'89 XJ shuts off after starting

I've been struggling with an interesting "no-run" problem for some time now. I have an '89 4.0 XJ. The Jeep starts fine, but turns right off (seems to be after it starts to come down from the Renix high rev). I believe the problem to be air-related, because I've discovered by accident that I can make the Jeep run if I pull the vacuum line off the MAP sensor and start that way. It will run with a very rich mixture, sucking the air through the disconnected vacuum line. I have done the following:

1. Pulled off T-body and cleaned all cavities within with carb cleaner.
2. Have new NAPA IAC (no effect), cleaned contacts on the plug.
3. Have new NAPA MAP Sensor (no effect).
4. Cleaned and replaced some vacuum lines.
5. Replaced CPS (noticed lower cranking to start time)
6. Cleaned and replaced ground/hot cables.

I'm getting a DRB II Scan tool to check if there are any codes set by the computer. In the meanwhile, does anyone have ideas about what else I could try to get it to stay running?

Thanks

If mine was doing that, I'd jump the ballast resistor just to eliminate that. Check the front harness from the front of the fuel rail down to the O2 sensor, knock sensor and engine temp. sender. Swap the relays around and make sure the EGR isn't hung open. Then work from there. The maybe check the TPS (the half closest to the TB for the auto) for supply voltage. Check the supply voltage to the MAP.
For every sensor I've ever changed, I've found 2 or 3 things (actually conservative, probably closer to 5-6) wrong in the connectors or the wiring.

yep, check the ballast resistor on the driver's side inner fender. there are 2 fuel circuits, one for startup and one after startup, the one after startup goes through that ballast resistor.

8Mud, thank you for replying:
1) I've actually cut out the ballast resistor and soldered wires together.
2) Replugged every connector around the relay area. Tested and swapped around relays.
3) Checked supply voltage to MAP: have 5V on supply and ~.57V on output (seems to be within specs according to FSM). I've turned the switch to "on," unplugged MAP sensor and checked voltage on the computer-end -- I assume it remembers the last value supplied by the sensor.

A) Will check voltage on TPS and take off the EGR Valve to see if it is clogged

serge said:

A) Will check voltage on TPS and take off the EGR Valve to see if it is clogged

Click to expand...

When you check the TPS, check the three plug connector 4.65 volts in and .6 volts out. You adjust the TPS at the four pin connector, but test for ECU motor inputs at the three pin connector.
It's hard to do, but try to get a pair of needle nose pliers on the EGR shaft and pull it out and push it in. The shaft is on a spring. Also try plugging the vacuum line just before the transducer. If there is no power to the EGR solenoid, it will default in the open position. An open EGR valve at idle will stall you right out.
Can you work the throttle a little and kind of coax it to life? Try starting it by holding the throttle plate open about the thickness of a nickle or maybe a little more.
Just thinking out loud, but when you pull the MAP plug, you throw it in open loop. The transition from open loop to closed loop happens pretty fast on my Renix and the O2 sensor takes over most all of the fuel metering chores, pretty quick. Somebody did a good post on a O2 sensor signal test a while back.
I keep a fuel pressure tester in my tool box, mine has never shown any fuel pressure problems. But I slap that sucker on there, just to eliminate that as a possibility. Like mentioned before, the start circuit and the run circuit, supply voltage to the fuel pump through two different circuits. At start it's around the relays (and ballast resistor) at run it's through the relays (and ballast resistor).

serge said:

Checked supply voltage to MAP: have 5V on supply and ~.57V on output (seems to be within specs according to FSM). I've turned the switch to "on," unplugged MAP sensor and checked voltage on the computer-end -- I assume it remembers the last value supplied by the sensor.

Click to expand...

The MAP is said to take a atmosheric pressure test as soon as the key is turned on, as a reference.
There is a second connector, between the MAP and the ECU, which is worth a look see. My local Jeep dealer had plugged it together and flattened a pin. That took awhile to find!

8Mud said:

Can you work the throttle a little and kind of coax it to life? Try starting it by holding the throttle plate open about the thickness of a nickle or maybe a little more.

Click to expand...

I think you have me onto something here -- I've depressed the gas pedal slightly: she started up and didn't shutdown. As soon as I let the pedal up, the rpms fell to ~150 and engine shut off.

