[no694terry]

So i broke something, lost rear wheel power and wasn't convinced it was a speed sensor. I picked up a bad but FREE used transmission that should still have a good extention/tail housing, clutch pack and solenoids.

1) First get the vehicle as high as possible, you will be under it the whole time. I have ramps in my garage that i drive up onto and lifts whole car the about 18 inches under the tires. 24 would have been nicer.

2) while my exhaust cooled done i drained the transmission fluid, and disconnected 02 sensors.

3) drop the exhaust, mine is all one big welded piece so i unbolt it at the heads, undo the rubber mounts and the entire system slides out of the way.

4) time to situate the rear driveshaft, I unbolted the center carrier bearing, and removed the rear u-joint shield hoping to get enough drop to pull driveshaft out of transmission without unbolting the u-joints, no such luck, it had to wait untill i removed transmission cross member.

5) i removed the 4 nuts holding transmission mounts to cross member, i then slid a heavy board from ramp to ramp to place a jack on to support transmission while i unbolt the cross member.

6) now with the cross member out of the way i used the jack to lower the transmission just enough to slide the drive shaft the rest of the way out and the used a bungee to hold it out of the way, then removed the transmission mount from cover mostly just for more room

7) remove the 10mm bolt holding in the speed sensor, place sensor out of way and remove rear cover and find your surprise. mine was that the stamped gear that drives the clutch pack had broken off the output shaft.

8) replace necessary parts clean all the gasket surfaces and prepare to reassemble.
when recombining the clutch pack and basket i found it easiest to carefully slide basket into the clutches, it took a while and was pretty frustrating trying to get all those clutch discs lined up so it all went back together, but with some pressure and twisting motions it all fell in place.

I got all my parts out of this donor transmission

9) i put output shaft and clutch pack back into transmission as one piece and then slid cover back over it, careful to tuck the solenoid plug into the transmission. i didn't and crushed my plug when i put cover back on and that sucked to fix.

10) finish reassembly.

The most time consuming part was cleaning gaskets and then another half hour replacing solenoid plug so in all it took me about 3 hours, I used about 6 quarts of dex III but have to do i final level check yet, but 6 came out so i put six back in. the fluid change seemed to help the shifting, the clutches grab sooner than my old worn out set so it all seems to have gone pretty well.

 

[plain OM]

Thanks for posting this DIY thread. In the past there have been requests for "instructions" on dropping the tail section. The excellent pics showing the interior with the clutch pack and duty solenoid, along with the notes, will be a great help.

the stamped gear that drives the clutch pack had broken off the output shaft

Was the hub (the "stamped gear") originally welded to the reduction gear (photo above) or was it only pressed on?

Sort of suggests that if the transfer clutch locks up, then in addition to the common wheel hop/skip and hesitation symptoms of "torque bind", the transfer clutch hubs might be vulnerable. Or, perhaps the hub was breaking loose and that caused the binding, at least until the hub broke free.

Ref: http://www.subaruoutback.org/forums/66-problems-maintenance/27832-my-awd-just-quit-i-think.html )

 

[no694terry]

The metal on the stamped gear itself tore away where it was weakest at the three indents. As far as being one with the reduction gear or a seperate piece, I will have check and photo later.

edit:

it turns out the hub/stamped gear and reduction gear are supposed one piece. the reduction gear is cast and the hub is steel but they are welded together and share the same cast splines. they are held onto the output shaft by a c-clip. the hub could be replaced with the reduction gear as easy as replacing the duty c by removing the clip and pulling the gear off the shaft. a couple good smacks on a hard surface and the gears and bearing came off the shaft.

 

[plain OM]

The photos of the hub, especially the second-to-last above, show deep grooves where the driven clutch plates bear against the outer "teeth" of the hub. Those grooves can cause the plates to bind by not allowing them to move forward and backwards along the teeth.

That hub is what transfers the tranny output shaft power to the driven clutch plates, and through the clutch to the rear propeller shaft and wheels. The welded joint between the hub and reduction gear can therefore be subjected to a lot of stress, all the more so if the clutch plates are binding.

This is a great thread. Again, my personal thanks for posting this.

Edit: How many miles on the transmission?

 

[no694terry]

i thought i felt something in the drivetrain acting up the last couple months. this transmission has 162k miles and saw a lot of offroad use and a lot of heavy loads, some of those loads close to 2000 pounds in the cargo area. i used this vehicle primary to retrieve rocks and boulders from out of the woods for building ponds and other landscaping. the car has paid for itself 10 times over

http://picasaweb.google.com/116638870540616010470/Pond?feat=directlink

 

[CNY_Dave]

That is some pretty gnarly grooving.

Dave

 

[CNY_Dave]

I just looked at what you do with that poor subie- you're a very very bad man!


Dave

 

[no694terry]

i've been better to subie lately, getting some new 215/75/15 firestone destinations a/t's for her next week. the 08 12k mile forester suspension has been a real nice addition

 

[Mi_ka]

To kick a dead thread...

...what surprises me is the difference in the depth of this super deep indentations: This suggests difference of clutch plates loading!
In similar motorcycle clutches I have never seen this.
Ideas?

 

[ntippet]

Thanks for reviving this thread.
I have not yet seen a hub with such deep grooves.

My take on the differences in groove depth:
Looks like a slight hydraulic pressure drop across the plates.
The rear plates are able to grab quicker and have less of a chance to "slam" into the hub.
But* front plates receive the power first from the transmission, & the pressure from engagement comes from the rear.
Looks like the deepest groove takes the most shock & power from engagement.
It probably exacerbates itself once a small groove has developed.

Didn't see a photo, but were there grooves in the outer basket as well?

 

[no694terry]

Thanks for reviving this thread.
I have not yet seen a hub with such deep grooves.

