• 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.

AW-4 gurus, need advice.

T

tonyxj2k

NAXJA Forum User
Location
WV
Hi!
I'd been doing a little searching, and found some information that has helped, but nothing specific to what I'm trying to accomplish.

I'm looking for a way to force a downshift from OD to drive electronically in my 2000 XJ. The way I understand it, that would involve breaking the circuit on the overdrive sense wire to fool the TCU into thinking I'd manually downshifted, but I'm wondering if I'd also need to apply power to the 3 sense wire at the same time or is there already power on it when the shifter is in OD?

I'm not sure if anybody has tried this before, but what I'd like to do is use a microswitch or adjustable vacuum switch to control OD so that I can set it to downshift sooner (1/2 to 3/4 throttle instead of WOT). I'm hoping this will eliminate the yo-yo between 3 and OD I'm seeing on hills by downshifting sooner and staying in drive longer.

Thanks in advance!
Tony
 
I think I answered my own question. I checked the 3 sense wire at the TCM connector and it only has 12v when the shifter is in drive. So, I think what I'm going to need to do is make use of both the normally open and normally closed contacts on the vacuum switch. If I've thought it through right, I'll need to cut the OD sense wire and connect the transmission side to the center contact on the switch and the TCM side to the NC contact. Then I'll need to tap into the 3 sense wire and connect a jumper to the NO contact on the switch. This should allow everything to function normally until vacuum drops below the setpoint of the switch at which point it will break the circuit to the OD sense wire and complete the circuit through the jumper to the 3 sense wire. I guess it'll also need a diode in the jumper to keep from feeding back into the OD sense wire at low vacuum when the shifter is in drive.
Can anybody see any problems I've overlooked with this?

Thanks!

Tony
 
It sounds like you need gears :p
 
By the way, i don't think this is really relevant in your case but i think it would be kinda neat to build a box that controls the AW4 manually. The shift solenoids are just an "add power = on" kind of thing. With a few micro switches it would be pretty easy to build a full manual control box.
 
It's pretty easy to control (I built a controller with a double throw 6 pole rotary switch from Radio Shack and some wire), but on a 98-01 XJ the TCU will realize it's not in the right gear and set a CEL unless you fool it with some signal generators and stuff.

However, have you tried just putting the shifter in 3? That should downshift to 3rd and keep it there (or in 2nd/1st as required.) Downshifting into 1-2 will give you 1st if the RPMs would be under 4500, or 2nd if they would be over 4500 in 1st.
 
Porch951 said:
It sounds like you need gears :p
Click to expand...
That's it exactly! And not only do I need gears, but I'm also in that special place where OD is too high for the hills, but D is too low to run all the time. I'll regear eventually but was looking at changing the OD shift point as a temporary fix.
Porch951 said:
By the way, i don't think this is really relevant in your case but i think it would be kinda neat to build a box that controls the AW4 manually. The shift solenoids are just an "add power = on" kind of thing. With a few micro switches it would be pretty easy to build a full manual control box.
Click to expand...
It's been done. Like the rotary switch controller kastein mentions. I even saw one in the course of searching that uses two momentary pushbutton switches. How cool would it be to use a controller like that and repurpose both of the cruise control rockers in the steering wheel to serve as shifter buttons!!??:D

kastein said:
It's pretty easy to control (I built a controller with a double throw 6 pole rotary switch from Radio Shack and some wire), but on a 98-01 XJ the TCU will realize it's not in the right gear and set a CEL unless you fool it with some signal generators and stuff.

However, have you tried just putting the shifter in 3? That should downshift to 3rd and keep it there (or in 2nd/1st as required.) Downshifting into 1-2 will give you 1st if the RPMs would be under 4500, or 2nd if they would be over 4500 in 1st.
Click to expand...
I've been manually downshifting into 3 on the hills. That's what gave me the vacuum switch idea. I figure if I can adjust the setpoint of the switch right, a downshift into 3 will happen at the right time, every time, automatically. Hopefully using the sense wires instead of direct control of the solenoids will let me avoid a CEL as the computer will simply think I've moved the shifter to 3 and command a downshift as a result. Of course this all assumes that no hydraulic changes are being made by moving the shifter from OD to 3 to 1-2 and that manual gear changes are made by the computer based on which sense wire has power.

Thanks for the input guys!

Tony
 
I'm not sure what you mean by sense wires. Are you talking about the NSS leads or the ISS and OSS? The ISS and OSS are analog signals, they vary from low voltage pulses to high voltage pulses (16 per revolution for the ISS and 4 per revolution for the OSS) depending on input and output shaft RPM.

As for using the cruise control rockers - I have actually considered this, it's an idea I am a huge fan of. Problem is, they are resistively multiplexed. There are only 2 wires for all those buttons. The way I'm thinking of attacking this is to use a DPDT switch to route the 2 wires from the steering column to either the ECU (to allow cruise to still work) or the device you're building, which would use a resistor chain and some comparators (I'd suggest LM324 or LM339, they are cheap, durable, and versatile) to determine which button was pressed.

