'92 ECU in '94 XJ...weird stuff happening.

My stock '94 ECU fried after 180k flawless miles...after mounting it horizontally with the coilover swap, it filled up with water. Oops.

So I grabbed the ECU out of my '92 parts XJ, and everything worked fine off the bat. After a few trips it threw a trouble code for 'ballast resistor relay circuit open/shorted', not surprising because I don't have a ballast resistor, as they were eliminated prior to '94.

Lately it's been doing this weird thing where if I turn the key on, it'll run the aux fan immediately, regardless of the engine temp or A/C switch position. If that fan is running, it won't give power to the fuel pump or ignition coil. It's almost as if the auto-shutdown relay has been tripped due to engine temps being too high, except I checked the shutdown relay and the temp sender at the front of the block, and both are OK.

If I cyle the key on and off enough times it eventually stops turning the fan on, and when that happens, it'll fire up with no problem. Almost like the temperature is back in operating range, as described above. While this is a decent temporary solution I'd rather not rely on this to get going again in the middle of a trail.

I doubt it makes a difference, but my '94 ECU was for a 5-speed, and the '92 came from an automatic. The fact that it's currently a 4.6L stroker shouldn't matter either as I used all '94 sensors and wiring on it.

As much as I've done to this thing (see sig), I've been very fortunate in that the few OEM parts left have been pretty trouble-free, so I don't have a lot of experience tracking down engine and electrical gremlins. :dunce: Anyone ever come across this particular issue before?

I'm wondering if swapping out the ASD relay would help.

My in-cab 'kill' switch is spliced into the feed under the dash for the ASD relay. The key still overrides everything, but I have manual control over the ASD as well with the key on.

If it's doing the weird fan thing, manually cutting the ASD relay via the switch does turn the fan off, as I'd expect because it kills the ECU anyway. So the relay itself seems to be operating OK.

Maybe you need to consider adding some diodes, could be some kind of feedback causing problems.

The switch for the ASD is just a simple interrupt in the normal wire under the dash; I just cut it and put a switch in the middle. I've been running the same setup for about 14 months now and didn't have any issues until swapping in the '92 ECU...and even then, had no issues until it started with this goofy ballast resistor trouble code. What concerns me is the computer is trying to flip the ballast resistor relay on, which means it's trying to do something with a wire/pin that I either don't have at all or is a completely different circuit in a '94 harness. Weird.

What I really need to do is make it act up again and get a scanner on it, and see what it's reading from the temp sender. I'm guessing the temp sender resistance decreases as the temp goes up, so if there's a short in the signal wire somewhere it could be interpreting that as a very high temp signal...

I'm all for testing stuff, including diodes, but I'm sort of at a loss as to where I'd put 'em...

Well, I don't know what your issue is as the ballast resistor that was used in the XJ early on had NO connection to the ECU/PCM, and can be bypassed with ZERO issues. It is nothing more than a resistor inline between the fuel pump relay and the fuel pump, nothing more.

Good luck.

The ECU does a few things to the ballast resistor. On normal operation, it pulls down the ballast resistor relay so that the fuel pump is "quieter, and lasts longer". It stays in that mode for regular driving habits. There's a bypass for the ballast resistor relay that triggers when the O2 sensor detects a lean condition at full-throttle, which jumps the fuel pump supply to full voltage to ensure adequate fuel supply. I believe it also bypasses the ballast resistor at startup, normally.

Neither of which matter in my case, as my fuel pump runs at full-bore 100% of the time anyway.

I started it half a dozen times tonight with no issue. Damn these intermittent problems.

For what it's worth, when I cut out and replaced the firewall bulkhead connector on my '92 a couple months back, I also took the opportunity to rid myself of the ballast resistor (whch had split apart) by cutting off the connectors on the two wires that run to it and butt-splicing them together. Haven't noticed a change in noise, and haven't had any fault codes for it.

You said your original PCM filled with water - was the harness connector on it similarly flooded? The connector job I mentioned above came about because I had a couple wires on my bulkhead connector completely rot away recently. It'd occasionally make the entire vehicle black out - darn near left me stranded just off the Schukyll Expwy on the way home from the pick-and-pull M&G back in the spring (there was also the incident at the RC trail cutting in March - remember the "chapstick fix"?), but sometimes it would work just fine. Frustrating, to say the least, and at times downright unnerving.

Replacing the whole connector finally fixed it; before I buttoned it up, I unloaded at least half a can of dielectric grease on the back of the connectors to prevent mud/water from getting into where the wires enter on this one, so hopefully I won't ever have to do this again.

Given your '94 is almost as old as my '92, you may want to go over the firewall bulkhead and the PCM connectors - between dirt, water, and corrosion, you might find problems like I found, or perhaps shorts.

Another place to look - go over all the contacts in the PDC. After I got my bulkhead connector replaced, my A/C was intermittent. I took it to someone due to time issues (also, if it needed a charge they'd already have it), and they found oxidized contacts in one of the fuse sockets. A quick cleaning later, and I have consistent, cold A/C again. Maybe you need to clean your underhood relay and fuse sockets?

The connector wasn't flooded...basically I have the ECU mounted flat, and the connector plugs into the top of it. Given that the whole inside is potted I didn't think it'd be a big deal, but this is how we learn things.

You wouldn't have a fault code for the ballast resistor because you probably still have the relays in place, even if the resistor itself is gone. It still changes the circuit but to no effect now.

The PDC is all good but I guess I could check the bulkhead connector...to my knowledge it's never been apart since the factory put it together. I haven't had any other intermittent issues though. The plug IS a little more...'exposed' to the elements now though.

True on the ballast code.

I've replaced a failed PCM as well - they do sometimes go bad. In my case, the potting aged and cracked, and I assume moisture infiltration did the rest. I also have seen cracking on my '84 Omni's spark computer potting, used one of the varieties of GE RTV to fix it (whichever one DOESN'T give off acetic acid while curing). Seems to be working fine, but that computer's practically using discrete logic gates... :eeks1:

Both your PDC and bulkhead look quite exposed to me too - can't hurt to look. My bulkhead connector was like yours - once assembled at factory, undisturbed to the best of my knowledge after that. It may have been pulled apart in 2008 (had a windshield guy pull the whole dash to replace the air duct-cowl seal just before leaving for Moab, don't know if the fuse panel had to come out for that or not), but otherwise nobody fooled with it until I started having problems with it. It did show signs of having been splashed with mud, though, even in an entirely stock compartment with fender liners intact. Most of the wires showed either dried mud, oxidation, or both, and two (one of which was that 10-guage one in the bottom half) were so far gone the part of the pins that crimp onto the wires was completely gone.

Much as I hate to say it, unless the wires fall completely out in your hand, it may be hard to tell whether any of the wires are intermittent. It would probably be easier to go to the boneyard, hack the connector assembly off a C4C rig that's obviously never been off the street (I snagged the one off a 93 and it matched my truck right down to the wire colors), and go to town with a pair of crimpers, heat-shrink splices, and a can of dielectic grease.