• NAXJA is having its 18th annual March Membership Drive!!!
    Everyone who joins or renews during March will be entered into a drawing!
    More Information - Join/Renew
  • Welcome to the new NAXJA Forum! If your password does not work, please use "Forgot your password?" link on the log-in page. Please feel free to reach out to [email protected] if we can provide any assistance.

98 aw4 tc lockup switch

amp7876

NAXJA Forum User
Location
Greenville, IN
I have searched and come up with conflicting answers.

Can you wire the 97 plus tcu to lock up the tc?
One article says a check engine light will trigger
another says you can not do it all, yet another says it will work.

One article says the 97 and up does not have a seperate tcu.

So what is the verdict can this be done successfully on the 97 up aw4?

:flame:-away
Thanks
 
not really concerned about the check engine light.
Just concerned with reasons people say it can't be done on the 97 up.
Check engine light is already on for another reason, "no 02 sensor"
Thanks for the link though, they have a couple of great ideas.
 
It will work, you can do it, it will turn the CEL on. You just want TC lockup and no gear switching? That's pretty odd, the stock unit does a decent job of keeping the TC locked at the right times, most people want to shift manually instead.

If you want to avoid the CEL coming on, wire like so:
97uptcoverridecelsafe.png


The switch is a standard 3-position slide switch you can find at most electronics stores that aren't Radio Shack, you should also be able to find them online in various places, let me know if you need a part number and a recommended distributor.

With the switch in the top position the torque converter will lock up when the TCU wants it to. With it in the middle position the torque converter will always be off. With the switch in the bottom position it will be locked on.

You *might* have to replace that 100 ohm, 5 watt resistor with a 12 ohm, 20 watt resistor, but I'm pretty sure a 100 ohm will be enough of a current draw for the TCU to be happy.
 
No i want both. Just did not want to complicate it. Figured if I could do one i could do the other.


I am familiar with circuits etc... but after reading multiple things on the net, i had some conflicting info.
I have access to all the electrical items needed, I am a distributor.
I happen to be screwing around at work looking for another weekend project and was going to pick up the items needed before leaving for home.
The XJ is sitting in the garage so i could not go look at it.

Thanks again.
Tony
 
Last edited:
ah, cool! Good luck then. Didn't realize you were familiar with electronics :cheers:
 
It will work, you can do it, it will turn the CEL on. You just want TC lockup and no gear switching? That's pretty odd, the stock unit does a decent job of keeping the TC locked at the right times, most people want to shift manually instead.

If you want to avoid the CEL coming on, wire like so:
97uptcoverridecelsafe.png


The switch is a standard 3-position slide switch you can find at most electronics stores that aren't Radio Shack, you should also be able to find them online in various places, let me know if you need a part number and a recommended distributor.

With the switch in the top position the torque converter will lock up when the TCU wants it to. With it in the middle position the torque converter will always be off. With the switch in the bottom position it will be locked on.

You *might* have to replace that 100 ohm, 5 watt resistor with a 12 ohm, 20 watt resistor, but I'm pretty sure a 100 ohm will be enough of a current draw for the TCU to be happy.

kastein,

Have you verified that this technique works to prevent the CEL from coming on when using an aftermarket manual shifter on a 97+ XJ? I've seen some speculation elsewhere in this forum, but I haven't seen anyone claim to have actually tried it.

I have Rail Shifter on order with RADesigns Products (should be here any day now!), and I was hoping to employ this technique using 3 20-ohm 25W resistors acting as dummy loads. I REALLY want this to work WITHOUT the TCM/DCM throwing a code and turning on the dreaded CEL.
 
kastein,

Have you verified that this technique works to prevent the CEL from coming on when using an aftermarket manual shifter on a 97+ XJ? I've seen some speculation elsewhere in this forum, but I haven't seen anyone claim to have actually tried it.

I have Rail Shifter on order with RADesigns Products (should be here any day now!), and I was hoping to employ this technique using 3 20-ohm 25W resistors acting as dummy loads. I REALLY want this to work WITHOUT the TCM/DCM throwing a code and turning on the dreaded CEL.
I have not.

However... I got the CEL to go out on my 98 recently! Only a few more fixes and it is inspectable. Expect testing in the near future now that I have a rig to do it on without driving to a friend's place and worrying about blowing theirs up.
 
I have not.

However... I got the CEL to go out on my 98 recently! Only a few more fixes and it is inspectable. Expect testing in the near future now that I have a rig to do it on without driving to a friend's place and worrying about blowing theirs up.

