Misfire just after start-up then runs fine.

Gentlemen,

I'm about to lose my mind dealing with this thing.

I've had this issue for a while now but it's become worse in the last month.

This is on my wife's 98 XJ Sport 4.0.

This started happening occasionally about 6 months ago usually only on a cold start. Within the last 2 months it's started doing it every time it's started after it sits for 30 minutes or so.

The Jeep runs perfect, except for the following: When I start it up, no matter whether the engine is warm or cold as soon as I start to drive away I get plagued with the engine misfiring. The more gas I give it, the worse it seems to get, falling on it's face and barely able to accelerate. Giving it gas results in the engine backfiring. By the time I get to the stop-sign at the end of my street it sits and idles, still misfiring, but will all of the sudden "fix" itself and idle perfectly. After that happens, I know that I'm ok to drive. The car will run perfect for the rest of the trip until I stop again. I can literally watch the idle bounce around during misfire and then when it fixes itself it revs up a bit and that pins itself right around 700rpm where it then idles perfectly.

If I happen to be moving when it starts chugging/missing, I can floor it to fix the issue, but if I give it any other gas aside from WOT it backfires.

While it's missing I obviously smell fuel.

If I start it and let it sit and idle for a while it goes thru the same cycle without ever touching the throttle. Start Engine, runs fine for 30 seconds, starts misfiring, misfires for ~1 minute (While fluctuating rpm between 500rpm and 1000rpm), Jumps up to 1500rpm then settles into ~700rpm then runs perfectly.

-My Check Engine Light is not on, and I did scan for pending codes... no codes are found.
-I've replaced the plugs
-It has a new fuel filter and fuel pump.
-New CPS last year.
-I've tested every sensor that I can.

Anyone solved this issue?

New CPS has gone bad? Coil? Something to do with open/closed loop system in PCM? O2 Sensor?

I don't usually start threads like this but I've about had it.

Thanks in Advance, Andy

I've read until my mind is numb. This seems to be a common issue with 98+ Grand Cherokee 4.0s.

Are you able to scope secondary ignition waveform while it misfires?

offroadordnance said:

Are you able to scope secondary ignition waveform while it misfires?

Click to expand...

To be completely honest... I thought I knew some stuff about Cherokee's and engines in general but I have no idea what "Secondary ignition Waveform" means.

I would suspect the O2 heater circuit, fuse, relay, or sensor, or wiring. Easy to test with volt meter. The timing is the issue. If you can get a live data ODB-II scanner on it, it will tell you a lot more, open loop, fuel trims, O2 data, air and coolant temps....Just look for what changes with the missfire/no missfire...

Also look for something that shorts out when damp, and unshorts when dry (warmed up).

My bet is that you have a leaky injector. It floods one cylinder when it sits. The rest of the cylinders start the engine and with a few revs, the gas gets cleaned out and the cylinder starts to fire again.

My first thing would be to run a can of BG44K through the fuel. That will clean the crud out of the injector and most likely make the injector start working.

As a check, pull the dipstick and see if you smell any gas. If so, bingo. If you don't, you may have found it before it got to the point of washing past the rings.

I would also pull the plugs and do a reading on them. If nothing else, post up the pix.

Good Luck.

Ecomike said:

I would suspect the O2 heater circuit, fuse, relay, or sensor, or wiring. Easy to test with volt meter. The timing is the issue. If you can get a live data ODB-II scanner on it, it will tell you a lot more, open loop, fuel trims, O2 data, air and coolant temps....Just look for what changes with the missfire/no missfire...

Also look for something that shorts out when damp, and unshorts when dry (warmed up).

Click to expand...

Unfortunately, I don't own a scanner. I'll have to look in to find someone with one.

I haven't looked into the O2 circuit yet.

old_man said:

My bet is that you have a leaky injector. It floods one cylinder when it sits. The rest of the cylinders start the engine and with a few revs, the gas gets cleaned out and the cylinder starts to fire again.

My first thing would be to run a can of BG44K through the fuel. That will clean the crud out of the injector and most likely make the injector start working.

As a check, pull the dipstick and see if you smell any gas. If so, bingo. If you don't, you may have found it before it got to the point of washing past the rings.

I would also pull the plugs and do a reading on them. If nothing else, post up the pix.

Good Luck.

Click to expand...

Will do. I plan to swap the plugs out again within the week.

Check and clean the battery posts, and clamps and grounds too.

http://www.naxja.org/forum/showthread.php?t=1012701&highlight=oxygen+test

How to test O2 sensors, and wiring, etc.

If you have a laptop, there are some $20 Ebay rigs for use with smart phones and laptops for OBD-II live data viewing.

