Well, this whole crazy thing started when I started researching the similarities between the Jeep AW4 and the Mitsubishi V4AW3 transmissions. I kept my eyes open and found an NP231 from a 2001 Wrangler with a Slip Yoke Eliminator (SYE) already installed. I also found an AW4 tailhousing from an unknown year Cherokee.
My first steps were back in December when I tried mounting the AW4 tailhousing and np231 to the V4AW3. The tailhousing bolted up just fine and the splines ended up being the same size and count. This initial "test fit" did let me know that if I wanted this to work I would need a "short" input shaft for the NP231. Over the years NP231's have come with different input shafts varying in spline count and lenth. There are three lengths and two spline counts: 21 and 23 splines. The V4AW3 is a 23 spline tailshaft. I also used this opportunity to check the clearances with the floor, exhaust, crossmembers, and t-bars. I felt confident that the NP231 was a good fit and at the most would need minor clocking. After this wrenching session, I reinstalled the Active-trac t-case so I could continue driving the truck while I shopped around for more parts. I would recommend messing with the Active-trac t-case as little as possible! It is a heavy beast to bench press into place.
Since I first started dreaming of this mod, I knew that I wanted to make a dual transfer case setup. I felt that the dual case provided the following benefits:
1)Allowed for selectable ranges, ie:1:1, 2.72:1, & 7.4:1. With these options, even the middle speed is lower than the stock 1.90:1!
2)It was something I could piece together as time and money permitted.
I had three options to choose from in trying to create a dual NP231:
1) Build the components myself. I didn't like this option simply because I knew it meant welding a mid-shaft together. I wasn't crazy out a welded shaft being in the heart of my drivetrain.
2) A Mad Rooster Offroad crawlerbox
3) Or a D.D.Machine "Box4rocks"
I decided on the D.D.Machine kit. I chose this kit because the guy that makes them seemed to really stand behind his product and know what he was doing. Also this kit had provisions for oil filling and draining. I was also less than thrilled with how the Mad Rooster guy handled himself when answering an email of mine.
I was lucky enough to find a Box4rocks 23 spline kit second-hand that had not yet been installed and came with all the parts that would be needed from a donor NP231 to make the doubler. I got even luckier when I found out that this doubler donor was a 23 spline short input shaft!
Here are some pics of the Box4rocks and doubler parts:
The doubler cut, welded, and guts assembled (you can see the midshaft poking out in the middle):
The plate that seals the rear of the doubler:
The clocking ring:
Once I had the doubler parts I wanted to start test fitting things to make sure everything would fit before I got too far into the project. I began by temporarily assembling the two cases. I soon found out that the assembly was almost too long and there was some interferance with the crossmember. Not only did I want the assembly to clear everything, but I knew that I would need some additional room for the drivetrain to "float" and enough clearance to reach the drain and fill plugs.
To clear the crossmember I did two things. The first was to remove this extra tab on the rear case housing:
It ended up looking like this:
Removing the tab helped the assembly to clear the crossmember, but it still didn't give me enough room to access the drain plug. My solution to this was to shorten the entire assembly 1/2" by removing the clocking ring. The clocking ring is a nice feature that is included in the kit, but unfortunately I could not afford the extra length that it caused. I experimented with various clocking positions and finally found one I was happy with. I settled on the final position because it was nicely tucked in above the t-bar without hitting the bottom of the floor.
To eliminate the clocking ring, I marked it's location and used the ring as a template to drill through the rear plate of the doubler case. Once these holes were drilled, I countersunk them so that bolt heads would clear the internals of the doubler case. The result looked like this:
I bolted the assembly together only to find out that once the clocking ring was removed that the rear case would no longer clear the bolt head that holds the rear part of the doubler shifter rail from floating around. I solved this problem by countersinking that bolt head as well:
Once that was completed, I assembled them again for another test fit and was happy with the results.
Now that I was happy with the dual transfer case assembly, I wanted to focus on rebuilding the cases, shifting them, and finishing the electrical work. The cases were rebuilt will all new bearings and seals. I won't go into any detail here since there are several good sources already on how to rebuild an NP231. I would like to mention that that this step should be considered when trying to budget this project.