Really? How did you do that. In my brief testing this weekend, I had a devil of a time getting the CEL to go out once I set it by disconnecting the TCM from any/all of the three tranny solenoids. I was hoping to find a way to do it without having to disconnect the battery. I'd love to know your secret! :laugh3:

If I can find the resistor locally, I hope to test them out this week -- schedule permitting. I'll be sure to post my findings.
 
Really? How did you do that. In my brief testing this weekend, I had a devil of a time getting the CEL to go out once I set it by disconnecting the TCM from any/all of the three tranny solenoids. I was hoping to find a way to do it without having to disconnect the battery. I'd love to know your secret! :laugh3:

If I can find the resistor locally, I hope to test them out this week -- schedule permitting. I'll be sure to post my findings.
Sorry to confuse - I bought the vehicle as a "some assembly required" project. It had a really, really bad radiator, a broken wire in the coolant temp sensor wiring, no e-fan, a bad upstream O2 sensor, and the interior was disassembled / airbags weren't connected when I got it. It also had a problem in the evap system. "getting the CEL to go out" involved fixing all of that - I bought it as a replacement for my '96 and also as a prototyping rig for all the stuff I've been trying to design. Just had to get it back to OEM condition before starting on everything, otherwise I wouldn't know if the problems I was seeing were due to pre-existing conditions or due to a mistake in my designs.
 
Sorry to confuse - I bought the vehicle as a "some assembly required" project. It had a really, really bad radiator, a broken wire in the coolant temp sensor wiring, no e-fan, a bad upstream O2 sensor, and the interior was disassembled / airbags weren't connected when I got it. It also had a problem in the evap system. "getting the CEL to go out" involved fixing all of that - I bought it as a replacement for my '96 and also as a prototyping rig for all the stuff I've been trying to design. Just had to get it back to OEM condition before starting on everything, otherwise I wouldn't know if the problems I was seeing were due to pre-existing conditions or due to a mistake in my designs.

Gotcha. I tried various strategies, with limited (and un-repeatable) success. Once, I got the CEL to go off after having pulled the fuse in the Power Distribution Block that supplies Bat+ to the TCM (motor off). However the 2nd (and 3rd, and 4th...) time I tried that, it didn't work. I'm still not sure why. Obviously, I'm missing something.

The 2nd time I got it to go off (and stay off) was purely by accident. I was attempting to retrieve the code that had been set. I cycled the key 3 (or 4?) times hoping to see the code on the odometer or via CEL flashes. Instead, I got no code at all and the CEL went OFF. I started the Jeep, and the CEL was off and stayed off. I did not have the chance to repeat that procedure to see if worked again.

Good grief. :confused:
 
Success!

**Not sure if this should be posted here or in a separate thread. I'll start here and re-post elsewhere if requested/required.**

I performed a few experiments last night, and was very pleased with the results. While I still have a few more tests to run, the important questions have been answered in the affirmative. And that means I will be able to install a full-featured aftermarket manual shifter in my XJ (98 model, stock 4.0L & AW4) that will not set a diagnostic code and/or light the CEL. :yelclap:

I received my RADdesigns Rail Shifter yesterday, and set straight to work. First, I needed to confirm two things on which my plan depends:

1. Can simple resistors be connected between the 3 solenoid outputs of the Transmission Control Module (TCM) and ground to act as dummy loads without the TCM throwing a code and lighting the CEL?

Yes. This was tested by simply cutting the 3 solenoid output wires in the harness near the TCM plug, and directly wiring a 12 Ohm, 50W resistor between each wire (on the TCM side, obviously) and ground. Started the Jeep and waited. Drove up and down the driveway a few times. No CEL. No problems. It works!


2. Can the TCM be switched back and forth between the dummy resistors and the tranny solenoids with the motor running and the transmission in gear without the TCM throwing a code and lighting the CEL?

Yes. This was tested by using the schematic below. A DPDT relay is used to switch each TCM solenoid output between the real and dummy loads. Note the break-before-make configuration assures that the TCM output transistors never see both loads simultaneously, which would obviously double the current draw. They do, however, see no load at all for a fraction of a second -- and that's what this test was all about. Would that brief no-load condition cause a code/CEL? Fortunately, it did not. The relays are controlled by rocker switches. Drove around the block several times, switching back and forth between manually sifting with the RADesigns Rail Shifter, and letting the TCM do it's full-auto thang. No CEL. No problems. It works!


I ran out of time, and was unable to hook up and test the TC part (orange wires, SW3, RY3, etc.), but it should work fine. It's fundamentally the same circuit.

A few additional observations:

1. Manually shifting the AW4 is SWEET. Now I understand why so many have reported in these forums that they actually prefer shifting it manually, even during daily driving. It's something you simply have to experience first hand.

