OB99W

It's somewhat unusual for the complaint to be that shudder is a problem when warm, but not cold; often, it's the opposite. So, I'll go out on a limb a bit and make some perhaps-unorthodox suggestions. In particular if this model has a clutch damper, if the hydraulics were opened make sure that things were bled correctly afterward. Is there any possibility if there originally was a damper that it was removed? Maybe the break-in didn't go well; you could try intentionally slipping the clutch just a bit and see if that cleans up the face and improves things. At 148k, if the engine/trans mounts are original, they might need replacement; maybe they got damaged or something was left loose when the clutch work was done.

It's often the front one (although with champagne/beige color, maybe not ). I suggest using an OEM unit for at least the front sensor.

Most cars are designed with hubcentric wheels because forces are better transferred between wheel and hub, and the wheel/tire are more-accurately centered (less radial runout). Even when the centerbore is too large, a wheel will "fit" if the bolt pattern and circle match the original. However, the wheel will be lugcentric, resulting in more stress to the studs and lugs than when the centerbore and hub match (properly transfering load), as well as possibly causing eccentric running and related problems.

The fluid has very high viscosity; it's quite different than gear oil, and is isolated from it. There have been some variations, but one type uses a silicone-based fluid. The fluid's usually not changed, and the later units can't even be disassembled. However, apparently some Australians have been known to open the earlier units... ...see this ("Liberty" = "Legacy"):http://www.mrtrally.com.au/performance/viscous_coupling.htm

If you haven't already done so, verify that the tires are properly inflated and the same circumference (within 1/4 inch). Okay, although the viscous coupler is a separate sealed unit and doesn't share oil with the trans.

If it really is torque bind, some forum members have reported alleviating it in manual-trans models (with viscous coupling) by doing a dozen or so figure-8's while in reverse in an empty parking lot or other safe area.

http://www.troublecodes.net/Subaru/

Probably ; see: http://endwrench.com/images/pdfs/EWPreOBDAug05.pdf

You might find the following informative: http://endwrench.com/images/pdfs/2.2SingleOverWin01.pdf http://endwrench.com/images/pdfs/2.2Liter.pdf http://endwrench.com/images/pdfs/TBeltEWWin05.pdf

You might want to have your tires checked for defects; a shifted or separated belt can cause pulsing/wobbling in the steering and rhythmic noises. If such a tire was originally on the front but then rotated to the rear, the symptoms would change.

I live in a semi-rural area. In order to (hopefully!) avoid hitting deer, I and many others around here often use high-beams at night. Most of us go back to lows when approaching another car, but I certainly have had the need to divert my vision to the roadside at times. Not a bad idea; night vision issues can sometimes be the result of something as simple as a vitamin-A deficiency.

Without knowing details of the leak, here are some things to consider: http://www.ultimatesubaru.org/forum/showthread.php?t=61162 http://www.ultimatesubaru.org/forum/showthread.php?t=67000 http://forums.nasioc.com/forums/showthread.php?t=1016573 Also, a while back there was a run of oil filters that leaked .

Exactly. For example: http://www.sjdiscounttools.com/kdt60612.html

I've got two tires mounted on alloy wheels that currently leak at the bead; there were three that leaked originally. All were dismounted, corrosion (they use a lot of salt on the roads here) "removed", and sealant applied. Two were done by the dealer, and one of those started leaking again shortly after the repair. The third was worked on at a tire shop, and was never "leak free", although the air loss was slowed. The leaky ones loose about 1.5 psi per day. Enough to cause an effective (loaded by vehicle weight) difference in circumference of 1/4 inch. No, I'm not going to get further into that discussion again.

VSS location and exact interaction with the ECU and TCU seem to vary with model type and year. For some info, see: http://endwrench.com/pdf/feb2004pdf/4EAT.pdf http://endwrench.com/images/pdfs/4EATInfo.pdf

That should be telling you something. As I mentioned in your earlier thread, it's unlikely that a tune-up will correct certain problems indicated by the trouble codes you previously reported. In particular, the vehicle speed sensor codes should be addressed. The ECU can't properly manage the engine systems if VSS data is faulty, and it can make it appear as if a tune-up (or more) is needed. When the ECU receives sensor data that is so far out of the usual range that it obviously indicates a problem, if possible the ECU ignores the data and substitutes default values. That is what "limp mode" is, and if your dealer doesn't know that, find someone qualified to work on your car. Since some sensor faults are caused by electrical system problems, I'd start by verifying that the alternator, battery, and associated wiring are in good condition. A poor electrical connection can cause excessive ripple voltage that can lead to sensor data corruption, for example. Not to , but I'd next check the VSS situation, since the speedometer not working is far from the only thing that is related to those sensors. Once that's resolved, move on to the other symptoms if they still exist.