My take on the differences in groove depth:
Looks like a slight hydraulic pressure drop across the plates.
The rear plates are able to grab quicker and have less of a chance to "slam" into the hub.
But* front plates receive the power first from the transmission, & the pressure from engagement comes from the rear.
Looks like the deepest groove takes the most shock & power from engagement.
It probably exacerbates itself once a small groove has developed.

Didn't see a photo, but were there grooves in the outer basket as well?

I'm going to have to say the grooves were limited to that baskets. I probably would have photo's the other grooves if any. But I don't remember exactly. My brother recently has been doing very strange things with his 98 impress outback with drivetrain changes. I can't even keep up on how many models make up his car. His last mod he was working on was putting a still LSD in his rs rear end?????

 

[Mi_ka]

Oups! Some bells ring!
These grooves suggest a possible harshly abrupt MTP engagement that may be happening to this car and has passed unnoticed due to being loaded. Maybe the car is eligible for some TCU reflash according this old TSB also mentioned in another thread:

NUMBER: 16-66-02
DATE: 03/15/03

APPLICABILITY:
1999-2002MY 4EAT Vehicles

SUBJECT:
TCU Code 45 and AWD Harsh Engagement

INTRODUCTION

If you encounter a TCU Code 45 (intake manifold pressure signal), and have confirmed the wiring and connectors as outlined in the appropriate Service Manual, it will be necessary to replace the TCU with one listed below. Please note that if the same vehicle is experiencing an ECM code for the pressure signal, replacement of the TCU will not correct it. You will need to diagnose and repair according to the Service Manual.

These new TCUs also include a slight logic change to address AWD harsh engagement. Some customers may comment, when they are driving in snow and ice and they try to accelerate during slow speed driving, that they feel a bang from the rear of the vehicle. This is caused by the transfer clutch being applied quickly when the system detects slight wheel slippage. You should confirm the AWD is working properly prior to replacing the TCU.

If you never have noticed any harsh AWD engagement then this TSB should not apply to you and these grooves are just a symptom of the system being stressed too much by your off road rock carrying expeditions. I wonder if any Baja owners have had any similar trouble as the Baja is supposed to be a more agricultural machine - oh wait is the Baja a VTD tranny?

 

[no694terry]

No harsh engagement, weak a slow if any prior to repair. After repair the awd responded much better with seamless engagement. And its a MY96

 

[Herman Felker]

Similar kick.

I suspect I have a similar problem in my 96 outback.

We lifted the car and put in drive we can accelerate hard and the back wheels slightly turn, but can be held still by hand.. (at an auto college in front of instructors haha)

There is a somewhat loud woooing/whirring sound when driving above 35-40 mph.

Has anyone else experienced such symptoms?

Please help! I plan on using the AWD option.

 

[Mi_ka]

I've been off Subaru forums for a while now and forget things but I think that if you lift the car the TCU sees complete lack of engine load and does not actuate the MPT much so not a lot of torque goes to the rear.

 

[CNY_Dave]

The older and non-VDC cars are not that smart- they just go by gear selection and throttle position at higher speeds, and at lower speeds they also look for front or rear wheel slip.

 

[Mi_ka]

Still, I can't remember where I read that car lifted, accelerating moves no real power to the rear wheels.
As logging folks in the forum have found (if I remember well), the MPT is commanded by the TCU to grab more in relation to speed difference of the TC input-output to assess engine load in coherence to throttle opening and speed.
With the car lifted, engine load is low thus weak MPT engagement.
Maybe someone who has a known good MPT and has lifted his car can help more fellow member Herman_Felker.

 

[CNY_Dave]

I have run mine lifted up, all 4 wheels off, and at idle and just off idle in any gear the rear wheels were not driven, any more throttle than that and the rears would decisively engage and spin. The engagement is low, 10%, but with no load they spin right up.

If the rears don't spin at all, there is no question something is wrong.

 

[Mi_ka]

Have you tried a little handbrake while lifted (since holding them by hand can by tricky and dangerous) to verify transfer of adequate torque? This what my reading through the years at the forester forum has recollected, that the torque transfer is minimal while lifted. I never got my dream '07 forester and never will to test this...

 

[canubaru]

95 - 99, there are several situations that will create full power to the rear. (60/40 max.)
WOT
gear selector in 1 (maybe 2) or reverse.
speed differential between the front and rear.

but under normal operation only 10% of the engine power is sent to the rear.

it has taken years to collect this little bit of info.
i do not know how this info relates to the 00 - 04 years, but i do not believe those years are dramatically different.
and even less knowledgeable about it for an H6 VDC.

if you lift the car the TCU sees complete lack of engine load and does not actuate the MPT much so not a lot of torque goes to the rear.

again, for 95 - 99 cars, this is only true as far as the standard 90/10 split goes.
10% is not much compared to 90%.
but the way the system works:
there is NO power to the rear when there is NO fluid pressure on the clutch plates / spool valve. (engine off or fwd fuse in.)
when there is FULL pressure , no regulation by the duty C solenoid, the front and rear are LOCKED together. this causes binding in turns on dry pavement and feels like a 4WD truck with locked hubs on dry pavement. (bad duty c or no power to it.)
regulated pressure, by the duty c and the computer, gives you two things, variable power to the rear from 10% to 40% as needed and an allowance for ''slip'' between the front and the rear in tight turns. this usually eliminates any binding situation in turns. it can not completely compensate for mis-matched tires.

if you think a 90/10 split isn't much and mostly a waste, put in the FWD fuse and drive the car.
you will feel the difference.
without the fuse it feels like a subaru awd.
with the fuse it feels like a honda fwd.
very noticeable in my opinion.

when the car has all four in the air, put the selector in 1 and see if that makes a difference to the rear wheels.