2001 PCM/ECU?

[Root Moose]

I swear there is something goofy encoded in the maps of the 2001 (2000 as well?) Jeeps.

First the description and symptoms. TLDR;?

Case 1:

The Jeep is towing heavy, Jeep weighs ~5000 lbs loaded, 32" tires, 4.56 gears, trailer is ~3000-3500 lbs, towing in heavy grade coastal mountains at under 1000 ft altitude. All new exhaust, high flow cats at all three locations, new fuel pump, Banks Torque Tube, etc., etc.

The Jeep will occasionally attempt to overheat when going up the hills but it's weird. It "smells" like a software fault to me.

Once while climbing at ~3200 RPM, second gear, the gauges spiked instantly to overheat and I pulled over to the side of the highway. Ambient temp was around 90ºF and humid, four lane highway across the back of Nova Scotia.

Had been climbing for several minutes (5?) and was slowing from ~65 mi/h to steady state at around 40 mi/h.

Once pulled over I let the Jeep idle for a minute and then I shut the Jeep off; waited another minute and turned it back on (no engine start) and the gauge was still pegged. Instantly turned it off again, waited less than a minute, turned it back on and the gauge was at the high side of normal like normally when pulling grades with the trailer. I restarted the Jeep and we continued the climb for another 4 minutes or so.

Case 2:

Another long pull up a similar mountain (Kelly's Mountain, NS), ambient is about 80ºF. Similar engine load conditions. Gauge is reading high normal.

Pulled off onto one of the scenic look-out sidings to take pictures... Maybe 5 minutes max. Got back in and started the Jeep (no engine start) and it is pegged to overheat. Popped the hood, waited a minute, gauge stayed where it was. Turned off the Jeep, closed the hood, started the engine and the gauge is reading high normal. Drove up the rest of the mountain (maybe three minutes) without incident.

Case 3:

Long pull around Bonne Bay, NL. This is the steepest, longest grade encountered. Similar engine/driving conditions for speed and ambient temperature. Jeep is well heat soaked after several hours of driving.

No overheat occurs, just keeps plugging up the mountain and over.

So, these overheat events... are they real? I didn't have a pyrometer with me in my boonie box (auto-x box has it), only the factory gauge available.

Is there a way to verify what is really happening, I mean reading the computer somehow?

I wish I had time to install the LM2 before our trip, that may have provided some decent hints.

Is it possible to Megasquirt these Jeeps, completely bypass the factory coding the only plug in the factory PCM on e-test day?

Am I chasing the wrong gremlin?

FWIW, cooling system appears to be fine. Newish water pump and t-stat, original 2001 radiator. Installed a transmission cooler before we left.

Maybe I should just give up and get a V8 WJ, diesel WK or a UZJ100 Land Cruiser? But I like my XJ.

 

[yossarian19]

I think the pyrometer is a smart place to start.
IIRC an 01 XJ has only one sensor for temp - it feeds the computer and the computer feeds the gauge. You could look at the computer's inputs / outputs with a scanner (if you have a smart phone, I recommend a wifi or bluetooth adapter & a $3 app) but it's only going to tell you what it's telling the gauge.
Pull the connector off the coolant temp sensor & look at the terminals. Any coolant seep or corrosion is a problem. While you've got the connector off, resistance test the sensor and compare it to expected values in the FSM (if you don't have these, PM me and I'll find the info)
I've many, many times considered getting the V8 WJ. It's honestly what I'd rather have. The XJ was paid off in 2009, though. Maybe later...
EDIT: OH! With the temp gauge pegged (and reading true), you'd expect to see the coolant reservoir damn-near-overflowing AND you'd be able to smell coolant. You'd also be hit with a helluvalotta heat when you popped the hood - more than normal. Was that the case? If not... I'd suspect the CTS even more than I do now.

 

[Root Moose]

There was no spilling coolant or coolant smell, heat didn't seem crazy hot when I opened the hood.

I'm at the top to Newfoundland right now... not sure I'll be able to get replacement parts if that ends up being an issue. I have the FSM on my iPad so I'll futz around with it tomorrow if I get a chance.

 

[Hoooper]

do the 01s have an idiot light style gauge? Some newer vehicles have a "gauge", but its really just an idiot light. It will seem normal in operation, but it pretty much only has cold, just right, and hot, so it would show normal until the engine got warm, and then it would shoot to showing hot.

I dont know if thats the case with the 01s, but seems possible. In that case, it wouldnt have to be overheating to show pegged on the gauge, but you wouldnt know unless you checked the computer reading

 

[MoparManiac]

I'd replace the coolant temp sensor first.

A Standard brand is $14 on RockAuto. Cheap money for the insurance, especially if you've never replaced it.

 

[Talyn]

do the 01s have an idiot light style gauge? Some newer vehicles have a "gauge", but its really just an idiot light.

Not really. The gauges read correct until it gets to an extreme range. Sort of like an advanced idiot light. Typically when that happens the "check gauges" light will illuminate and peg the gauge showing the problem. I found this out when my alternator took a crap and I still needed to get home on battery. Around 11 or so volts the check gauge light appeared and the volt meter went to 0 even though I still had enough of a charge to get me around for a bit.

What may be happening is that he is getting marginally hot and the PCM is doing its look at me routine. The issue after parking was most likely an engine heat soak. I don't see any possible way of the coolant temp going from a bit hot to full over heating in an instant.

One other theory is that when the system fully pressurizes, which doesn't happen for a while if using the stock style radiator cap, that coolant could be forced into a faulty(leaking) coolant sender unit.

I'd replace the coolant temp sensor first.

I would try this.