You have a Private Message.

Subaru specifies a 50/50 mix. That not only ensures sufficient freeze-point protection, but also that anti-corrosion and lubrication levels are adequate as well, as has been mentioned. Unless you're willing to monitor pH, etc. of the coolant and adjust additive levels, keeping a 50/50 mix is prudent, simple, and especially if your Subaru is under warranty, meets the requirement. It's true that a 50/50 antifreeze (ethylene glycol) mix has only about 85% of the heat transfer capacity of pure water. However, given that using pure water really isn't an acceptable option, that fact that the boiling point is elevated about 12 degrees F in a 50/50 mix is not moot.

The "100% ready to use antifreeze" you refer to is apparently about 40% antifreeze and 60% water, based on freeze protection to "(-26 celsius)". When you buy antifreeze that isn't premixed, it is 100% antifreeze, and therefore is usually diluted with an equal amount of water to obtain a 50/50 mixture. Yes, it's good to use de-ionized water, and probably better to use distilled; ordinary tap water in many areas is "hard", and disolved minerals (mainly calcium compounds) will come out of solution and form deposits that can at least partially clog passages and compromise heat transfer.

There's potential for lots of problems under this condition. In comparison to a 50% water/antifreeze mixture, oil has only a bit more than half of the specific heat capacity and only about a fifth of the thermal conductivity, in addition to being more viscous. That makes it a very poor engine coolant, so whatever the original cause of the oil in the coolant, the longer the engine was/is run this way the more likely there will be additional damage.

I'm not suggesting that cats never fail; obviously, they sometimes do, but the more-modern honeycomb type are less prone to than the older pelletized kind. When they do go, the honeycombs tend to develop holes, rather than plugs like the pellets. This is where I expect Nipper to suggest that the cat be banged on (not too hard !) to see if it rattles. Honeycomb-type shouldn't (the pellet-bed ones should somewhat, or it usually indicates the pellets have fused together).

True, an exhaust leak is another possibility. A clogged cat causes what any exhaust restriction would -- poor engine performance (limited power, especially at higher RPM, and reduced gas mileage); a vacuum gauge can be useful in making the diagnosis. As I said before, if rear sensor output varies much, that's what sets the P0420 code. So measuring that sensor's output voltage can verify that the ECU is accurately reporting what the rear sensor is doing, but not necessarily determine whether the cat is the cause. Cookie mentioned that P0420 was the result when his front sensor became "slow". That's because its poor response allows more-sustained swings in mixture to occur, resulting in exhaust variations which the cat can't sufficiently cope with, and some of the variation will therefore be seen at the rear sensor. If you're going to measure O2 sensor response, I'd start with the front one. However, be aware that a voltmeter probably won't give easily-interpreted results, since a properly functioning front sensor's output will be constantly cycling at a rate of about twice a second (when installed; they can switch in less than a tenth of a second on a test bench). Therefore an oscilloscope or "scopemeter" is the better tool for the job. If you have access to the equipment, then doing some further troubleshooting makes sense. However, if you have to pay someone to run the tests, replacing sensors might actually be less expensive.

Yes, I read (and agree with) what you wrote, but I wonder who read what I said in post #5 of this thread.

P0420 is the code for "Catalyst system efficiency below threshold ". The ECU determines that by monitoring the voltage output of the second (rear) O2 sensor. The output of the front sensor is supposed to oscillate (once the engine is warm enough for the system to be running closed-loop). If the catalyst is working sufficiently well, the rear sensor output should be pretty constant; if it's not, the ECU flags that as catalyst inefficiency, and sets the P0420 code. Unfortunately, the problem can have a few causes. It could be due to a bad O2 sensor (often the front one, but the rear ones do fail as well), or sometimes the cat itself. A cat doesn't have to be in a state of meltdown to be defective; it can be poisoned by various things in the exhaust, most of which shouldn't be there if the engine is running well, has no internal leaks, no one has used inappropriate "cleaners", etc.

Or not ... for the reasons already mentioned.