96 XJ - will start but dies right away

Ok, important stuff - 96 cherokee sport, AW4, 4x4, 4.0, 160k

Been running fine for the most part. I say most part, because if you accidentally hit the gas pedal while starting it, it would become rather hard to start. Otherwise, turn the key and she starts right up, plenty of power, no overheating, nothing to cause any concerns aside from the blasted windshield washer fluid low light that always seems to be coming on.

Today, when I go to leave work, I accidentally bumped the gas pedal, and she started her acting up crap again, except this time, I can not get it to start and stay running. Normally, I just crank it with the gas pedal floored and as soon as she starts, let off the gas, and all is fine.

I can get it to start with the pedal floored, but it will only run for a few seconds, and no way will it idle....almost like it is starving for fuel, and I can hear the fuel pump running all the time (as in it never seems to build up pressure. I also notice a lot of smoke out the tailpipe (mostly blueish), and a wierd smell from the exhaust, almost like burnt gunpowder.

I figure the fuel pump has come apart, so I have her towed home.

Put the gage on the fuel rail to check, and I get 50 psi, and the fuel pump is still no turning off. Reading while trying to crank it bounces from 40-50 psi.

Hmm, now where to go ? I have spark, but not sure if it is consistent or not though.

Anyway, looking for ideas on where to start. I was sure it was the fuel pump, but not so much now......

The "throttle position sensor" is the engine management sensor that is a bit suspect. They can be tested. Here is more.
-----------------------------------------------------------------------
The throttle position sensor is connected to the throttle shaft on the throttle body. It sends throttle valve angle information to the PCM. The PCM uses this information to determine how much fuel the engine needs. The TPS is really just a simple potentiometer with one end connected to 5 volts from the PCM and the other to ground. A third wire is connected to the PCM. As you move the accelerator pedal with your foot, the output of the TPS changes. At a closed throttle position, the output of the TPS is low, about a half a volt. As the throttle valve opens, the output increases so that, at wide open throttle, the output voltage should be above 3.9 volts. Testing can be performed with an electrical meter. Analog meter is best. You are looking for a smooth sweep of voltage throughout the entire throttle band. While slowly opening and closing the throttle, take note to the movement of the voltmeter needle. There should be a direct relationship between the needle motion to the motion of the throttle. If at anytime the needle moves abruptly or inconsistently with the movement of the throttle, the TPS is bad

You should have 5 volts going into the TPS. At idle, TPS output voltage must be greater than 200 millivolts. At wide open throttle (WOT), TPS output voltage must be less than 4.8 volts.. The best is to use an analog meter (not digital) to see if the transition from idle to WOT is smooth with no dead spots. With your meter set for volts, put the black probe on a good ground like your negative battery terminal. With the key on, engine not running, test with the red probe of your meter (install a paper clip into the back of the plug of the TPS) to see which wire has the 5 volts. One of the other wires should show .26V (or so). The other wire will be the ground and should show no voltage. Move the throttle and look for smooth meter response up to the 4.49 at WOT.

Perform the test procedure again and wiggle and/or tap on the TPS while you watch the meter. If you notice any flat spots or abrupt changes in the meter readings, replace the TPS.

The TPS is sensitive to heat, moisture and vibration leading to the failure of some units. The sensor is a sealed unit and cannot be repaired only replaced. A TPS may fail gradually leading to a number of symptoms which can include one or more of the following: -

NOTE: The throttle position sensor is also DIRECTLY involved with transmission shifting characteristics! It should be verified early in the troubleshooting process, when a transmission issue is suspected!

• Poor idle control: The TPS is used by the ECU to determine if the throttle is closed and the car should be using the Idle Air Control Valve exclusively for idle control. A fault TPS sensor can confuse the ECU causing the idle to be erratic or "hunting".
• High Idle Speed: The TPS may report faulty values causing the engine idle speed to be increased above normal. This is normally found in conjunction with a slow engine return to idle speed symptom.
• Slow engine return to idle: A failing TPS can report the minimum throttle position values incorrectly which can stop the engine entering idle mode when the throttle is closed. Normally when the throttle is closed the engine fuel injectors will be deactivated until a defined engine RPM speed is reached and the engine brought smoothly to idle speed. When failing a TPS will not report the throttle closed and fueling will continue causing the engine to return to idle very slowly.
• Engine Hesitation on Throttle Application: The TPS is also used by the ECU to determine if the driver has applied the throttle quicker than the Manifold Air Pressure sensor can read. The fueling is adjusted accordingly to cope with the sudden increase in air volume, however a faulty sensor can cause the ECU to ignore this data and the engine will "hesitate" when applying the throttle. In extreme cases with the engine at idle, a sudden application of full throttle can stall the engine.
• Engine Misfire: A fault TPS can report values outside the denied acceptable range causing the ECU to incorrectly fuel the engine. This is noticeable as a slight misfire and can trigger the misfire detection software and/or Malfunction Indicator Lamp (MIL) light on the dashboard. Extreme cases can cause excessing misfires resulting in one or more cylinders being shut down to prevent engine and catalytic converter damage.

