You need to read the first post, the entire issue is detailed there. The company tested the TCU I sent them and said it was good. They replaced the electrolytic caps anyway. Didn't make a difference.

 

I re-read the OP. I'm fairly familiar with the AW4 electronics and TCM. Based on the symptoms of:
- Originally working, (rules out miswiring)
- TCM not powering #1 solenoid on pin C16 when stopped, but operating the t/c lockup solenoid as expected (verifies speed sensor and brake sensor)
- Manually applying power to the solenoids works (verifies the solenoids and wiring)
- You verified power on pin D14 and D16, and ground to D7 when connector plugged in. (eliminates high resistance in the power wiring dropping voltage)
You've done your homework. Would you agree this seems like #1 and possibly the #2 solenoid driver in the TCM is blown. What's causing the failure is the mystery. Intermittent short to ground or power on the solenoid wiring? Inadequate grounding to the engine block and trans. Shorting any of the NSS lines to power or to ground shouldn't do anything, as the get pulled up by the NSS. I'm stumped.

If I still have extra TCM on the shelf, I'd send you, but I sold it last year to someone in Europe who had transplanted an AW4 into something else. They usually go for about $50 on eBay. The local pull-n-save yard charged $8 for them when they still had Renix-era Jeeps on the lot. Now I can't find anything older than late 90's.

 

FINALLY got the chance to go through the wiring harness -- literally wire for wire! Found and fixed two issues: the power/comfort switch resistor had one leg broken (previously mentioned), and the TPS signal wires were reversed. I don't think that really mattered, as I'd originally cut (then soldered) the wires when I was adjusting the TPS way back when. Why? So I could get a good reading with my volt meter. Not finding any real big issues was both a relief and a disappointment, but I didn't think there were any real wiring issues based on it previously working (back in 2001).

Did some more checking of the NSS before installing the wiring harness. I couldn't get a signal in neutral across the designated terminals, but got it in every other range. Park uses the same terminals as neutral, and I was getting a signal in park. So I took the old NSS apart again. The terminals inside for neutral are very narrow, so it was obvious that it had to be exact to get a signal. An adjustment of the NSS fixed that problem. I rechecked the old NSS. I had attempted to clean and grease it (dielectric grease) before replacing, as many suggested. I couldn't get a neutral signal from it no matter what I did, so it was obviously defective, though the TCU doesn't use a neutral indicator -- just a signal from the 1-2 and 3 positions. Otherwise it assumes Drive (OD).

The final issue is actually the real culprit. I discovered a faulty connector for ignition power (D14, I believe). It's a bullet connector. If I pushed it in or wiggled it, I would get power on D14. Let it sit for a day or so and it would push itself out and break contact. That was obviously the entire issue the whole time, based on the way the TCU was acting. I was checking all connections (twisting and/or pushing the bullet connectors) before checking voltage, so until this got loose enough that it wouldn't stay in at all (when I was reinstalling) it was eluding me! I got everything put back in and fixed yesterday, but didn't have time to start and drive the car, so I'm still awaiting a test drive, but I'm pretty confident that intermittent power connection was the whole problem for all these years. Funny how something so simple can elude you when trouble shooting!!

 

Nicely done!! Anxiously awaiting the follow up.

 

Just had to stop by my shop (7 miles away on my brother's farm) plug in the TCU, and take a test drive! Good news is that it starts in first as it should with the TCU plugged in, so the ignition power wire was part of the problem. Bad news is that it doesn't up-shift -- just stays in first. Page 34 of the early Jeep Training Reference Manual ("AW-4 Automatic Transmission Operation/Diagnosis", Part Number 8981 321 183 on back -- not the AW-4 Service manual, which has nothing on the electronics) states that the TCU will not up-shift out of first if the speed sensor fails. I'm pretty sure I replaced it years ago, but looks like I'll have to pull it and check it for operation. Easy to do with a magnet and ohm meter or powered test light, it's just a magnetic switch. If it passes that test then I have to install it and jack a wheel up so I can turn the driveshaft. The magnet may be missing from the rotor on the output shaft -- if so the switch won't activate. I've got a good electronic VOM with a tone for continuity, so it will be an easy test.

Unfortunately I have to go out of town Saturday to a Carolina AMC Club meeting, and want to take my Rambler! So it's unplug the TCU and plug the switch box back in tomorrow evening. With work it may be another week or two before I can check the speed sensor. I'm really suspecting the rotor has lost its magnet somehow, but I may have a bad sensor or wiring to the sensor (maybe the connector?)... will have to wait and see!!

 

Yup, first thing to check is the speed sensor. Easy enough to check with the meter as you noted. A bad TPS indicating wide open throttle might make it stay in 1t until pretty high rpm. A bad #2 solenoid would also cause not shifting into 2nd.

 

Just an update: I tested the speed sensor and it's fine! Does anyone know where the magnet is inside the trans? It's possible that my trans has lost the magnet. I might try an external speed sensor with a magnet mounted on the driveshaft, the kind used for older model cruise controls... Can't find a source for the output speed control rotor, may have to go to a dealer or a used trans. It's in the output shaft housing or transfer case adapter on 4x4s. Don't know how it could have lost a magnet, but I got it used -- maybe it was removed? Appears to be held in place by snap rings and a key in parts illustrations... maybe key broke or is missing? I haven't taken mine apart, just going by parts illustrations.

 

Just an update. I finally pulled the output housing and checked the rotor. Magnet is fine, it can't come out anyway! So I'm back to square one. I'm just about of a mind to just build a new wiring harness -- it's not that much to it. I'm going to go through and check that I have power where I should and test the existing wiring for continuity to ALL the sensors again. I have the factory service manual and will hit it when I find the time first. Finding time is the hardest part of all this!!

 

You test the output speed sensor from the trans computer. Should see it closing once per revolution of the output shaft. Can test the TPS signal from there too.