Frank Swygert

First a little background. I installed a 1988 4.0L and AW4 in a 63 Rambler back in 2002. Just lifted everything from the donor XJ and plopped it under the hood of the Rambler, complete wiring harness, diagnostic connectors, etc. So it's essentially a stock XJ installation. That's why I posted here. I use a column shifter instead of floor, but everything else is identical to an XJ.

For the first 3-4 months everything worked fine. Then it suddenly stopped shifting. Picked up a factory AW4 manual, already had a factory wiring diagram. It would start in whatever gear the lever was in -- 4th or OD in D, 3 in 3, 1 in 1-2, just like it should when manually shifted and no power to TCU. Everything I've read, both in the manuals and on-line, indicate that if it's in manual (or limp-home) mode it was due to no power to the computer. Checked the fuse in the wiring harness, and even used a meter to check for voltage on both pins of the harness connector (one switched, one hot all the time). Had power, so that was good. Checked all the sensors, replaced the TPS and speed sensor, and ohmed out the solenoids. All was good, so I picked up another TCU from a salvage yard. That one worked for 2-3 weeks before it quit. WTF??? At that time new solenoids were a good bit pricier than now, but I checked thinking I might replace them anyway. Once I got the cost from a Jeep dealer (no aftermarket that I could find). Checked price on a rebuilt TCU, and that was a bit high at the time too. Due to the cost I did some more reading. Would be simple to make a toggle switch box and check shifting that way, so I did. Couldn't find anything wrong, so I made another switch box using a rotary switch instead of toggles (four position rotary and a couple diodes, and a rocker switch for lock-up) and drove it like that for the next 17 years. The switch wore out recently, and I decided to have another go at it.

So here is where I'm at now:
1. Sent my TCU to a company in Florida for rebuilding. They got it and said it was fine, but due to age I should consider replacing the electrolytic capacitors. I'm aware those can be a problem at some point so agreed. $300 with shipping both ways. Didn't do a thing.

2. Checked both power wires again. I already knew I had switched power, I'm powering the manual shift box from the TCU connector. Just tinned the ends of the wire and shoved them in the connector ports for switched power and the three solenoids. So I know the wiring down to the solenoids is good as well.

3. Checked all the sensors. Adjusted the TCU a bit to get closer to desired range (wasn't off much). I haven't checked the speed sensor yet, will do that soon. I suspect it's fine, because with the trans in D and power to the TCU (TCU connected, of course) it will automatically lock the converter, but it still starts in 4th gear in D. I didn't check for automatic lock-up in other gears.

4. I cleaned the neutral safety/backup switch. No joy, so I just decided to replace it since it's old anyway. Still nothing! I checked output on the NSS and found I wasn't getting anything. Turns out there was no power to the NSS! From the wiring diagram in the AW4 manual and factory wiring book the NSS gets power from the TCU fuse. Tahter than unwrap the harness I just cut the wire a couple inches back from the NSS connector and installed a jumper wire from the fused side of the TCU fuse and the NSS connector. I checed the NSS and I get the proper connections on everything except neutral. I get no connections in neutral at all (should be the same as park). Could be due to the column shifter, but I get the proper connections in every other position. From the AW4 manual only the 1-2 and 3 positions are sent to the TCU, nothing else is, so this shouldn't be a problem. I don't have back-up lights nor do I have the actual NSS portion connected (it will start in gear). No issue driving sine I'm the only one who drives this vehicle, and as far as I can tell from the AW4 manual and wiring diagrams it should have no effect on the TCU.

5. I've checked and cleaned all the grounds. I actually crimped a new connector on the TCU grounds. they are all on the block at the same location -- stock 87-90 XJ.

I'm at my wits end here!! the only clue I have is the dead power wire to the NSS. My only thought is that maybe it shorted and has a couple wires crossed, and the TCU is getting mixed signals and doesn't know what to do, so shuts down. Right now I've replaced the worn rotary switch and am using the manual bx again -- shifting with the column lever is a PITA (would be with the floor lever in an XJ as well long term). Anyone have any similar experience with an early AW4?? I've asked on the AMC Forum that I frequent since several have installed 2WD AW4s in cars (some on AMC V-8s), but no one has had any issues or they are using a later OBD compliant computer. I stopped at a local trans shop to see if they had any clues or could read the data from the TCU, and they were no help at all. Seems like there is only 8-10 codes from the early TCU anyway, and they wouldn't help much.

