Normally mine runs at 13 on the gauge. Will go up to 14+ if I've run the battery down some. From what your saying above it sounds normal so far.

 

Mine normally runs around 14 on the dash, it definitely looks lower than normal.

Okay, I have a new theory.

Every time I've replaced the alternator, it fixed the problem. So I assumed I made the right choice by replacing the alternator. 4 times in a row. This last time around I pulled out the alternator and got it tested, and it passed. So I put it back in, and the problem went away.

Every time I fixed the problem, the battery was disconnected for some period of time and reconnected.

My theory is that the internal regulator is failing due to some issue with the PCM. By disconnecting the battery it forces the PCM to reboot, and the problem is temporarily resolved.

My new plan is to keep driving the vehicle until the problem happens again. When it does, I'll pull over and disconnect the battery for a minute and reconnect it. If that fixes it, I can be pretty certain that it's an issue with the PCM.

 

I like your creativity, and it's worth doing, however there might be a gap. It would be just as likely that the computer is receiving a false reading from something like that ambient temp sensor right (what we've been calling the battery temp sensor)? That a PCM reset, and total system power reset could be causing to self-correct for a while.

A bad battery temperature sensor hasn't been ruled out at this point yet right? (can't remember).

One thing that stinks is that the ambient temperature sensor data is not in the OBD2 for the 96. So you can't watch a scanner while you drive to see if it has weird or missing readings during (or preceding) the problem.

I assume an actual battery temperature problem has been ruled out? due to this happening over multiple batteries? I assume you popped the hood and didn't find anything incredibly hot right?

One strange thing is that 14+ volts is pretty high for a 96. And if that is what you are seeing at the gauge, it is likely even a little bit higher under the hood. Normally you'd only see that if it was actively charging a less than fully charged battery. Does it stay that way all the time when the problem is not happening? Or does it eventually drop down into the 13.5-13.8 range?

 

When I say "battery temp sensor" I'm talking about the spring-loaded thermocouple that's in the battery tray.

I've unplugged it and measured the impedance which is about 13.5K ohms at 70 degrees Fahrenheit. A little high, but that should just make the charge rate go higher, right? Or is it lower? I forget. I've heard of people putting fixed 50K ohm resistors in there to bypass the sensor, so you'd think 13.5K would be fine.

I also tried disconnecting the battery temp sensor while the problem occurred, but that didn't change anything. But who knows what the PCM will do when it sees infinite ohms. Also, I challenge your assertion about OBDII and the battery temp sensor. When I unplugged it, I got a check engine light.

Anyway, regardless of what's causing the regulator to glitch...if that IS what's happening, I can fix it with an external regulator. Might not be optimal, but it should work. And for a 22 year old vehicle, that's good enough for me.

 

Same sensor. I was just trying to use the term that the FSM uses (I think they refer to it as the ambient temperature sensor). Edit *** Correcting myself here - just checked the FSM and they call it "Battery Temperature Sensor". Sorry.

You are correct on the OBD2. I didn't mean that the computer didn't monitor, and throw codes when it is missing. I just meant that the computer doesn't include that information in the OBD2 data stream. So it doesn't appear to be a value that can be monitored in real-time using a scanner. At least it is not present as a data item on my setup. I can monitor engine coolant temp and the intake air temp sensors, but the ambient temp sensor reading isn't there (I believe there is a PID for it on the list, but the computer isn't providing a value).

I wasn't trying to be difficult or drag things out. I think your plan is a good one.

 

You're kidding, right? You're a total stranger and have taken time out of your day to help me multiple times over the past week. That's more than anyone else has done. I believe I'm on the right path because of the help you provided me. So, thank you.

 

Appreciated. I just know that even those with good intentions (me) can be frustrating when every little detail is repeated, or brought back up. Its rare that someone posting here takes the time to actually test things and provide data (you), so I was taking a special interest in helping as much as I could on this one.

Also, as I mentioned, I just happened to have been going through some very similar, but not exactly the same, troubleshooting on my 96 right at the same time.

 

I had my battery temp sensor out a couple of weeks ago to measure, and the resistance measurement was in spec. I warmed the sensor up with my hand while measuring, and the resistance decreased. So, as the temp. increases the resistance decreases, and the alternator charge voltage also decreases.

Your sensor resistance is reading higher than normal so, if my observations above are true, yours should be charging at a higher voltage than normal. In 90 degree temps, mine typically charges in the mid-13v range, and doesn't reach 14v until ambient (battery) temperatures are down into the low 70's.

 

LOL sorry couldn't help myself.

 

If you took the alt out and put it back in (...the fourth alternator...) and it worked I would be looking at the connections to the alt. and ground, the little plastic piece with 2 fields and the ground. Sounds like a bad field terminal connection, those super small ones. Are they nice and shiny because you already polished them up? How are you tightening them , with pliers or with the proper sized 1/4" ratchet and socket?

Also don't waste time on the bat (ambient temp sensor) because it will NOT cause the alt to stop charging. Also 13..k is good. If you replace it with a 50k resistor you will reset the pcm and fail inspection because the obd2 modules wont run when it's telling them it's 50 below out.

 

The saga continues....

I've been driving the jeep daily since the last post. Everything has been fine until yesterday morning. Start the jeep, no charge (running at around 12V).

So this is my chance to test me theory. I shut off the jeep, disconnect the battery and wait one minute. Reconnect the battery, start it up, still 12V.

At this point I'm thinking the voltage regulator is just bad. So I call my dad up and it just so happens that he has an external regulator. I install it, and still only have 12V. Now I'm questioning whether or not this external regulator is working, so I disconnect it and hook the field terminals directly up to the + and - of the battery. Still 12V. Now all evidence points to a bad alternator.

This is the SAME alternator that passed the test at Advance Auto a few weeks ago. I pulled it out, took it to the same Advance Auto. This time it FAILED.

So tomorrow I'm going to take the alternator back to the shop and see if they will replace it under warranty.

The next one will be alternator #5 in 2 1/2 years. Wish me luck.

 

Took it back to the alternator shop, they confirmed that it's bad, but they can't figure out why. Getting a new one under warranty tomorrow.

 

Sounds like somewhere you have a constant drain. Alternator is probably working harder than it should, making it fail faster than normal. Unfortunately the process to figure that out is lengthy. Have to go through and check all components. I would assume this isn't a bad ground but a hot wire grounded somewhere or not fully connected. Not fully connected increases resistance, thereby increasing amperage

 

I already checked that awhile back and didn't find anything with the ignition in the "on" position (although engine wasn't running). It's only drawing something like 50mA if I recall correctly.

 

Curious, where did you measure that at? Your fuel pump alone should read more than that on start-up (or at least when you first switch ignition to "on")