2. The RADDesigns Rail Shifter is smaller than it appears online. It should be easier to mount/locate than I had feared.

3. The 12 Ohm resistors get really hot. I'm hoping that kastein's theory is correct, and that higher value resistors can be used to reduce current draw. Using the metal knee blocker under the steering column as a heat sink will probably alleviate the problem, but using higher-Ohm resistors would make for a far more elegant solution, if they work. Will test and advise.

4. The schematic can be easily adapted for use with any sort of manual shifter and/or a home-brew creation of your choice.

5. It should be noted that the testing was very brief -- less than 20 minutes driving time total. As such, I can't swear that this is the final solution but it sure feels like a major step in the right direction. I thought it was worth sharing.

Schematic:

ShifterSchematicV2.gif


Discuss.
 
:cheers: awesome! Glad you got it working.

My reasoning on the 100 ohm resistors being good enough (heck, probably even less current draw / more resistance would be fine, but I don't like to push my luck) is based off reverse engineering part of the stock TCU. It appears that as usual, the checking consists of:
* is anything drawing current at all?
* is it not drawing so much current it should be on fire?
* all's fine!

From what I could tell when I pulled the thing apart, they used a Sanken SI-5154 positive-side transistor driver for each solenoid, with some other peripheral components connected. Since the SI-5154 has no actual current measuring in it, just "is current being drawn" combined with feedback from the control pin for fault detection, it follows that unless I completely missed something, it's just looking for some load... any load at all really.

The only reason I threw so many caveats and warnings about this being untested info on there is because I haven't tested it, and I haven't yet fully reverse engineered the board, just gotten a bill of materials and traced the wiring on the top/bottom and checked to see if any wiring was hidden on the middle two layers (which I couldn't be sure of.)

I have picked up a set of connectors I can use to build a test/breakout board to go between the TCU and the transmission as of this morning though, and the only other project on the plate for tonight is installing a dryer for my landlord, so hopefully I'll have more solid info soon.
 
My reasoning on the 100 ohm resistors being good enough (heck, probably even less current draw / more resistance would be fine, but I don't like to push my luck) is based off reverse engineering part of the stock TCU. It appears that as usual, the checking consists of:
* is anything drawing current at all?
* is it not drawing so much current it should be on fire?
* all's fine!

From what I could tell when I pulled the thing apart, they used a Sanken SI-5154 positive-side transistor driver for each solenoid, with some other peripheral components connected. Since the SI-5154 has no actual current measuring in it, just "is current being drawn" combined with feedback from the control pin for fault detection, it follows that unless I completely missed something, it's just looking for some load... any load at all really.

I see. I know nothing of the 5154. :confused1 Care to hazard a guess as to the minimum current draw that will reliably trigger "yes, it's on!" state?
 
The SI-5154 datasheet is not clear on this... At least as far as I can tell. It appears to require a few milliamps of current to consider the output loaded. To be on the safe side and not risk a spurious CEL I would aim for a load drawing approximately half a watt (at 14 volts, ohm's law gives a current of about 36mA for this amount of power consumption) and use a 1 watt resistor to keep a good safety margin and lower the temp a bit. This gives a resistance of about 386 ohms, since that is not a standard value I'd try a 360 ohm 1 watt resistor which should draw about 39mA at 14V and dissipate about 540mW. Again, I haven't tried this yet, but I would bet on it not triggering the CEL when set up that way.

I have a bucket of assorted power resistors and a wiring harness pigtail sitting here now, so I'm going to solder up my test breakout cable and put it in tomorrow before my drive to work. Would do it tonight, but it's pouring rain and I have no garage.
 
Great write up guys, I've been playing with resistors too but without repeatable results, probably not the right ohms. This type of mod is simple most guys can just buy the parts and do it themselves. I get a lot of questions about fooling the OBD II ECU/TCU systems. I hope this leads to a positive answer.
Thanks again "raneil" for the great info, also to you "kastein" you're always a help with the electronics questions on the forum.
 
The SI-5154 datasheet is not clear on this... At least as far as I can tell. It appears to require a few milliamps of current to consider the output loaded. To be on the safe side and not risk a spurious CEL I would aim for a load drawing approximately half a watt (at 14 volts, ohm's law gives a current of about 36mA for this amount of power consumption) and use a 1 watt resistor to keep a good safety margin and lower the temp a bit. This gives a resistance of about 386 ohms, since that is not a standard value I'd try a 360 ohm 1 watt resistor which should draw about 39mA at 14V and dissipate about 540mW. Again, I haven't tried this yet, but I would bet on it not triggering the CEL when set up that way.