Hope someone here can be of more help before I rip the wiring harness out and just make a new one! I'm really going to be upset if I do that and it doesn't solve the problem!!

lawsoncl

You're checking power at the TCU with it plugged in? There is a fuse in the original wiring harness near the TCU connector. If the connection is corroded, you might see voltage on the connector with it unplugged, but not have enough voltage with a load on it. With it parked, is the TCU sending voltage to tthe #1 solenoid to command 1st gear?

Frank Swygert

I've got power at the TCU connector on both the switched and always hot lines. Fuse has been checked and cleaned as well. I haven't checked to see if the TCU is sending the signals, but it's pretty obvious it's not because it starts in 4th with the selector lever in D. The lock-up converter works as it should though, locking up at speed and disengaging with the brake switch. I'm using the TCU connector (therfore factory wiring) to manually turn the solenoids on and off for shifting now, so the wiring is good to the solenoids, and solenoids are working.

cruiser54

Original TCU, correct?

The plastic connector on the TCU, what color is it?

lawsoncl

The toggle switches are drawing power from the connector, right? That would seem to eliminate power, and solenoids and point at the TCU, which you're also eliminated. You've got a stumper here.

You said you checked grounds at the engine block. Have you verified a solid ground at the TCU? I'm just thinking the solenoids ground through the trans housing, but the TCU has it's own ground in the connector. Measure voltage on the ground pin at the TCU connector with it plugged in, or measure resistance back to the battery. The renix wiring harness has a crimped connection in the wiring loom in the engine bay against the firewall above the oil dipstick that sometimes goes bad. It's half-dozen ground wires coming together, including the TCU, TPS, as I recall. (Yeah, the renix had lots of poor grounding designs.)

Frank Swygert

Hmmm... it's a replacement TCU, but it's the correct number for the 87-88. The connector on the TCU itself is an off-white or tan IIRC -- I'll have to check to be sure. The wiring harness end that plugs into the TCU has a black and sort of tan color on it -- black on the C side, tan (maybe a dark off-white.. or faded white?) on the D side IIRC... might have that reversed.

I've read your site and bookmarked it! I was about to ask if there is a ground splice in the TCU harness. As you know the 87-90 (well, at least 87-88...) models have an independent TCU harness, not integrated with the engine/chassis harness like the 91+ models do. I think that eliminates the ground splice that lawsoncl mentions, maybe not.

To measure voltage on the ground with the TCU plugged in I'm assuming back probe it with the common probe of the multi-meter on a good ground. Ohm back to the battery by using a jumper wire for the ground back to the battery, probe the connector itself. I'm familiar with electrical troubleshooting, just making sure we're on the same page! I was thinking I might have to check grounds, and maybe pull the harness and unwrap it to look for shorts or splices. May be easier to do that (since it's a stand-alone trans harness) and start ohming wires from the TCU connector to where they are supposed to be.

The biggest stumper is it worked for months, then suddenly quit, but worked again for a couple weeks with a used replacement before quitting. the harness and trans did sit for at least a year between being pulled and then sold to me, maybe a couple years.

00t444e

How about one of these. http://www.radesignsproducts.com/rail-controller.html

lawsoncl

Yup, sounds like we're on the same page for troubleshooting the ground side. The TCU harness is mostly independent of the engine harness. As I recall it had it's own wiring to the dual-output tps, it's own brake pedal sensor, but that power and ground tied into the main harness under the hood. This is the stupid in-harness splice where multiple grounds come together I was thinking of. That one's been fixed, but the oem is just a the wires crimped together without even a butt connector. Roughly centered above the valve cover.

Frank Swygert

Well, the ground sounds like the most likely culprit at this point. Could have been marginal, and maybe I moved the harness a bit and it worked for a couple weeks with the replacement (years ago!) before giving up. I'll have to check the grounds first chance I get, which may be a couple weeks. Will post back here.