There are some hydraulic changes based on shifter position from D to 3 to 1-2, the 1-2 selection is the most different. I forget the differences between 3 and D but they aren't very significant.
 
kastein said:
I'm not sure what you mean by sense wires. Are you talking about the NSS leads or the ISS and OSS? The ISS and OSS are analog signals, they vary from low voltage pulses to high voltage pulses (16 per revolution for the ISS and 4 per revolution for the OSS) depending on input and output shaft RPM.

As for using the cruise control rockers - I have actually considered this, it's an idea I am a huge fan of. Problem is, they are resistively multiplexed. There are only 2 wires for all those buttons. The way I'm thinking of attacking this is to use a DPDT switch to route the 2 wires from the steering column to either the ECU (to allow cruise to still work) or the device you're building, which would use a resistor chain and some comparators (I'd suggest LM324 or LM339, they are cheap, durable, and versatile) to determine which button was pressed.

There are some hydraulic changes based on shifter position from D to 3 to 1-2, the 1-2 selection is the most different. I forget the differences between 3 and D but they aren't very significant.
Click to expand...

By sense wires I mean the TRS T3 (OD) sense wire on pin 9 of the TCM connector and the TRS T1 (3) sense wire on pin 22. The way I understand it, these wires (along with the 1-2 sense wire) are how the TCM knows which position the manual selector is in. Each sense wire has 12v+ on it from the transmission when the shifter is in the range that coresponds to that wire. If I understand it right, applying 12v+ to pin 9 at the TCM will cause the transmission to behave as though the shifter is in the OD position and applying 12v+ to pin 22 (while opening the circuit on pin 9) will cause it to behave as though the shifter has been moved to the 3 position. The main thing I'm not clear on is whether or not applying power to pin 22 while removing it from pin 9 would be enough to get the TCM to command a downshift or if there also needs to be a physical change to the position of the manual shift valve.

I'd be really interested to hear if you make any progress with using the cruise rockers as buttons for a manual shifter. That might be a project I'd like to tackle in the future but my skill level would probably only allow for buying a completed unit and wiring it in.

Thanks for the help!

Tony
 
Last edited:
tonyxj2k said:
By sense wires I mean the TRS T3 (OD) sense wire on pin 9 of the TCM connector and the TRS T1 (3) sense wire on pin 22. The way I understand it, these wires (along with the 1-2 sense wire) are how the TCM knows which position the manual selector is in. Each sense wire has 12v+ on it from the transmission when the shifter is in the range that coresponds to that wire.
Click to expand...
100% correct so far.

tonyxj2k said:
If I understand it right, applying 12v+ to pin 9 at the TCM will cause the transmission to behave as though the shifter is in the OD position and applying 12v+ to pin 22 (while opening the circuit on pin 9) will cause it to behave as though the shifter has been moved to the 3 position. The main thing I'm not clear on is whether or not applying power to pin 22 while removing it from pin 9 would be enough to get the TCM to command a downshift or if there also needs to be a physical change to the position of the manual shift valve.
Click to expand...
Should be, I think - I'd bet on it working for D/OD vs 3 selection. Note, if you set it up wrong and it sends power to more than one sense line at a time, it will almost certainly set off a CEL. I haven't tested this but I can't imagine the TCU not complaining loudly about being told to go into two different gears at the same time.

Technically speaking, the transmission goes into whatever gear it is told to by a combination of the manual shifter and the solenoids. The TCU is responsible for interpreting the ISS, OSS, TPS, NSS, and brake pedal switch inputs and determining (1) what position the gearshift is in (which limits what gears it can electrically command the tranny to shift into) and (2) what gear it should electronically select from those available to it using the solenoids. If you fake it out and tell it that the shifter is in 3 when it's actually in D, it should simply limit itself to 1/2/3 gears and never select OD.

tonyxj2k said:
I'd be really interested to hear if you make any progress with using the cruise rockers as buttons for a manual shifter.
Click to expand...
Nothing solid so far, just theory. What I just posted will hopefully get one of the other electronics nerds on here thinking and maybe they'll post something up... I've been considering this idea for a while like I said, but have been too busy with other stuff to get anything done with it.
 
kastein said:
100% correct so far.

Should be, I think - I'd bet on it working for D/OD vs 3 selection. Note, if you set it up wrong and it sends power to more than one sense line at a time, it will almost certainly set off a CEL. I haven't tested this but I can't imagine the TCU not complaining loudly about being told to go into two different gears at the same time.
Click to expand...
As long as it doesn't actually try to do it.:scared: :) I'm getting bad mental images of a transbrake effect @ 70mph and pieces of AW4 all over the road.
Technically speaking, the transmission goes into whatever gear it is told to by a combination of the manual shifter and the solenoids. The TCU is responsible for interpreting the ISS, OSS, TPS, NSS, and brake pedal switch inputs and determining (1) what position the gearshift is in (which limits what gears it can electrically command the tranny to shift into) and (2) what gear it should electronically select from those available to it using the solenoids. If you fake it out and tell it that the shifter is in 3 when it's actually in D, it should simply limit itself to 1/2/3 gears and never select OD.
Click to expand...
Perfect! That's what I was hoping for. Then I can use an adjustable vacuum switch and have control over how much throttle/load it takes to downshift out of OD by adjusting the setpoint on the switch. It should be very similar to using a vacuum switch to control a lockup converter when installing a trans with lockup into a car without a computer.:)

Nothing solid so far, just theory. What I just posted will hopefully get one of the other electronics nerds on here thinking and maybe they'll post something up... I've been considering this idea for a while like I said, but have been too busy with other stuff to get anything done with it.
Click to expand...