8Mud said:

When you check the TPS, check the three plug connector 4.65 volts in and .6 volts out. You adjust the TPS at the four pin connector, but test for ECU motor inputs at the three pin connector.

Click to expand...

I was reading 5V reference and around ~.51V out. I got this on ECU-end of connector that was disconnected from TPS with key in run position. Tomorrow, I will try to get the readings with voltmeter plugged into the back of the connector and see if readings change as I move the throttle around.

Thanks!

8Mud said:

Can you work the throttle a little and kind of coax it to life? Try starting it by holding the throttle plate open about the thickness of a nickle or maybe a little more.

Click to expand...

I think you have me onto something here -- I've depressed the gas pedal slightly: she started up and didn't shutdown. As soon as I let the pedal up, the rpms fell to ~150 and engine shut off.

8Mud said:

When you check the TPS, check the three plug connector 4.65 volts in and .6 volts out. You adjust the TPS at the four pin connector, but test for ECU motor inputs at the three pin connector.

Click to expand...

I was reading 5V reference and around ~.51V out. I got this on ECU-end of connector that was disconnected from TPS with key in run position. Tomorrow, I will try to get the readings with voltmeter plugged into the back of the connector and see if readings change as I move the throttle around.

Thanks!

I haven't read all the post in this thread so excuse me if you've already figured this out, but I just went through this last week, My problem turned out to be cps (crank position sensor) that was after I replaced the o2, tps, fuel filter, fuel pump and finally took it to a shop and they replaced it for me. no everything is great

If you get it coaxed to life, look down the IAC air hole and watch the IAC piston as you work the throttle a little (it should move some). The O2 sensor is most of the input for the IAC, in conjunction with the TPS ( I'd guess).
Don't try the IAC out of the TB, I launched my IAC piston trying that.
Clean up your old IAC and let a drop or two of light oil run behind the piston and give it a try. The grease behind the piston (on the rod) gets old and thick.
My idle would fluctuate from near stall to around 700 RPM, it turned out to be mostly the O2 sensor. But everything from my ground connections to the cleanliness of my connectors had an influence.
On the drivers side of the TB, there is an orifice, that appears to control idle some. There is an adjustable torx headed screw, under a metal cap on the drivers side of the TB. The orifice on mine was completely plugged. I counted the turns and removed the screw, cleaned everything out and then reinstalled it.
I also adjusted the throttle plate stop, so the throttle plate stays open a crack (10 thousandths or so). Then readjusted my TPS. As a little insurance, if the orifice gets plugged up again. My absolute minimum idle is around 400 RPM, on a hot motor. Though it does change some depending on the engine temp. and MAT inputs. I had my idle set up higher, but noticed a decided reduction in gas mileage.
Have you tested your MAT yet? Have you done a resistance test on the CPS to ECU wires? I found two loose tabs in my ECU connectors, which can't be good.

az4wheeler2002 said:

I haven't read all the post in this thread so excuse me if you've already figured this out, but I just went through this last week, My problem turned out to be cps (crank position sensor) that was after I replaced the o2, tps, fuel filter, fuel pump and finally took it to a shop and they replaced it for me. no everything is great

Click to expand...

CPS was one of the first things I've replaced. My FSM had a resistance check procedure -- and old CPS passed it. But then I found this great website http://www.lunghd.com/Tech_Articles/Engine/Basic_Sensors_Diagnostics.htm that had me measuring voltage, while cranking the engine. Well, I did and saw no voltage spikes and decided to replace the CPS. After I've gone through the pain of replacing the CPS (makes you think why wouldn't they build a little hatch inside the Jeep to access it). I realized I should have used an analog voltmeter, not multimeter. But, it's done and now I have a new CPS sensor in the Jeep.

8Mud said:

If you get it coaxed to life, look down the IAC air hole and watch the IAC piston as you work the throttle a little (it should move some).

Click to expand...

IAC I've replaced, since I didn't know how old it was -- got the Jeep at 90K and added another 50 to it. Noticed also, that contacts needed major cleaning on the plug going into the IAC.

8Mud said:

The O2 sensor is most of the input for the IAC, in conjunction with the TPS ( I'd guess).

Click to expand...

Also, replaced with Bosch unit. The old one had open circuit. :-|

8Mud said:

On the drivers side of the TB, there is an orifice, that appears to control idle some. There is an adjustable torx headed screw, under a metal cap on the drivers side of the TB. The orifice on mine was completely plugged. I counted the turns and removed the screw, cleaned everything out and then reinstalled it.

