECEPTICLE (https://www.te.com/en/product-171662-5.html?compatible=2822363-1)
Purchase: https://www.digikey.com/en/products/detail/te-connectivity-amp-connectors/171662-5/1891347

WATERPROOF WIRE CAVITY SEAL (https://www.te.com/en/product-2822357-1.html)
Purchase: https://www.digikey.com/en/products...-1/5357192?s=N4IgTCBcDa4BxjAZgKwHYC0BGEBdAvkA

...and you will of course need some TXL wire (check confirm the correct gauge), buy some at AutoZone.

***You need all 3 to complete fabrication***, plus some 18awg TXL (or silicon insulated) wire from AutoZone, and a decent crimping tool. Don't buy parts directly from TE (commercial quantities only), use the purchase links or look up the parts elsewhere (Mouser.com).

I will continue to monitor the results of installing a proper locking connector for the IAT, and if this turns out to be yet another goose chase I will update this posting.

Hi, mine shows fine in scan tool live data and freeze frame and in recorded data

 

[jeepwarrior]

If the canister cracks, it will breathe in the atmosphere directly.
Only when the valve is open.
Listen and visually check.

If there is a blockage in the evaporation system, the pressure in the fuel tank will be disturbed.
As a result, fuel pressure becomes unstable.
In Liberty's case that includes a vapor vent filter.
This is suspected if there is a large amount of air coming in and going out when you open the fuel cap after driving.

So the engine would receive air not fuel air vapor and possibly misfire?

Why wouldn't it misfire all the time then? Why only going downhill coasting?
.why wouldn't I have error code for evap?

What about temporary disconnect purge valve from intake so it's venting outside vs into intake, see if misfires return

What about smoke test tank and see there's any leaks?

 

[ikuo78]

So the engine would receive air not fuel air vapor and possibly misfire?

Why wouldn't it misfire all the time then? Why only going downhill coasting?
.why wouldn't I have error code for evap?

What about temporary disconnect purge valve from intake so it's venting outside vs into intake, see if misfires return

What about smoke test tank and see there's any leaks?

It's about possibility.
The conditions that may indicate a canister failure are described above.

I think you can think about it by recording the correlation between live data of purge pressure and intake pressure while driving.
I've never done it though.
It appears that the Liberty's fuel pressure regulator does not mechanically synchronize the intake pressure.
There may be electrical synchronization, but I have never investigated it that deeply.

In Japan, a professional car mechanic tells Sunday Mechanic, "If you underestimate the canister, you'll end up with a hole in the fuel tank."

I've never heard of this actually happening, but
Canister effects often lurk behind mysterious phenomena.

Has anyone replaced it?

In Japan, it is a legal inspection item, but too many Sunday mechanics ignore it.
The 2007 model is definitely at the end of its lifespan.