I forgot to mention I am gonna check the TPS tomorrow. I am tired of messing with it tonight after my day at work.

But thinking about it while reading your post, the trans has been doing some wierd shifting lately, like it seems sometimes when you let off the gas just a bit, it would shift to the next gear too soon. It usually did this opn the 1-2 or 2-3 shift.

And when it acts up and is hard to start, it sometimes has a slow return to idle.

Sounds to me like you have a fuel injector stuck open dumping raw fuel into it (pump keeps running and the nasty exhaust....) . Check the oil for gasoline!!!!

Also check the vacuum line connection to the MAP sensor, for a vacuum leak. Check it good, even the bottom side. Also a major possibility for your latest symptoms.

I also notice a lot of smoke out the tailpipe (mostly blueish), and a wierd smell from the exhaust, almost like burnt gunpowder.

Click to expand...

TRCM said:

I forgot to mention I am gonna check the TPS tomorrow. I am tired of messing with it tonight after my day at work.

But thinking about it while reading your post, the trans has been doing some wierd shifting lately, like it seems sometimes when you let off the gas just a bit, it would shift to the next gear too soon. It usually did this opn the 1-2 or 2-3 shift.

And when it acts up and is hard to start, it sometimes has a slow return to idle.

Click to expand...

Sounds exactly like the TPS problems I'd been tormented with in the past. Funky shifting, random torque converter unlocking while cruising, unintended acceleration while stopped at red lights, surging and stalling at idle- just some of the fun that awaits! It all started after I had the factory recall done, that included a re-flash of the ECU. Which, by the way, the contents of which are totally unknown to dealer techs. And, try to find an un-molested '96 ECU at the u-pull-it yards...They're all gone!

That being said, go to a u-pull-it yard and stock up on previously loved TPS units. '91 to '96 is the same part #, including '93 to '96 ZJ 4.0s. Beware the late '96 'Classic' model, as they may have the '97 wiring harness, which has different plugs on various sensors, including the TPS. I've had good luck with real TPS's, not so much with aftermarket pieces. Good luck.

Ok, here's what I got.....

I can not get any codes to show using the check engine light. I tried the on-off-on-off-on thing in many different ways, and I got nothing.....period. No 12, no 55, and nothing in between.

All fuel injectors read 12.1 ohms (did not backprobe them, just disconnected the wires and read across the injector spades)

Coolant & air temp sensors - O/L (outside air temp here is 74)

MAP sensor - 4.7v, no leaks (tested it with a hand vacuum pump)

CPS - reads O/L (digital meter, so no scale for reference)

TPS - idle is .7v, wot is 4v


Pulled the plugs for #2, #3, & #4 cylinders. All are black & sooty (kinda expected from the way it was trying to run yesterday), and 1 seemed wet & oily (did not smell like gas) on the threads, but not the electrode itself, altho that plugs electrode was shinier than the others

Oil on dipstick is normal for level and smell (no obvious gas smell), and it's consistency seems ok, but I do need to change it as it's about time

No obvious signs of problems under the distributor cap, altho I will probably replace it and the wires since 2 of them came apart @ the plugs from the rubber sticking to the plugs themselves

I have a Renix, but had a similar issue, would fire up then die right after. If I held the key between run/start it would stay running.
I have since by-passed the ign switch and all is well. So ign. may be a culprit as well.

My 87 suddenly decided to stop running, and I had very similar issues to yours, most of them anyway, and the rotor and cap were just on their last legs.

We tend to assume that by the time people are digging around the sensors issues, they covered the plugs, wires, rotor and cap issues, and checked to see if the laptop is plugged in, LOL.

You may just need a tune up for the most part? Also might have leaking valve seals, leaking oil into the head onto the plugs? I had to replace them on two of my cylinders recently. Only got 289,000 miles out the old ones, LOL.

Does the on-off-on trick code check trick work on OBD-II 96 models?

What does O/L mean on the meter data posts?

O/L means Out of Limits....IE, open on a resistance test

All I know is Google and everyone here says it works (the key trick) on 96, but it don't work on mine.