The rotary switch box I made up works the same as the RADesigns shifter, using about $20-25 of parts and about an hour of time. It's just a little 2"x3"x1.5" deep handy box with a new aluminum plate cover (cover extends over one side of the box for about 3/4" and has about a 1/2" lip bent down to hook over edge of my console), with a 6 position rotary switch (only four used, of course) with two diodes to control power flow, a rocker switch for the lock-up solenoid, and some wire. Just takes a few minutes to solder the wires and diodes to the switches. Four wires -- three for solenoids, one for power. I just tinned the ends of the wires for about 1/2" and insert them into the pin holes of the TCU connector. Full manual shift, no TCU you could even eliminate most of the wiring if you wanted. Just put it in "D" and use switch to shift. I started to re-wire it to shift 1-2-3-3 with lockup-4 with lockup. I usually just leave the lock-up switch on, but there is a hard shift between 2 and 3 with the converter locked. Let off gas when making the shift and it's not so hard, or turn of the lock-up when shifting. Doesn't shift hard between 3 and 4 with the converter locked, only 2 and 3. Wiring it so it only locks in 3 and 4 would be more convenient, and only take another diode or two. I didn't change anything because I'm hoping to get the TCU working again. mine doesn't look as cool, but does the exact same job for much less.

Frank Swygert

I have the wiring harness out of the car now. After pulling the covering I discovered there are NO SPLICES in the grounds, but there are two splices -- one on the power wire going to the comfort/power switch and an resistor, and another that I think is on a power line... I forget now exactly what line it's on. I did discover that if that resistor in the power/comfort switch line is disconnected you'll blow the TCU power fuse! Broke the resistor off one side when messing with the harness and started blowing fuses.

I soldered the two splices, but still not certain that will work since I have been getting power through the harness to manually control the solenoids. I'm beginning to think the company I sent the TCU to didn't fully test it. I'm going to see if the soldered connections helped (maybe only one portion of the TCU was getting power?), but I'm going to check each wire with an ohm meter first to make sure there are no breaks. If this fails I'm either going to just continue to use the manual switch or re-wire it to use a MicroSquirt. The MicroSquirt will cost me about $400 (and the time to wire)... not sure I want to spend more money on the thing right now!

There are a couple things that really puzzle me. First, the trans starts off in fourth gear (OD) with the selector lever in Drive even with the computer connected. It should only do that if there is no power to the TCU. I've been using cavity F16 to power the solenoids, but there are three different power signals -- one on F8 and E8 (labeled "fused B+"). So maybe I'm not getting power to one of those. The second is that the converter lock-up seems to work fine though with the computer hooked up and .shifting manually with the gear selector. Hopefully those two power splices are the culprits...

lawsoncl

Popping the fuse if the resistor is removed doesn't seem right. Shorting across it would likely blow a fuse though. Is the TCU even trying to power the solenoids?

Frank Swygert

The resistor may have shorted against something, hard to tell now that the harness is out. One leg was broken though, and it started blowing the fuse when I was messing with the harness and connectors. Nothing else appears to be out of sorts, just one leg of the resistor broken, so I suspect it. I haven't checked all the wiring yet though. I may have nicked a wire when cutting the covering off...

The TCU doesn't seem to be doing anything with the trans shift solenoids, only the lock-up. I know the wiring to the solenoids is good because I've been using that all along with my manual controller -- wires tinned and pushed into the TCU connector. Been getting power from the main line too. At first I thought it was the neutral safety switch, as that also tells the TCU where the gear selector is, but that doesn't seem to be the case. The old switch wasn't making good contact and was pretty rough inside, so I replaced it with a new one. I don't seem to be getting a connection in neutral, but the other ranges are correct, so that shouldn't be an issue. It IS making the proper connection when in "D", so the TCU should be starting the trans off in first when in D, but it's not. It's acting as if the TCU has no power at all (blown fuse). According to the manual the TCU won't shift OUT of first if one of the other sensors are bad, specifically the speed sensor and TPS. I'm about ready to just give up on this thing, but not quite yet -- still want to go through the harness and check wire by wire in case there is a short, and make sure those soldered power splices weren't causing an issue.

lawsoncl

I'm leaning towards a bad TCU. Does you're manual switch setup disconnect the solenoid from the TCU? I'm wondering if applying 12-volts to the TCU solenoid outputs killed the drivers.

Frank Swygert

No, the manual shifter completely takes the place of the tcu, not plugged in with it. The TCU was sent off to be rebuilt and should be good, unless a wiring issue blew it. I don't think that's the case, but I have yet to check all wiring.

lawsoncl

Why was the TCU rebuilt? They rarely go bad. I'm not even sure what a rebuld would consist of. Maybe replace any electrolytic caps, and replace any blown driver chips. I would assume they do some basic testing and would have caught that it's not driving the shift solenoids.