2004 Liberty, signs of a failing PCM or bad grounds?

[Retroshaft]

Hey all,

My 2004 Liberty Gas A/T stopped working out of the blue one morning. It started for half a second then died, wouldn't even crank anymore. Then I tried again a few hours later and then it would sometime crank (but no start) or not crank at all. I tried pulling out a potential code error from the dashboard by doing the quick 3 key turns to on trick, which usually works fine, but this time all I got was the dashboard light flicking on/off 3 times afterwards.

Battery is fine, starter relay is fine, shorting the starter motor directly to the battery make it crank, the engine frame is to ground, also I do have half the gas tank full just in case...

I thought maybe the starting condition weren't met since sometime(!) it cranks but sometime it doesn't... So I pulled up the wiring diagram from a 2003 Liberty and I checked the voltages around the Starter Relay. I found that while trying to start, I get 0.5V on pin 86 and no potential difference in relation to either the positive or negative side of the battery on pin 85 (see attached file!). From my understanding, if everything was fine I should have gotten +12v to ground on pin 86 and +12v to battery positive on pin 85 correct? Is this a sign that something's wrong with the PCM or the park/neutral sensor maybe? Or maybe I got a bad ground somewhere, and in that case, how do I go on checking if that's the case? Any tips on what I should check from there?

 

[FCSSensai]

I'd focus on fixing the no start before worrying about the intermittent no crank.
What keeps a car running is fuel, ignition/spark, and air.
I'd check things fuel related first, personally.

 

[derekj]

Were you checking voltages at the relay with it plugged in or did you have it out? I would pull the relay out and make sure you have 12 volts at pin 86 and a good ground at pin 85 - the bottom of the picture is cut off but I assume after it goes through the range switch it goes to ground? I like to load test the circuits with a old sealed beam headlamp - shows that you have good current flow in the circuit. If you jump pins 30 and 87 at the relay does the engine crank over?

 

[Retroshaft]

Were you checking voltages at the relay with it plugged in or did you have it out? I would pull the relay out and make sure you have 12 volts at pin 86 and a good ground at pin 85 - the bottom of the picture is cut off but I assume after it goes through the range switch it goes to ground? I like to load test the circuits with a old sealed beam headlamp - shows that you have good current flow in the circuit. If you jump pins 30 and 87 at the relay does the engine crank over?

I checked all voltages without the relay in, so I indeed do NOT have 12v at pin 86 and ground at pin 85. It doesn't explicitly shows the park sensor going to ground when OK so maybe it doesn't... I didn't try to directly short 30 and 87, but testing on pin 30 I get 12v, and I tried putting 87 directly to the battery and it did crank over without fuzz. I'll try shorting them directly as soon as I have some free time!

Edit: Just tried shorting 30 and 87, engine crank but nothing else.

 

[Ksat]

From the diagram, it looks like you're supposed to get ground on 85 and power on 86 (the latter when the key is in start position). That would cause the starter motor relay to energize, bonding pins 30 and 87 together, which would send power on down to the starter solenoid.

 

[FCSSensai]

From the diagram, it looks like you're supposed to get ground on 85 and power on 86 (the latter when the key is in start position). That would cause the starter motor relay to energize, bonding pins 30 and 87 together, which would send power on down to the starter solenoid.

Regardless, if he's shorting it and getting the starter to crank... there's no reason to address that part further imho until the major "not running" problem is resolved. It likely is just not wanting to start because it knows it won't.

 

[Retroshaft]

Update: I checked my voltages around again today and surprise, now I get 12v on pin 86 on start... The only thing stopping it from cranking is no ground on pin 85 which mean the PCM' seeing something it doesn't like and is telling me to F- off I guess. Now without being able to get any diagnostic code I don't have much to narrow down the issue...

 

[Ksat]

Maybe there's a problem with the electrical part of the ignition switch? Also, are you sure the key is turning fully to the start position and there isn't a problem with the ignition actuator pin?

 

[blackhawk]

Well, give the kj what it needs. Post a ground cable from battery and energize the 85 pin to see the reaction. Then, just start ..... You must connect all fuses and relays.

 

[blackhawk]

Here you are the starter diagram for 2004 kj a/t.

 

[blackhawk]

The last 3....
I have not way to get a capture of whole diagram without pixeling

 

[Retroshaft]

Well, give the kj what it needs. Post a ground cable from battery and energize the 85 pin to see the reaction. Then, just start ..... You must connect all fuses and relays.

Tried that today, I got a consistent crank with no engine start. Guess when I have some free time I'll check the other relays to see if they have starting condition and try to narrow down the issue this way.

Thank you very much for the wiring diagram btw, I'm using the 2003 and 2005 service manual to guide myself but really wish I could get my hand on the 2004 proper.

 

[Retroshaft]

Update: Bought a Bluetooth OBD2 reader (3210RS) and... "Could not link to vehicle". I tried the reader on another vehicle and no problem getting a reading there, so I can rule out user error. I checked the power and ground pin of the OBD2 connector and all fine. I'm starting to lean on the PCM being the issue here.

 

[Ksat]

Check for power on pin 16 and grounds on pins 4 and 5 if you didn't do that already.

Have you checked if there are any blown fuses?

 

[Retroshaft]

Check for power on pin 16 and grounds on pins 4 and 5 if you didn't do that already.

Have you checked if there are any blown fuses?

First thing I did and all fine. Also yes, all fuses are good.

I managed to find a copy of the 2004 service manual, so next I'll be checking the PCM connectors if the power/ground going to it is clean or not.

 

[Billwill]

The 3.7 Gasser KJs have their main grounding coming off the rear of the heads and connecting to the Fire-wall.
Check these for corrosion.

You can make sure the Starter motor is grounding itself to the chassis by taking a thick cable such as a Jumper cable as used for jumping a flat battery.

Just use one of these leads and clamp one side to the Negative pole on the battery. Clamp the other end of the cable to the body of the Starter motor and see if the engine turns over!

Also the Actuator Pin assembly where the ignition key plugs in has a known problem of failing or just getting too much worn out.... as Ksat has mentioned!

Remove the plastic cover over the Ignition Switch.
You will see that the switch assembly is held in place by a single s.c.r.e.w which is quite difficult to remove. Once you have pulled the assembly away from the column...check that the end of the Actuator Assembly has not bent or worn out so that it is unable to rotate the ignition switch properly. Leave the Ignition key in its normal slot so that it activates the security system and rotate the Ignition Switch itself ...mounted on a small PC board...with a flat screwdriver while in Neutral etc....hopefully the engine will turn over!
You can get a new Ignition Actuator assembly from Dorman!
This used to be a common problem of the ignition assembly falling apart!

 

[Retroshaft]

Been quite busy with work but... I found out the issue and... it's a faulty fuse...

Fuse 24 to be exact, the one going to the PCM and Fuel Pump Relay. The fuse was doing fine while the load was light, but once the ignition was on RUN, the resistance shot up like crazy. Resulting to the PCM/Fuel Pump Relay getting 12v while OFF and dropping to 4v on RUN/START... And I'm guessing since the fuse resistance wasn't constant it might explain why sometime it would crank and sometime not, since the PCM only need 6v to work. Guess I clearly understand now why you should test circuits with a load applied.