Click to expand...

Hmm, will have to check that out.

8Mud said:

Have you tested your MAT yet? Have you done a resistance test on the CPS to ECU wires? I found two loose tabs in my ECU connectors, which can't be good.

Click to expand...

MAT and CPS checked out OK. On a cool engine, they actually show same resistance to within Ohms. How do I check the resistance of CPS to ECU wires? One contact on the CPS connector --> another on the ECU connector?

serge said:

I was reading 5V reference and around ~.51V out. I got this on ECU-end of connector that was disconnected from TPS with key in run position. Tomorrow, I will try to get the readings with voltmeter plugged into the back of the connector and see if readings change as I move the throttle around.

Click to expand...

Well, I've plugged the leads into the back of the TPS connector and got the readings. With key in "on" position, I've observed stable .51V on the C wire. By moving the throttle plate around, that value changed pretty smoothly to ~4.5 on WOT. That seems to indicate that TPS is functional. Here is the interesting part -- when started (playing magic with the gas pedal), the voltage jumps to .67V and stays there? Does this mean that ECU detects incorrect reading from TPS and overrides it with a default value? Also, which voltage is default for TPS (ECU end) at idle? 8Mud indicated .6, alldata says .8 and I can't find it in the FSM... Also, does any one have a good guide for adjusting the TPS?

The out voltage of the three pin connector part of the TPS is supposed to be 17-18% of the input voltage. A typical input voltage, would be around 4.65 volts. output near 0.80. The TPS is normally (with the auto trans) adjusted at the four pin connector. Input voltage around 4.65 (book calls for 5 volts but it isn't often there) and an output of 82-83% of the input voltage or 3.8-3.85 volts. And the readings at the three pin connector rarely end up ideal, often around .65 volts or so.
The TPS with the auto trans is two halves and is adjusted at the tranny (TCU) half. The ECU seems to make adjustments for initial low input voltage from it's half. Another oddity is the ECU input voltage to the TPS and the TCU input voltage to the TPS are often somewhat different, so perfect results are uncommon.
I checked all of my CPS wires, TPS wires and others for line loss, due to resistance. From the sensor to the ECU, there numerous cable connectors that can cause trouble. Many of the voltages are fractions of a volt, doesn't take much resistance to screw up the values.
The CPS has a history of signal loss and trouble with weak inputs to the ECU. I checked resistance between CPS connector and the ECU connector. I jumped the connector at the CPS and checked ohms at the D-1 and C-1 pins on the ECU connectors and found a resistance I finally traced down to a connector. I also took an ohm reading of the CPS, plugged it together and took an ohm reading at the ECU CPS connection pins, just to see if it was the same.
You said your IAC is new , but is it working? :exclamati If the piston isn't moving as the throttle is moved with the engine running, it probably isn't. The IAC typically parks itself partially open when the motor is shut off or the piston isn't seated all the way into the seat. Try a little experiment, unhook the 3/8" vacuum line from the top (rear quarter) of the intake and have someone start the engine, try to control the idle speed, by how much of the hole you cover with your finger. With that line removed and the hole completely open, it will typically idle at around 3000 RPM.

Clear as Mud?

There have been a bunch of TPS adjustment write ups, it is a whole lot easier to do, than to explain.

My 89 XJ had the same identical problem. It wound up being one of the fusible links located near the battery. I bet that's what it is, as you have a start condition while in the crank mode then the fuel pump cuts out when the ignition is in the run mode due to the fuel pump fusible link being blown. A quick way to trouble shoot a fusible link is to gently see if they stretch. If they do stretch, then it's bad and must be replaced.
Good luck, please let us know what it is.
Adam

albush1 said:

My 89 XJ had the same identical problem. It wound up being one of the fusible links located near the battery. I bet that's what it is, as you have a start condition while in the crank mode then the fuel pump cuts out when the ignition is in the run mode due to the fuel pump fusible link being blown. A quick way to trouble shoot a fusible link is to gently see if they stretch. If they do stretch, then it's bad and must be replaced.
Good luck, please let us know what it is.
Adam

Click to expand...

