Roger Lister's Guide to Chrysler, Dodge, and Plymouth Minivan All Wheel Drive
The AWD setup in the Chrysler minivans is setup with a Power Transfer Unit that replaces the right transmission housing extension. What do I mean by that? If you pull the halfshaft from the right side, you will find a bolt on housing. Take that off, and then bolt on the PTU. Then you replace the halfshaft and FWD works as before. Oh, you will need the specific K member for this to work, because the original K member won't allow the the Torque Tube to pass to the rear.
The Torque Tube transfers the power from the PTU, back to an overruning clutch unit (which allows the front and rear wheels to spin at different speeds). Within this unit is a vacuum powered selector that engages the Rear Driveline Module.
The Rear Driveline Module has a differential in it, and sends the power to the rear wheels with halfshafts. The RDM is bolted to the chassis of the van and rubber isolated.
Now for the rear axle. All the minivans use LEAF springs in the rear. Since the rest of the EEK based cars use coils, a little imagination metal fabrication is needed. I would strongly suggest custom built trailing arms to mount the axle and provide a sturdy perch for the springs. (ever since I got my Daytona, I've been seriously considering just this idea...but I have a Turbo Z. If I'd gotten a BASE Daytona...) The hubs ride in knuckles not unlike the front, just different mounting.
Limitations. The AWD system is based off the 3.3 V6 version of the A604. I do not know if the 3.0 version of the 604 has the removable housing or not. If it does, then I would suggest a rebuilt A604 from Forward Motion (about $850). This tranny uses components from the A606 found behind the 3.5 V6 which stock has 214 HP.
Next limitation. According to a Chryco engineer, the AWD system was not incorporated for the turbo cars due to the torque the turbo motors developed. But since the AWD is an option with the 3.8 V6 and it puts out better than 180 HP, I would think it should handle 200-250 HP. As long as you don't beat on it...
I've driven a short wheelbase AWD Caravan. The owner has altered the vacuum actuation of the overrunning clutch, so he can engage AWD whenever he wants. I will say that on a rough dirt road, it was impressive. I think the rear wheels get less than 40% of the power though. I couldn't get the rear to come around when speeding through a flat corner. However, it did a VERY controllable all wheel drift through it... I only got to spend about 15 minutes with that van and the owner. He doesn't like the 3.3, and plans to drop in a 3.8 when the 3.3 finally dies. He has had the PTU and the RDM replaced in almost 100k miles. He likes it.
Also, I believe the rear axle for the vans is wider than the rest of the EEK based cars. Rear halfshafts may need to be fabricated...
Steven Naugler added:
There seems to be some confusion on how the AWD minivans works. I have a 1997 AWD Grand Caravan and have recently replaced the driveshaft and serviced the rear differential and over running clutch. The viscous coupling is mounted on the forward end of the drive shaft. All power directed to the rear axle runs through the viscous coupling and there is no clutch that either bypasses it or locks it up.
Just forward of the rear differential mounted between the driveshaft and the differential pinion is a unit with an overunning clutch and a dog clutch. When in any forward gear the dog clutch is disengaged. Whenever the front wheels are the same speed or slower than the rear wheels the over running clutch is slipping and no torque is supplied to the rear wheels. Whenever the front wheels are faster than the rear, as when they slip on ice or snow, the over running clutch locks up transferring torque to the rear axle, but only as much as the viscous coupling will supply. When the front axle is only slightly faster than the rear, only a small amount of torque is transferred. When in reverse the dog clutch engages which locks the driveshaft to the rear differential pinion because the over running clutch function is backwards for driving the van in reverse.
I am not a big fan of this low cost AWD design, but it works. I can get around in the rare Delaware snow storm easily when other Chrysler minivans suffer from wheelspin or plain can't move. If only the system could teach physics to the drivers of SUVs that think 4WD and ABS renders them immune to those same laws.
The AWD system does not split torque unevenly left to right. The left and right half shafts have a normal differential identical in function, if not truly identical, to the FWD differential. The driveshaft to the rear is turned by a ring and pinion driven directly from the ring gear (actually a helical gear) of the front axle differential. The ring gear drives the pinion, unlike normal ring and pinions where the pinion drives the ring. This whole ring gear assembly is hollow so that the a shaft from the front axle differential to the right side front axle half shaft can pass through the ring gear without affecting or being affected by power to the rear axle. There are no limited slip differentials in the transaxle.
Bob Sheaves noted:
AWD is NEVER locked up, by definition. 4WD can be either. Experience at JTE (Jeep-Truck Engineering) and others, performing measured testing, has proven that AWD is far more stable than either front or rear wheel drive. Real world testing in racing venues (remember the Audi Quattros, Mitsu EVOs and Subaru WRXs in WRC racing, among others?) has shown that AWD will always outperform equivelent RWD or FWD, due to the very factor I mentioned, balanced tractive effort at the ground contact patch.
The Acura MDX clutches are a cost reduced (read that as "cheap") design obtained from Isuzu. The reaction time of the clutch assembly is between 50 and 72 MS. The viscous type reacts faster, altho the heat buildup from such a launch is abuse. Because of that slippage, the AWD in the Voyager is safer for the untrained driver. The stability program that controls the MDX is "toned down" from the Isuzu application for just that reason.
Having spent seat time in the EVO in Japan, I am fully aware of that system, and how it differs from the MDX and the competition WRC cars.
A viscous coupling cannot "lock up" AT ALL due to fluid shear. There will ALWAYS be slippage of some measurable amount. That is what generates the heat-the shear effect on the fluid molecules.
The electronically controlled clutches will ALSO slip at a certain torque input, before driveline damage occurs. The only way to "lock up" any drive component is to mechanically couple them together, as in a dog clutch (Detroit Locker type), for example, that incorporates mechanical interference in the design. Friction alone will not "lock" anything at ultimate loads. Friction will generate HEAT in a clutch pack also.
To get back to the original point, instead of all this dancing, is that YES I would purchase a Pacifica with a viscous coupling. It is simple, less parts, more reliable, and less cost than ANY electronic system.
FYI- this design was chosen from 3 competing designs, and was done by Steyr in Austria (then owner of what became the Eurostar assembly plant a bit later in Graz).
Simple, cheap, and relatively reliable are all its pluses. Limited power output, component heat, and weight are its downfalls. You pays your money and you takes your choices.
For my wife, it is the best of all worlds-keeps her on the road and doesn't detract too much from fuel economy, compared to other systems. BTW- ours has the infamous A604 and has been absolutely bulletproof since we bought it (it has over 140,000 miles now).