Rather I'll skip straight to the cool stuff... The shifter! I had a few conversations with Landraider over this topic and he was a great person to talk with for brainstorming ideas. There are so many different routes a person could take here, but I wanted something that was both functional as well as not too out of place in the truck.(I almost have to laugh at that one)
I started out with an auto shifter from a Dodge Omni:
With all respect, it probably could have been made to work, but I didn't like the idea of two of those shifters side-by-side. To be honest, as I mulled over various options, I wasn't too happy with any two shifters side-by-side. It was from this and talking with Curtis, that I realized a viable option: a manual transmission shifter from a FWD car.
This option had real potential because most FWD manual shifters use two cables. One cable is for the side to side movement, and the other cable is for the front to rear movement. I was curious if I could use one of these shifters to use each cable independently to shift each case. After spending an entire Saturday morning at the junk yard, I found a shifter that I thought I could make work. It was from a Hyundai Elantra, and was chosen because it was in great shape, had boots that were still intact, and had open eye rubber-bushed ends that seemed easy to work with. I nabbed the shifter, base, cables, firewall grommet, cable brackets, and any other hardware I felt might be useful.
Once I had the parts home I began mocking up the cable geometry. I had to weld an extra tab onto the shifter bracket for the rear case. The front case shifter bracket worked out ok. I cut apart the cable brackets from the donor car and welded them to new brackets that I made to bolt to the cases. The resulting cable brackets looked like this:
I hooked everything up to the shifter and tested it out. It worked just fine.
With this arrangement, the front doubler case is shifted when the shifter is moved from side to side and the rear (full function) case is shifted when the shifter is moved front to back. Because both cases shift independent of each other, I have the option of 2L if I really wanted it.
Mounting the shifter in the cab was a bit of a challenge. I ended up removing the SRS control module to buy me some routing space for the cables, and I also had to remove the duct for the rear passenger floor heat from under the driver's seat. I reused the firewall grommet fromt he donor car and passed the cables through the floor on top drivers side of the transmission tunnel near the firewall. I also had to trim the center console to fit the shifter in its new location as well as heat the shifter shaft and bend it slightly so that it wouldn't interfere with the stock transmission shifter. Ultimately, I ended up relocating the locker switch further back on the console just under where the cup holders are. I'm not sure if I like this location yet, and I might end up relocating this switch up to the dash instead. Here's the shifter installed:
Here it is as viewed from the driver's seat:
I still need to enclose the shifter pivot and get a boot on it, but for now that's just a cosmetic issue.
I was able to reuse the transmission shift cable bracket with minimal work. I simply used an existing hole on the new AW4 tailhousing and made a bracket that bolted to the front case for the top hole. Once it was installed, I adjusted the cable so that it shifted properly in its new position.
While playing around with the shifter for the cases, I realized the doubler case could be smoothed out a bit by removing the top-mounted vacuum switch. This was an easy decision to make since I no longer needed this switch. I measured the threads (5/8-18) and was able to locate a replacement plug for it at NAPA.
The plug as-is was too long for the case, so I had to grind it down to make it shorter. Once that was done, it fit perfectly and made the shifting of the doubler case much smoother.
The next step was to try to make the electrical items work. Early on, I realized that the pulse sensor for the transmission would most likely turn out to be a pain in the . That fear was soon realized. Like mentioned before, the AW4 tailhousing was from an unknown year, and didn't work with the Montero's stock pulse sensor. It was evident that the two sensors were two different sizes, so I tried to modify the AW4 to accept the Montero sensor. This soon proved to be a bad idea. I had to build up material with weld and machine flat surfaces as well as bore out the hole. In the end, I questioned the sealing ability of the bore and the clearance between the pulse "gear" and the sensor. Frustrated, I did more research. To my suprise I found that in '96 Jeep changed the AW4 pulse sensor to a four pulse rotation sensor (same count as the Montero's). I decided since I screwed up the tailhousing that I had that I might as well try for one of the '96+ housings. While I was at it, I decided to try for the housing, sensor, and pulse gear from the same tranmission in hopes that the combination would be a direct swap onto my transmission.
