Strange/Annoying electrical problem

Lately I've been having some very strange and very annoying electrical problems with my '97 XJ.

About a month back the voltage gauge started reading 19 volts but the Jeep ran fine so I thought nothing of it. A week later it starts reading 19 volts and not wanting to shift. It slips in and out of 3rd and just randomly dives into overdrive. I unplugged the TCU and shifted it manually and all was well, I think this pretty much rules out the AW4?

My first thought was a bad battery so I had AutoZone test it, it passed, but I threw a new one in just to be safe. Fired the Jeep up and it was just as bad as before, so I just disconnected the TCU and drove it manually for a while.

Fast forward to this weekend and now the lights start flickering and I can't run my lights and the radio at the same time. During the day if the radio is on and I hit the turn signal the radio cuts out.

In the past month I've put in-
Brand new alternator
New battery
Refreshed all of the grounds from the battery to the chassis and ran new cables to the starter/alternator
Got rid of the crappy braided firewall ground for a 4 gauge cable.
Refreshed all of the grounds I could see for the PCM

To be clear, it's a 1997 XJ with an AW4 and about 165,000 miles on it.

I'm pretty fed up with this thing at this point, any help would be much appreciated!

Hmmm, interesting. Perhaps your tcu is bad? Have you checked the voltage with a meter just make sure your gauge is right?

Have you verified the 19 volts with a volt meter?

Nimlas,

It would seem that with a true 19 volts on the bus you would get many problems. You would get an MEL light I would think.

Like others have said, use a good voltmeter and check the actual bus voltage (engine running). Also do an instrument cluster self-diagnostic test, watching the voltmeter and MEL light (bulb burned out?).

(1) Begin the test with the ignition switch in the Off position.

(2) Depress the trip odometer reset button.

(3) While holding the trip odometer reset button depressed, turn the ignition switch to the On position, but do not start the engine.

(4) Release the trip odometer reset button.

(5) Compare the operation of the Voltmeter with the following:

Every two (2) seconds the voltmeter needle will go to a calibration point. They are at:

9v 9v 14v 14v 19v 19v 14v 14v 9v 9v

(6) The cluster will exit the self-diagnostic mode at the completion of the test, or if the ignition switch is turned to the Off position.

I don't have a multimeter so I haven't been able to read the actual voltage, but I highly doubt it's actually charging at 19 volts because every other day the battery is too dead to even turn the Jeep over.

Last night it got even stranger. Went to meet my parents for dinner and stopped for gas on the way there. On a whim I plugged the TCU back in and it shifted beautifully. It also charged right under 14 volts. However, the truck would not start again at the gas station, there was just enough juice in the battery for the solenoid to click. Got a jump and met my parents for dinner, same thing happened after dinner. Got a jump and drove to my girlfriends house, same thing happened as I was trying to leave.
All the while it said it was charging fine and it did shift wonderfully.

Sidewinder, I'll run the self diagnostic and report back. I've done that before to read check engine codes but since this problem hasn't thrown a code I haven't bothered.

Thanks all!

Charging a 12v battery at 19v will kill it.

The plugs on the alternator from the PCM are not for "sensing". That is for modulating the field. The temp sensor would prevent the charging system from making the battery too hot, it would not cause over charging. The PCM senses the voltage from one of its 12 + and - inputs.

While you might be right...... I have learned to never rule anything out by thinking "Oh that couldn't cause that problem"; when working on the Jeeps electrical system until I have verified and tested the connection or component thoroughly.
If the temperature sensor is not functioning properly or missing the potential is there for the computer to react incorrectly... As for the wiring that senses and excite the field... there is still the potential for incorrect operation....

I didn't say rule out the field circuit. If it is overcharging the field circuit should be checked. No wires "sense" the field, it is just modulated by the PCM. One wire is constant the other wire is PWM.

For an overcharging issue it would be a waste of time to mess with the battery sensor. It will not cause an overcharge issue. If it is faulty the battery could over heat or an under charge situation would happen. But the PCM wouldn't call for above 14v.