I've had a low-RPM, high-load misfire for a while on my 1991 4.0 XJ. Replaced the cap and rotor a while ago with no improvement. Then did the O2 sensor, MAP, TPS, coolant temp sensor, intake air temp sensor, wires, plugs, and a metric ton of beer.

Decided to pull the cap off again and take a look inside and found this mess. Any ideas?

 

What did you set the plug gap to?

 

I've done 040 and 030. It's back at 40 now. Funny thing is nothing changed when I closed the gap.

 

What plugs did you install?

You should be using Champion RC9YC plugs, gapped to .035".

An optional plug is the NGK FR5. This is a hotter plug than the RC9YC.

 

I'll check on the plugs again this afternoon. Do you really think they can cause a problem like this, though? To my simple mind, it looks like something's telling it to advance or retard the hell out of the timing and it's throwing the spark at the wall instead of at the terminals...

It's weird. The jeep runs great most of the time. Starts are a little long, but it gets a consistent 17 mpg and idles smooth. It's just when you roll into the throttle (maybe more than 1/2 or 3/4 pedal) below about 2500 that it misfires. Not so bad that it quits, but just bad enough to make me want to light it on fire. It's never stalled or exhibited any "classic bad crank position sensor" type behavior...

 

A larger plug gap and high resistance will cause the build up you're seeing in the cap. The fuel we are forced to use will also cause it because it is resistant to burn and puts a lot of stress on the ignition system.

 

The funny thing is that with all my experimenting with different gaps, the problem persisted. No effect.

I'm still thinking that it looks like the spark timing is being driven waaaaay out of whack...

 

Perhaps you should stop experimenting and install the recommended plugs set at the recommended gap.

Install a matched set of distributor cap and rotor (same brand). The rotors of some manufacturers differ in the distance from the cap terminals than others and may cause problems.

The Camshaft Position Sensor determines what cylinder to fire and when. By your description of "Starts are a little long, but it gets a consistent 17 mpg and idles smooth. It's just when you roll into the throttle (maybe more than 1/2 or 3/4 pedal) below about 2500 that it misfires" sounds like the timing is a little advanced (starts long), has good mileage, and idles smooth, but misfires at acceleration.

 

I have had two sets of matched cap and rotor (came from the same box) on this machine since I bought it. And the plugs are the correct Champions set to the factory spec .035 gap.


The experimenting was simply to look for a change. I had one data point (when it came to the plug gap variable). I wanted to establish a sensitivity curve with respect to the spark plug gap. I have now determined that, within reason, the misfire exhibits zero sensitivity to spark plug gap changes. You know, science-like.

I have also experimented with changes to most of the input variables in the system (simple-minded people might call this "sensor replacement") with no change to the problem. It's not better. It's not worse. It's exactly the same. That tells me that the problem may be in the transfer functions (which live inside the computer) or in some odd-ball input error to the ignition map transfer functions (like a wiring short, which would return an incorrect input value).

It should be mentioned that the only sensor I have not replaced that has any effect on the ignition map is the crank position sensor and only because it hasn't exhibited any of the common problems associated with CPS failure. I may replace it anyway, just to eliminate the last input variable from the matrix.

I have also not changed the ignition pickup, but past experience tells me those almost never fail and when they do they will fail mechanically, not electrically. It would also be unusual for the pickup to fail in such a way as to allow normal operation at some cylinder pressure/RPM combinations but not at others. I understand that the ignition system load is affected by the cylinder pressure, but this just doesn't smell like a pickup failure. Like the CPS, I may decide to replace it too, just to eliminate the variable.

/EndScience.

 

If by "Ignition Pickup" you are referring to the Camshaft Position Sensor, it may affect ignition timing and fuel injector pulse width. The flex plate has three sets of trigger holes, each set representing two cylinders. As the flex plate rotates past the CPS the CPS tells the ECU that the two cylinders are in the position for firing. The Cam Sensor tells the ECU which cylinder is to be fired, and injected, and when. It essentially is the ignition advance device for the ignition system (and fuel injection system). The Cam Sensor is a dick to change in your vintage Heep so it may be left to last in the sensor shotgun series.

There are electrical tests that can be performed on the Cam Sensor.

 

looks ok to me

 

maybe the wrong cap for the rotor, looks like metal shavings, maybe the rotoer is hitting the cap, recheck part numbers, or possibly the distributor is "wiggling"

 

It's all carbon tracing in there. I did ping a few of the old ignition guys here (I'm trapped as an engineer for one of the Big 3) and they say that's not an uncommon look for an old ignition system. Now that it has another fresh cap and rotor on it (yes, they are a matched kit) I'll see if it does an encore performance of the Carbon Painting Roadshow.

I never noticed any excessive distributor shaft play when I changed the rotors, but I'll check it more closely for fun.

Yeah, I started looking at the "cam position sensor" and decided it was going to be a *****. I agree that leaving it for last is best...