I'd love to try it myself, but I have absolutely no formal electronics training. Anything much past switches, relays, and the occasional diode or resistor is beyond me. I really should take some classes. I have a feeling I'd really enjoy it.

Thanks again!

Tony
 
It's simply impossible for it to try and get two gears at once... the signals to the solenoids are "encoded", i.e. 2 solenoids select 4 gears (on/off, on/on, off/on, off/off.) Shifting into 1st at 70mph could be, well, interesting, but 2nd and above would just peg the rev limiter.

Rather than using a vacuum switch, I would use another comparator and a variable resistor to set the setpoint relative to the voltage coming from the TPS. It's actually the way the factory TCU does it (though they implement it using an ADC and software running on a microcontroller rather than an analog solution) you just want to change the shift point.

All the electronics for your idea needs is -
* IF the D line from the NSS is at +12V, we send +12V to the TCU on the TRS sense line for Drive UNLESS the signal from the TPS indicates that you don't want it to go into OD ever, in which case, we send the +12V to the TCU on the TRS sense line for 3rd instead. If the 3 or 1-2 lines from the NSS are at +12V we should not change anything.

If you want to be able to control the transmission more extensively it gets more interesting to fool the TCU. I haven't fully worked out how to fool it completely, in fact.
 
Or just buy one of my shift controllers, assuming the CEL is no big deal. Otherwise pull the lever back sounds like the simplest. I had shift buttons on my steering wheel, I have a couple customers with joysticks to shift, many things can be done.
 
You know, it's really a shame that we even have to fool the TCU in the first place. In a perfect world, I could accomplish exactly what I want by tweeking a table and reflashing the TCU. I kind of miss the older Chrysler stuff where anybody with an eprom burner and software like ChEM or D-Cal could make whatever changes they wanted to their car's computer and everything you needed to learn to get started was posted on websites and forums for anybody to read. I guess all that went out the window with OBDII.:(
 
Buy a 95, and swap the frontend/console etc, then you don't have to "fool" the computer?
Really, these 4.0L's don't mind higher RPMs
 
R.DesJardin said:
Or just buy one of my shift controllers, assuming the CEL is no big deal. Otherwise pull the lever back sounds like the simplest. I had shift buttons on my steering wheel, I have a couple customers with joysticks to shift, many things can be done.
Click to expand...

That shift controller looks really nice! I may consider it in the future if I decide to go with a manual setup. Right now though, I'm just wanting to gain some control over the OD shift points.
 
Johnnie Walker said:
Buy a 95, and swap the frontend/console etc, then you don't have to "fool" the computer?
Really, these 4.0L's don't mind higher RPMs
Click to expand...

Right, we're only talking 2600 @ 70 running in 3. I just figured there might be a gas mileage improvement if it were allowed to shift into OD on flat ground and slight hills.
 
tonyxj2k said:
You know, it's really a shame that we even have to fool the TCU in the first place. In a perfect world, I could accomplish exactly what I want by tweeking a table and reflashing the TCU. I kind of miss the older Chrysler stuff where anybody with an eprom burner and software like ChEM or D-Cal could make whatever changes they wanted to their car's computer and everything you needed to learn to get started was posted on websites and forums for anybody to read. I guess all that went out the window with OBDII.:(
Click to expand...
yeah... I've actually looked into doing that on the TCU on these. It's a Fujitsu MB89665 mask ROM microcontroller. All I need to reverse engineer the firmware is the special programming adapter that makes it look like a 27C256 EPROM to my EPROM burner so I can dump the data out, obviously I would have to buy a new controller and solder it down after burning it but it'd be a start. Problem is, that controller chip is so obsolete I can't find anyone willing to sell me the programming adapter and getting a schematic to make my own is completely out of the question.
 
You must log in or register to reply here.

Similar threads

W
New build advice needed.
2
Replies
23
Views
5K
Goose_XJ88
Goose_XJ88
N
A strange start
Replies
4
Views
616
noah99944
N
B
Could AW4 solenoids/transfer case/differential be causing my rumble noise?
Replies
9
Views
1K
kastein
kastein
S
Need Recommendations For Reman Engine
Replies
3
Views
845
MotoBoss
M
B
Son's 97 XJ 4" build
Replies
6
Views
593
burntkat
B
Back
Top