Pretty much the same thing crossed my mind, but I was thinking of an interuption in the same circuit, in another location.
Your suggestion is probably the next on the check list. It takes awhile to check it all, been there done that. I'm on my third ohm meter battery :laugh3:

8Mud said:

Pretty much the same thing crossed my mind, but I was thinking of an interuption in the same circuit, in another location.
Your suggestion is probably the next on the check list. It takes awhile to check it all, been there done that. I'm on my third ohm meter battery :laugh3:

Click to expand...

I don't believe this to be a problem with fuel supply -- I can reliably have the Jeep running by pulling the vacuum line off the MAP Sensor. I've adjusted the TPS using the Renix manual (did it on the tranny connector). As a verification, checked the output on the ECU - now have a perfect .80 on terminal C (was .51 before). This is with key in Run and engine off. When I start the engine, it will shutdown again. Now here is the interesting part, when I pull the vacuum line off the MAP sensor and make the engine run, the voltage on the TPS ECU connector terminal C jumps to .91. The throttle is still closed. When I shut the engine off, voltage returns to .8 volts. Is this how it is supposed to work? Is the ECU going mad?

Thats only because your engine is running and your battery voltage is up.
If your fuel pressure is low, unpluging the MAP could keep it running.
Could be a bad pump or a pluged filter letting enough fuel through to start but not enough to reu.

serge said:

I don't believe this to be a problem with fuel supply -- I can reliably have the Jeep running by pulling the vacuum line off the MAP Sensor. I've adjusted the TPS using the Renix manual (did it on the tranny connector). As a verification, checked the output on the ECU - now have a perfect .80 on terminal C (was .51 before). This is with key in Run and engine off. When I start the engine, it will shutdown again. Now here is the interesting part, when I pull the vacuum line off the MAP sensor and make the engine run, the voltage on the TPS ECU connector terminal C jumps to .91. The throttle is still closed. When I shut the engine off, voltage returns to .8 volts. Is this how it is supposed to work? Is the ECU going mad?

Click to expand...

Langer1 was posting while I was writing, we are thinking along the same lines.

I never finished tracing the TPS and MAP circuit, but did notice they share some wiring.
Have you blown through the MAP vacuum feed line with it still hooked to the TB?
You may just be defaulting into open loop, by pulling the MAP vacuum line. I know pulling the electrical plug does this.
It's possible to have good pump pressure at start up, but have it fall almost too low to run the motor after the starter is released. Pulling the MAP vacuum line, could be widening the pulse enough to run the motor. Could be it just throws the system into open loop. Open loop operation, requires fewer sensors.
The fuel pump and the power for the injectors goes through the pump relay. Mentioned in another thread, is a low voltage condition to the pump, from a sorry splice at the fuel pump relay feed (which is bypassed during start). I'd leave a fuel pressure tester on there and keep an eye on it anytime I got it running.
I'm just guessing here, but do a feed voltage test on the MAP and a function check. The TPS and the MAP work together. I saw one MAP that had a 12 volt feed, that sure enough messed up everything.
If your IAC is stuck closed (or faulty for whatever reason) it will (might) start and die. It might start and die if the EGR is hung open. It might start and die if the fuel pump is starved for amperage. Volts X AMPS = watts, Watts is what you need for the pump to supply enough fuel. Voltage tests will fool you. Put a volt meter in the fuel supply circuit, watch the volts, while starting and again while you release the starter. Your fuel pump, could be working, but working poorly.
Make sure you are getting a steady supply of fuel during start and while running.
Make sure you are getting a steady supply of voltage to the pump and injector circuit from the pump relay.
Make sure you are getting enough air during/after start.
Make sure the EGR isn't dumping exhaust gas into your intake at/or near idle. If you can work the throttle and baby it till you get above 1500 RPM or so and the motor smooths out, the EGR won't affect things much. If the wire is disconnected from the EGR solenoid, the EGR my open at idle and cause a stall.
Next step is to check most all of the sensors.
Next step is to figure out which sensor is screwing up the works.
Did you ever try to start it with a vacuum line (or two) disconnected just to see if some extra air helped the whole process?

langer1 said:

If your fuel pressure is low, unpluging the MAP could keep it running. Could be a bad pump or a pluged filter letting enough fuel through to start but not enough to reu.

Click to expand...

Langer1: "by unplugging MAP" you mean the electrical connector? Actually, that has to stay on the MAP, otherwise engine will shut off. It is the vacuum line that has to be unplugged -- the engine sucks the air through it ...