I'd have thunk that if a tuneup was the issue, I'd have had more problems in the past. Everytime I've had a problem like that cause a no start, it had been giving me a LOT of little problems for awhile. This one has been running fine.

I am gonna replace the TPS, plugs, wires, cap & rotor, as well as oil/filter, and try again to see where I sit.

All the readings seemed ok, except the TPS, so it's getting replaced.

TRCM said:

O/L means Out of Limits....IE, open on a resistance test

Click to expand...

If that is correct, and the meter was set up right, right range????? and you had contact on the probes????, then two temp sensors are toast then, but that sounds ODD!!!

I am not a CPS test expert on the 96, but that O/L sounds like a bad CPS!!!!

Was the meter set to ohms or volts during those tests? Vehicle Power on or off?

Only meter I have is digital...no range to select

Set to ohms/resistance

yes, contact on the probes for the air/temp sensor, but I don't have a chart to check against for ohm reading @ actual air temp

the numbers I had for the CPS were open on 1k-10k range is good

vehicle power on or off didn't matter, as all were taken on the sensor itself, not the harness (well, except the CPS)

You should have gotten a number reading on the two temperature sensors at any temperature, that would be in range of the meter, so they must be bad. Still odd for two to be bad.

I think the same is true of the CPS, but one of the other HO experts should confirm that.

These two comments are confusing, and possibly contradictory?

Only meter I have is digital...no range to select

the numbers I had for the CPS were open on 1k-10k range is good

Click to expand...

A couple of questions? Do you get a CEL at all, does it ever light up? If not you likely have a sensor short, that is sucking down the sensor voltage.

I've had a few on my 96. the CPS was shorting, but tested good. It was shorting on the non test poles. I was getting no CEL at all when I turned the key to run. Another time my rear O2 sensor wire came loose and cooked on the exhaust. And if somebody got lazy and forgot to tie down the CPS wire to the hold down, it can flop over and cook on the exhaust manifold. Many of the sensors share the same power supply, one shorts and it throws others out of range and many times affects the CEL function.

I was having charging/electrical issues on two occasions. First was my alternator was going out, slowly loosing voltage and my IAC would act up. I could start it cold by giving 1/8th pedal, but had to constantly keep a foot on the gas pedal and work it a little to keep it going, it would die at idle. After the motor warmed up the IAC would work. It seemed my IAC would stick closed and only open after it heated up, likely grease and/or thick oil. Pretty much the same thing but a little different, my battery was failing and most every time I'd let my foot off the accelerator pedal the motor would die. My guess is the IAC acts up when it is dirty and not getting enough voltage/power so it works reliably. A sticky IAC won't give you a code.

The basics, you need fuel, spark and air. Something is going to happen if they are all present. Though they have to be in the proper proportions, in a rather broad range.

If you are in fact getting no 12 and 55 code and the CEL is lighting up when the key is in the run position, I don't have a clue what is wrong. If the CEL never lights, the bub is burnt out or you may have a sensor voltage supply circuit short someplace. It takes a certain rhythm to get the CEL to blink codes, I've screwed it up. Not too fast and not to slow, it may take a few tries.

My 96 is OBD 1 (transition), I've heard the difference in the two systems is most apparent at the gauges. Digital gauges means OBD 2 and the check engine light function is changed to a display, that displays "P"codes. Just what I've read, I have no first hand knowledge.

When my CPS was going it was a hit or miss type of thing. Sometimes it would run fine, then miss and almost die while driving along. Sometimes it didn't want to start. Sometimes it wouldn't idle and died. Kind of a craps shoot every time I tried to go someplace. The CPS always tested good, by the book.

Edit: Had another thought, my valve cover was leaking, my CPS was covered in oil. Motor was acting like the CPS was bad. I cleaned off the CPS, replaced the gasket and and all is well for the last few years. I really have no idea why hot oil would affect the CPS, but the empirical (evidence)) results seem to indicate it can and does.

Ecomike said:

You should have gotten a number reading on the two temperature sensors at any temperature, that would be in range of the meter, so they must be bad. Still odd for two to be bad.

I think the same is true of the CPS, but one of the other HO experts should confirm that.

These two comments are confusing, and possibly contradictory?

Click to expand...

Nope, but possibly confusing if you didn't read that I don't have a selectable range on my meter.,...the specs I was given said open in the 1k to 10k range, and I got an open.......but no range to verify with

8Mud said:

A couple of questions? Do you get a CEL at all, does it ever light up? If not you likely have a sensor short, that is sucking down the sensor voltage. Yes, but only when starting the vehicle.