Ecomike said:

Check and clean the battery posts, and clamps and grounds too.

http://www.naxja.org/forum/showthread.php?t=1012701&highlight=oxygen+test

How to test O2 sensors, and wiring, etc.

If you have a laptop, there are some $20 Ebay rigs for use with smart phones and laptops for OBD-II live data viewing.

Click to expand...

Ok, so this one is going to have 4 wires. One wire at 12-14 volts, one wire should read .1-.9 fluctuating at idle, and the other 2 wires should read less than 1 ohm to ground?

This Jeep has an exhaust leak but its in the back of the muffler. That couldn't effect this could it? It had a leaking exhaust manifold but I fixed that when this started happening.

Any suggestion on ebay special computers?

That is correct. You would need a high impedance volt meter for .1 to .9 volt sensor tests, but the rest can be tested with a cheap $5 digital meter.

No the muffler will have no effect on the O2 sensor, unless it is periodically blocking the exhaust internally, but that is usually constant or random, and yours is neither.

I bought mine over 2 years ago, and some things have changed since then. Check the tools forum here and see if they have any opinions, and look for my old thread here on the topic.

Just search "OBD-II Ebay scanner" and look for bluetooth and USB versions. Mine is USB.

Ok, I just had another guy with the same exact issue verify that his was the O2 sensor.

I'll verify the circuit when I get home and pick up an O2 sensor on the way.

Thanks for the help, guys. I'll report back asap.

One more thing.

There should be 2 black wires.

What are the colors of the other 2 wires? I assume they're some other color than black.

I realize it should be easy to tell them apart from each other by voltage but I don't want to confuse the signal wire for one of the ground wires when testing.

1 Orange, 1 Blue, 2 Black wires? Sounds right?

O2 heating circuit malfunction would throw a code since its OBD2. A scan tool will indicate O2 voltage toggling rich to lean when monitoring live data. Most chain parts stores provide the scan tool service for free. If it runs ok at start up, gets bad with higher RPM, and goes away suddenly I'd bet a weak valve spring allowing lifter pump up when the oil is cold and thick. By the time the engine warms up and in closed loop, the oil is thin, the lifter doesn't pump up, cause misfire, and store a code. A competent diagnostic shop with an O'scope can look at the ignition coil HV waveforms to identify which cyl is misfiring in 5 minutes and save you from throwing time and money at it.

More often than not, OBD-II seems to fail to throw codes when there are O2 sensor system problems. The stores code readers to not have live sensor data.

Welp, I went out. Started testing wires.

Looking at this diagram helped absolutely none.


The black and blue wire went to the ECM as well.

None of the wire colors matched. It specifies 91-01 wires are the same color. I beg to differ.

Either way. For whatever reason with the key on and engine off I could only get 4 volts on the only wire with voltage.

That may be because of my cheesy voltmeter.

So, with high hopes I chugged along and installed the new O2 sensor. After fighting with it for 30 minutes I finally got it installed.

Started the engine and for the first time in 2 months it didn't stumble. I drove it up the road and the issue did not show itself.

HOPEFULLY, that solved the problem.

We'll see tomorrow when the wife drives it to take the kids to school.

While I was under there I noticed a pretty serious valve cover gasket leak and leaking transmission cooler line. Yay for me.

I'll report back and let you know if it's holding up.

Which wire had 4 volts?

Was the test done with the O2 sensor connected (the 4 volt reading), or disconnected?

With the sensor disconnected it should read 12-14 volts. Also it should be the fattest wire of the 4 wires. It operates an electric heater in the O2 sensor, that is needed at idle to get accurate O2 concentration data. The sensor gets too cold at idle to read accurately if that electric heater is not working.

Ecomike said:

Which wire had 4 volts?

Was the test done with the O2 sensor connected (the 4 volt reading), or disconnected?

With the sensor disconnected it should read 12-14 volts. Also it should be the fattest wire of the 4 wires. It operates an electric heater in the O2 sensor, that is needed at idle to get accurate O2 concentration data. The sensor gets too cold at idle to read accurately if that electric heater is not working.

Click to expand...

All 4 wires were the same size.

The black wire with the green tracer had 4 volts. I should have snapped a picture.

The wire had 4 volts disconnected from the sensor as well as plugged into the uninstalled new sensor.

I assume we've been discussing the upstream O2 sensor in the down pipe closest to the manifold... I didn't even look for the down stream sensor.

Lowrange2 said:

None of the wire colors matched. It specifies 91-01 wires are the same color. I beg to differ.

Click to expand...

Let me guess: Haynes or Chilton, right?

joe_peters said:

Let me guess: Haynes or Chilton, right?

Click to expand...

Chilton... Can't afford all-data.