Coolant ECU reflash fixed 6 MT dead pedal feel

[DieselOx]

http://www.automd.com/tsb/bulletin_b346660/

Just got back form the dealer after they did the update for the "coolant warning light..." ECU re-flash. We just got the postcard about it this week from Subaru, that one, which ever it is.

I declare the dead pedal feel entirely cured. I thought I would have to live with this forever, that the 6-MT was such a low volume unit that Subaru wouldn't waste resources trying to fix it.

Here's how I know they fixed it. It all started when I continued to kill the engine 3 to 8 times a month when taking off from a stop, never learning the feel of it. I've been driving a stick for 30 years, and have never taken this long to get used to a new clutch, so I started trying to figure out what the heck was going on.

I'd been complaining about the dead pedal (1 to 2 second throttle response delay) for months; eventually, I carefully noted a peculiar behavior during a specific kind of idle dip when stopped with the clutch disengaged (repeatable by holding all 4 window controls up at the same time, while turning the wheel slightly, slowly back and forth until the idle dips dramatically, almost killing the engine-the power steering idle kick up cycle sometimes takes a second to catch up to the extra demand).

During the idle dip, the throttle is dead until the dip is over. I tested this dozens of times until I knew exactly how it felt. The ECU was not listening to the throttle at all during the idle dip, but gives throttle response back to you suddenly in a burst after catching up (by this time you have the pedal half way to the floor trying to get going).

After the update, my first inkling something was different was the clutch just felt like any old normal clutch, no more hesitation; my wife even commented that she thought it was much easier and smoother taking off from a stop, before I ever said anything, or had driven the car at all (she picked it up at the dealer and met me at home).

Amazed, I tested it by creating several idle dips and launched the car right in the middle to see if it could possibly have been fixed. Sure enough, the throttle response is immediate and sure. There is also almost no galloping during initial acceleration anymore, which was pretty common before.

In short, I don't feel like a teenager learning how to use a clutch every time I take off from a stop. I am fully convinced that's it's been totally fixed. I actually feel like there's a weight lifted off my chest I didn't even know was there. This is the first car I ever bought new, and I guess I felt worse about this than I was letting myself believe.

Sorry about the long post, and thanks for letting me get this off my chest. I've always loved this car, and now it's unequivocal.

If anyone with a pre 2013 6-MT still hates the throttle response, go get this re-flash, you'll love your car again.

[this car has a Jan 2012 build date]

[DieselOx]

After reading about the problem the brz had with idle control, it seems a reflash, even with the same software version, can reset the adaptive learning of the ecu, so it can un learn my wife bad habits (she has the lightest foot I've ever seen).

It colud be that ours is now re learning our habits, so she's agreed to drive like a teenage jerk for the next hundred miles (the learning phase) just to get the widest possible "normal" established.

Link to brz problem explanation, sounds very similar, minus the cam timing issues:

http://www.autoweek.com/article/2012...news/121009939

[Novablue]

That's exactly what I was thinking when I read your post. ECU's do weird things sometimes, we have this problem at work from one driver to another all the time.

[Ztech]

While the BRZ may have a different ECU, the Outback's ECU has only limited "adaptive" learning capabilities. Long term fuel trims and timing advance are the only parameters that can vary from the baseline table values. Long term fuel trims will be adjusted for persistent rich or lean conditions and ignition timing maximum advance will be retarded for persistent input from the motors knock sensor. An example for LTFT would be engine operation in much colder than normal temps, the colder air is denser which leans the AFR so the ECU bumps up the long term fuel trim to compensate. For knock retard, a tank of poor fuel might cause persistent knock so the ECU will lower the max timing advance as knocks sensors can't eliminate all pre-ignition in some conditions. The ECU will hold these values for a few hundred miles after the conditions that caused them are removed and then revert to baseline values.

Modern ECU's are fast enough to modify engine parameters dynamically or on the fly so to speak so there's no need to change baseline values for different driver inputs - ie granny vs leadfoot ect.

[DieselOx]

If you're right, then it's the software update that fixed it, because it's a totally different car now. What's your source for this info?