I've had a few on my 96. the CPS was shorting, but tested good. It was shorting on the non test poles. I was getting no CEL at all when I turned the key to run. Another time my rear O2 sensor wire came loose and cooked on the exhaust. And if somebody got lazy and forgot to tie down the CPS wire to the hold down, it can flop over and cook on the exhaust manifold. Many of the sensors share the same power supply, one shorts and it throws others out of range and many times affects the CEL function. CPS wire harness is in the loom and not flopping around


I was having charging/electrical issues on two occasions. First was my alternator was going out, slowly loosing voltage and my IAC would act up. I could start it cold by giving 1/8th pedal, but had to constantly keep a foot on the gas pedal and work it a little to keep it going, it would die at idle. After the motor warmed up the IAC would work. It seemed my IAC would stick closed and only open after it heated up, likely grease and/or thick oil. Pretty much the same thing but a little different, my battery was failing and most every time I'd let my foot off the accelerator pedal the motor would die. My guess is the IAC acts up when it is dirty and not getting enough voltage/power so it works reliably. A sticky IAC won't give you a code. No charging or battery issues


The basics, you need fuel, spark and air. Something is going to happen if they are all present. Though they have to be in the proper proportions, in a rather broad range.


If you are in fact getting no 12 and 55 code and the CEL is lighting up when the key is in the run position, I don't have a clue what is wrong. If the CEL never lights, the bub is burnt out or you may have a sensor voltage supply circuit short someplace. It takes a certain rhythm to get the CEL to blink codes, I've screwed it up. Not too fast and not to slow, it may take a few tries. I get no light, and I tried it fast, super fast, slow, super slow, and all ranges I could think of in between. I spent over an hour trying to get the codes, but nothing.....


My 96 is OBD 1 (transition), I've heard the difference in the two systems is most apparent at the gauges. Digital gauges means OBD 2 and the check engine light function is changed to a display, that displays "P"codes. Just what I've read, I have no first hand knowledge. I do not have digital gages, and I do have a check enegine light. I know AZ has pulled the codes for me in the past.


When my CPS was going it was a hit or miss type of thing. Sometimes it would run fine, then miss and almost die while driving along. Sometimes it didn't want to start. Sometimes it wouldn't idle and died. Kind of a craps shoot every time I tried to go someplace. The CPS always tested good, by the book.


Edit: Had another thought, my valve cover was leaking, my CPS was covered in oil. Motor was acting like the CPS was bad. I cleaned off the CPS, replaced the gasket and and all is well for the last few years. I really have no idea why hot oil would affect the CPS, but the empirical (evidence)) results seem to indicate it can and does. No leaks in that area, and none to speak of really anywhere

Click to expand...

When you turn the key to the run position, before start, you should get a CEL (check bulb function) every time. If not, it could be a short in the 5 volt supply circuit, one of your sensors is likely shorting out.
I usually check the 5 volt supply voltage at the cam position sensor, just because it is easy to get at.
I don't know for sure, but I believe there are two low voltage ( sensor supply) circuits, but they seem to be either the same or interconnected inside the ECM.

The only time I was unable to pull the MIL codes (CEL) was when I had a shorted CPS. The CEL refused to light when the key was in the run position. The CPS tested good, but when I unplugged the CPS the CEL came back to life. The CPS tested good by the book, but was bad.

I think we are having some minor communication problems here. What I've experienced may or may not be relevant. Sounds similar. but not exactly the same- Bare with me I'm old as dirt and not getting any brighter in my old age. But have been messing around with XJ's since 87.

no communication problem that I am aware of.....

TRCM said:

no communication problem that I am aware of.....

Click to expand...

I'm sometimes not sure if I'm communicating well, between the typos, spelling and brain farts.

Well, I just found out something that may explain why I can't get the engine codes using the Key.

My 96 has a 97 TPS on it, so it may also have a 97 computer.......which may complicate things a bunch if I have mixed & matched sensors.

It is a 10-96 build date...oh fun......................so it was a late 96 build.

TRCM said:

Well, I just found out something that may explain why I can't get the engine codes using the Key.

My 96 has a 97 TPS on it, so it may also have a 97 computer.......which may complicate things a bunch if I have mixed & matched sensors.

It is a 10-96 build date...oh fun......................so it was a late 96 build.

Click to expand...

Is it a 'Classic'? Ours, a 9-96 build, has the '97 harness under the hood, and '97 computer and plugs. By '97, they were fully committed to OBDII, therefore no OBDI code retrieval. Scanners don't cost that much, and the parts stores scan for free.

8Mud said:

I'm sometimes not sure if I'm communicating well, between the typos, spelling and brain farts.

Click to expand...

Say what? LMOA!!!!