I only had a short window of opportunity to play with it last night, and I didn't come to any solid conclusions. I did notice that I accidentally left the #1 solenoid circuit open for quite some time with the motor running, and the CEL never came on. Odd. I also determined that opening the TC solenoid circuit reliably lights the CEL within about 10 seconds. Is it possible we could get away with needing only one resistor in this setup? Surely not. But it's definitely worth checking out.

I have a bucket of assorted power resistors and a wiring harness pigtail sitting here now, so I'm going to solder up my test breakout cable and put it in tomorrow before my drive to work. Would do it tonight, but it's pouring rain and I have no garage.

I look forward to seeing what you find out. It looks like family commitments will keep me out of the garage for at least the rest of the week, so it will be a while before I can get back to it. Keep us posted on your progress!

@R.DesJardin: Likewise; I hope all of this will lead us all to a definitive solution. I really, really, really hate to see that CEL come on!
 
I only had a short window of opportunity to play with it last night, and I didn't come to any solid conclusions. I did notice that I accidentally left the #1 solenoid circuit open for quite some time with the motor running, and the CEL never came on. Odd. I also determined that opening the TC solenoid circuit reliably lights the CEL within about 10 seconds. Is it possible we could get away with needing only one resistor in this setup? Surely not. But it's definitely worth checking out.



I look forward to seeing what you find out. It looks like family commitments will keep me out of the garage for at least the rest of the week, so it will be a while before I can get back to it. Keep us posted on your progress!

@R.DesJardin: Likewise; I hope all of this will lead us all to a definitive solution. I really, really, really hate to see that CEL come on!
Found out last night after soldering everything up that I used the transmission wiring harness connectors that match a 97. 97 is the odd year out, 96 and earlier use a different connector and 98 and later use another different connector... and I drive a 96 (which wouldn't throw a CEL anyways) and a 98. Oops!

Picked up the right connector for 98-01 testing at the junkyard on my way into work... Weather report says very low chance of rain tomorrow, so I should actually get something done then.

The only resistors I could find that were close to the value I calculated were 332 ohm, 1% RN70B-class milspec 2 watt resistors I picked up as a surplus lot a few years ago. I also found some 4xx-ohm RN70Bs and a few higher values as well, so I'll test a range of resistances and see what the approximate minimum current we can get away with is.

Also, I think I can explain why the CEL didn't come on for you when you accidentally left a circuit open - first gear (which you would be in, idling with your foot on the brake) is sol1 on, sol2 off. According to the diagnostic function truth table in the SI-5154 datasheet (copy/pasted below), the VO (output drive) and DIAG (current draw feedback) signals are indistinguishable between "normal" with the solenoid energized (VIN = high, VO = high, DIAG = high) and "open load" with the solenoid energized (VIN = high, VO = high, DIAG = high.) If you made it to a speed/throttle position at which the TCU would attempt to shift into 3rd or overdrive (sol1 off, sol2 on/off respectively) I am betting the CEL would illuminate immediately. This was confusing me when I looked at the truth table since I haven't tested any of this yet, I was scratching my head going "but how can it tell..." - I guess the answer is that it can't. I can't explain the torque converter one, though, maybe someone with more working knowledge of the AW4's hydraulic valving can explain that.

If the above holds true, the TCU also should not be able to tell if you short sol2 to ground while idling in drive at a full stop, nor should it be able to tell if it's trying to set 2nd or 3rd gear and you have sol2 disconnected. Please don't try shorting it though, I don't want you blowing up your TCU and while the SI-5154 claims to have overcurrent/thermal protection circuitry, my experience has always been that those features are worth the paper they are printed on and not much more.

Code:
MODE     VIN VO DIAG
normal   L   L  L
normal   H   H  H
open     L   H  H
open     H   H  H
shorted  L   L  L
shorted  H   L  L
overheat L   L  L
overheat H   L  L
VIN is the input, VO and DIAG are the outputs.
Note that shorted and overheat are duplicates (i.e. shorting the output will cause a thermal override, makes sense) and that normal with VIN = L and shorted with VIN = L are indistinguishable, as are normal with VIN = H and open with VIN = H.

For reference, here is the SI-5154 datasheet: http://pdf1.alldatasheet.net/datasheet-pdf/view/38275/SANKEN/SI-5154S.html
 
Last edited:
That makes sense. Sounds like we're closing in on it. We just need to drill down on the resistor values and then test them out on multiple vehicles for validation.

I really need to shop for some quality automotive connectors/plugs. I would have accomplished a lot more last night if I hadn't been fighting so many bad connections, alligator clips, and messy, tangled wires. :rolleyes:

Any suggestions for quality automotive plugs/connectors? What does everyone else use?
 
Back
Top