[plain OM]

When the FWD fuse is in place the symptoms are gone.

This would still seem to be the key.

If there's no symptoms with the fuse in, then there's probably torque build-up in the drive train.

When the engine is running with the transmission is in Neutral, the duty cycle signal to the AWD solenoid is held at around 5%, which is very low. At this rate the clutch is probably not capable of transmitting much, if any, torque.

However, when the transmission is changed to a drive gear (D, or R), the duty cycle is raised to approximately 30%; it will go above this depending on driving conditions, but not below.

Normally, 30% should be sufficient to transfer power to the rear, but not be enough to lock the clutch. In other words, at the 30% duty ratio, the clutch will allow some slippage. This is necessary to accommodate steering corrections, bumps, and tight turning, where the front and rear drives turn at different rates.

If there's a "kick" when the gear lever is subsequently moved from D to N, then that could be the release of torque build up as the clutch duty cycle reverts to 5% and the clutch disengages.

The pressure on the clutch is controlled by the AWD duty solenoid in concert with the AWD transfer valve. The solenoid, which is pulsed by the duty cycle signal, directs hydraulic pilot pressure to the transfer valve. The pilot pressure operates the transfer valve which controls line pressure to the clutch. The transfer valve or solenoid could be defective such that either tends to over-respond at the 30% duty cycle rate, maintaining more pressure on the clutch than there should be.

The test for this would be to measure the hydraulic pressure to the transfer clutch at different duty cycle rates. The FSM has the specs, and there's a port on the side of the extension case specifically for this.

I don't have the details for the 2005 4EAT, but I've attached the pressure test page from the 2007 4-speed to provide an idea. The specs might be somewhat different for 2005, but it gives an idea of what to expect. If the readings are significantly higher at lower duty ratios, a faulty solenoid or transfer valve could well be the problem.

If this is the problem, then it seems to me there should also be torque bind symptoms when turning in a tight circle, and I can't explain why that doesn't seem to be the case. Nevertheless, it would probably be worthwhile to better establish if there's incorrect pressure on the clutch before jumping to disassemble the transmission or valve body.

I understand the solenoids in the valve body of the later Phase II 4EAT, including the AWD solenoid, cannot be changed individually. But perhaps someone has found a way.

p.s. when was the last time the ATF was replaced? Did you try draining and filling the fluid a few times to clear the original torque bind symptoms?

 

[plain OM]

There is definitely a bind in the drivetrain. The car will make sharp turns but when the wheel is cranked all the way the car has to be assisted through the turns with a little throttle.

What symptoms were there, if not these, that led you to replace the transfer clutch?

Back to some fundamentals. How many miles on the car? Have you owned it since new? Has the ATF been changed periodically? Was the ATF replaced when the extension case was removed to change the transfer clutch? Who did the AWD clutch replacement? Any other significant work done on the car? Are there any warning lights on, or flashing, in the instrument panel when the car is started or being driven?

. . . try changing the tranny fluid a few times

There would be no harm in doing several drain and refills of the ATF with a bit of driving between each, especially if the fluid hasn't been changed before, it's been a long time since the last change, or the car has been used in severe service such as towing. This might work if something, such as a valve, is sticking due to some sort of build up gums or varnish -- the hope is that the new ATF, with fresh detergents etc, will dissolve and remove the contaminant. There have been some reports of positive results so it's worth a try. But if the problem is mechanical there will be no real change. See http://www.subaruoutback.org/forums...anging-automatic-transmission-fluid-easy.html for more information on changing the ATF.

It might be helpful to also make sure that the duty ratio signal being sent to the AWD solenoid is good in the first place. If it's incorrect, then the pressure will be as well. The signal is generated by the transmission control module and can be read at the OBD connector by scanners able to access the TCM data (not all can). A dealer can do this with the Subaru Select Monitor. But you can do so as well. See http://www.subaruoutback.org/forums...426-freessm-complete-access-your-ecm-tcu.html, where two programs, FreeSSM and RomRaider Logger, were used to read the AWD duty cycle ratio and see how it changed under different conditions. They should work with your 2005. A quick test would be to see if the ratio is at 5% or so with the ignition at ON in Neutral or Park (no need to actually start the engine), and at around 30% when the lever is moved to D. (Then see how it varies under different engine on conditions, as in the thread.)

There's always the possibility that the clutch wasn't assembled correctly. For example, there's a spring that pushes back against the clutch piston so that the clutch is released when the hydraulic pressure drops. In addition, the clutch plates have to have a gap to allow them to separate. But if the spring is weak, or if the plates were installed too tightly, or if the inner or outer "drum" that the plates spline into is badly grooved, it might be possible for the clutch to bind even at lower pressures.

These are possibilities, not conclusions. I think some diagnostic tests to see if anything is clearly amiss would be a start.