Was I ever glad I made this decision! After many phone calls, I was able to track down a parts donor for my ever-mutating Frankenstein. The donor came in the form of a '99 Cherokee Sport. I knew at first sight of the parts that I was very lucky, the hole for the sensor was the same as the Monty's! Once I was home, I began comparing the two sensors and realized that the business end of the sensors were the same! In the picture, you can see the Montero sensor on the left and the Jeep sensor on the right:
The '99 AW4 tailhousing with Montero sensor in place:
I wasn't entirely lucky with these parts, however. The Jeep pulse gear was much longer than the Montero's and wouldn't fit on the V4AW3's tailshaft without some work. I ended up boring an internal shoulder and replacing the stock Montero circlip with an OD snap ring.
In this picture you can see the Montero gear on the left and the Jeep gear on the right. This picture was taken after I modified the Jeep gear.
The transmission pulse sensor was the last thing keeping me from bolting the case assembly up to the transmission.
Once that was accomplished I reinstalled the case assembly and realized that with the brackets for the shifter cables, I now had an interference problem. One of these brackets was dangerously close to touching the bottom of the floor. I could have ignored it and left it as-is, but I new that with any drivetrain wrap at all, the bracket would start tapping the floor. The solution? BFH!!!
It was not as bad as it might sound. In all reality, I only had to "massage" this area a bit, and it wasn't drastic at all. A body lift would have made this a non-issue, but what's the fun in that?
With the cases very close to being in the final location, I turned my attention to the transmission crossmember. I started by flipping it around so that the drivers side was now on the passenger side and bolted it back into place. I ended up welding up this little bracket which uses the isolator from the original transmission mount.
Most other dual transfer case applications will create another crossmember or mount to support the additional weight of a dual transfer case assembly. After lifting the old Active-trac assembly, I wasn't worried about the additional weight being a concern. I've also checked out the drivetrain wrap with the cases in Low^2 and the stock isolators seem to handle it just fine.
For the rear driveshaft, I decided upon a Tom Woods unit. I double checked the driveshaft geometry and settled on a double cardan front joint with a booted extra travel slip shaft. The extra travel is far from necessary with this application but I chose it because it provides more contact area on the splines (which should translate to longer life) and I might just change the rear suspension some time in the future.
I also had Woods machine the pinion flange so that it would accept a standard off-the-shelf Spicer universal flange.
The first spin around the block revealed that the speedometer wasn't working. At the time, there were several reasons why this could be.
1) Electrical problem (wires not connected in correct order, blown fuse, bad sensor, etc)
2) The sensor was providing the wrong type of signal
3) The speedo gear housing wasn't clocked properly
I started with the first item on the list. I checked the wiring and verified that the ground was good and the sensor was getting a full 12v.
Scratching my head, I decide to check the third item on the list. I pulled the speedo sensor/gear and reinstalled it while making sure the gears were meshing properly. Another spin around the block confirmed that the speedo still wasn't working.
At this point, I decided that the second item on the list was the culprit. This turned out to be the case. What little information I could find online suggested that the Jeep sensor was an eight pulse per rotation, square wave sensor. From what I could tell with an analog multimeter this was not the case. It appeared to be a sin wave sensor.
This is where I got the bright idea to check the stock Montero pulse sensor. It was clearly a four pulse per rotation with a square wave.
After reviewing all the speedometer gears, stems, and housings that I had, I realized that the threads on the Montero pulse generator were very similar to the threads on a stem I had from a cable driven NP231 speedo. I carefully threaded the two together and realized it could work! Unfortunately I needed an adapter key to mate the two shafts together. After measuring a stem from another part I had, I knew that I needed a double male .104 square adapter key. Most of you have probably read how I found that part.
Now that I had a complete sensor assembly, I installed it in the t-case and went for another test drive.
Success! The speedo was now working, but it was reading 15-20% too fast. I have still to fix this little problem, but today I received my SGI-5 Digital Dakota speedometer signal interface. This problem should be straightened out shortly.
Other than that, I still need to make a front driveshaft, and a skid plate to protect it all.
Can't wait to test it out on the trails!
Here's a few more pics of the cases in their new home :
The clearance between the rear case and crossmember:
Editors Note..... Pic links are DOA, but I am working on that