I have an AW4 transmission. It's in my 1972 AMC Javelin with a 304, but since the transmission, TCU etc. came from a '94 XJ, I thought maybe someone here might have seen this before. First of all, the AW4 works great with the 304. I've been running it for about five years now, but for the last few months the TC wants to unlock under light throttle applications. If I give it a little throttle it locks up the TC again. If I'm just cruising along at 45-55 mph it locks and unlocks constantly, but I can cause it to stay locked just by giving it some throttle. It's not really speed related though, just seems to have to do with whether I'm on the throttle or not. I replaced the TPS. The TV cable is properly adjusted and it shifts fine otherwise. I looked at the ground connections as I have some of them connected to the throttle bracket and wondered if I was causing a poor ground to connect whenever I stepped on the gas and possibly breaking the ground connection when I don't, but I jumpered those wires to a known good ground and it made no difference.

Any AW4 owners seen this behavior before? I do have access to a spare TCU I can try but will have to drive a few miles to get it.

The other thing that comes to mind is the brake switch, that is there to switch off lock-up when the brakes are applied, but that seems to be working normally and the only thing I can relate the behavior to is throttle position.

Thanks in advance!

 

TPS would have been my first thought. Brake sensor would have been my second after that. Engine torque moving a wire would have been my third.

Do you have a meter and can check the TPS signal at the TCU is changing smoothly? It's very rarely the TCU itself.

 

Yes, TPS looks smooth. I actually have two TPS here (different brands too) and both perform exactly the same. What puzzles me is that when the torque converter unlocks on light throttle, I can make it lock up every time just by giving it a little gas. Seems like if it were a wiring problem it wouldn't be consistent like that.

Which transmission solenoid controls the TC lock up? I might check the resistance on that.

 

Drove out and got my backup TCU and it performs exactly the same, so I've ruled out TPS and TCU, and I know when the TV cable is out of adjustment because then it shifts like crap, but it's actually shifting perfectly except for downshifting from O/D to 3rd on light throttle. Might just go ahead and buy a lock-up solenoid as they're cheap and I wouldn't mind changing the trans fluid and filter anyway while I'm there.

 

You "looked" at a ground and determined it was good?

Here's what that ground affects in an XJ:All the components listed below ground at the dipstick tube stud:

Distributor Sync Sensor, TCU main ground, TCU “Shift Point Logic”, Ignition Control Module, Fuel Injectors, ECU main ground (which other engine sensors ground through, including the Oxygen sensor, Knock Sensor, Cruise Control and Transmission Sync signal. All extremely important stuff.

 

Yeah, stared at it for a good twenty minutes too! Seriously, I tightened the ground connections and also ran a clip lead over to another ground just to be sure. Keep in mind, this is just a 1994 AW4, so no ECU (just a TCU), cruise control or sensor of any kind other than a TPS. Still, I know how common ground issues are so won't ever rule them out.

I've been doing a lot of reading of old posts and found several that mention the same problem with no apparent resolution. Here's a post from forum member Bobolink in 2015 which describes it very well:

Another from forum member arto-wa, in 2018 that is exactly what I experience:

I saw another post that said that that early XJs did this but later ones didn't, so before I go chasing my tail is there any truth to the notion that this is normal behavior on pre-1996 or so XJ?

I'm going to disconnect the brake O/D switch temporarily too, just to rule that out.

 

Ok, found two more ground wires that I'd forgotten that were separate from the others. Temporarily jumpered them to another ground location and also recrimped connections to the brake switch that looked like they could use improvement. Not sure which of these two things fixed it but it appears to have solved the problem. Went for a test drive and everything worked as it should-- no locking and unlocking of the TC under light throttle. Since I installed this transmission five years ago I've invested in some better crimpers, so I think I'll go through and redo any connections that look suspect.

EDIT: Well, this is beyond frustrating. Made all the changes I mentioned above and the problem is back. Repeated my test with jumpered connections as I had done successfully this morning and now that's not working either. What's changed since this morning? Temperature is the only thing I can think of, both ambient and also the TPS and transmission are now warmed up from running the car. The search continues...

 

I think you're on to something. Not temperature related I bet. connections, grounds etc were touched.

 

Any way to get a meter in or a test light hooked up to the output of the TCU for the lockup solenoid? I'm thinking some way to determine if it's the solenoid failing or the TCU is actually commanding it on/off. Or just add a switch to power it directly and see if it's still acting up.

 

That's easy.

 

I would add a switch as suggested with diode that alights, this will tell when PCM commands and also allow full manual control, which would be good for cruising on freeway and some off road applications

maybe there is a high resistance in the TPS -PCM wiring, try piggy back that one

I solder a pin or insulated paper clip to 5A wire

 

Yeah, that's doable. I did check the resistance of all three solenoids, which was 15.7-15.8 ohms at the TCU, which I think is probably a little higher than it would be if measured right at the solenoid. Resistance of the ground at the TCU (pin D7) also seems good. I also bypassed the brake switch and went for a test drive, but it didn't change anything.

 

additional,

what i said is slightly wrong a correctly wired setup will show when the TCM commands

I am 99% sure pre 96 has no PCM control over trans changes

if you google ****** go jeep, his webpage shows what to do

 

Yes, this is in a '72 AMC Javelin so no PCM. That's why I chose a '94 transmission. The AW4 has been in the car for five years now and only in the last few months I've noticed it unlocking TC at light throttle and locking up again when I add throttle. My first thought was dead spot in the TPS, but I've ruled that out.

 

I think I've figured out what's going on.

Because I'm using the transmission with a carburetor, I had to come up with a way to actuate the TPS in the absence of the XJ's throttle body. I came up with a little pulley arrangement and epoxied the pulley on a shaft into the rotating part of the TPS. I bought a little TPS bracket off Amazon and found a place to mount the TPS near the carb, then attached the other end of the cable to the carb throttle linkage. That's actually worked great for several years, but I recently changed the carb to a Holley Street Demon, and the linkage has slightly different geometry than the old Autolite carb I was using. That meant as the throttle was depressed the voltage curve of the TPS was different as the throttle linkage moved through its arc. In other words, the cable to the TPS was moving more than it was before when the throttle was depressed, so to get the voltage so that car both shifted normally and had the right voltage at full throttle, the cable had some slack in it at idle. At light throttle, the cable was still slightly loose, so the TPS voltage at times was reading as if the throttle was closed. That meant the TCU thought the car was coasting when it wasn't, and I'm pretty sure the AW4 is programmed to unlock the TC whenever your foot is off the throttle. You probably never notice that the TC has unlocked when the car is coasting, but when you're actually on the throttle a little it's very noticeable, and the slightest throttle movement would cause it to lock or unlock.

The new carb has three different attachment points for different transmission TV cables, so by using one of those I'm pretty sure I can get the voltage where it needs to be throughout its range. I temporarily attached the cable with tie wraps in different places for testing. I think I've found the sweet spot and now just need to come up with something permanent.