forester2002s

+1 on double-checking the bulbs. Have you tried switching the two bulbs from left to right? If the problem persists unchanged, then it's not the bulb. If, however, the bad low-beam switches to the other side, then it's the bulb.

Make marks on all of the toothed-pulleys, and on the engine block in line with these marks. And then make marks on the old timing-belt, at the toothed-pulley marks. And finally, take a photo of this set-up. Then mark the new timing-belt with marks to match the old belt. And away you go!

Could it be that the complete hatch itself is rattling? If so, wrap some electrical tape around the metal 'loop' that the latch mechanism locks on to.

The change in idle-speed with defrost/vent settings may be normal, and due to the AC compressor being engaged.

I once had an annoying slow leak on one tire that was difficult to find. This was probably not what happened to the OP, since that involved two sets of tires. But in my case, what I eventually discovered was a small bar-code label, that was accidently left stuck on to the tire's bead. At first, I couldn't find it. I had no facility to totally submerge the tire in a water bath - that's the best way to find a leak. So I tried hosing the wheel and tire with water, and checking for bubbles. At first nothing; and then about a year later, when I tried the hose again, I noticed a very small bubble forming at the rim-seal. Sure enough, when I deflated the tire, and broke the seal, there was the offending bar-code label (about 1cm x 4cm).

The diff could have been very low on oil. Low enough, that the remaining oil was not wetting the rotating parts. And that would have caused the diff to seize-up and 'explode'. And I can't see on the photo where the oil is actually leaking out; does the crack extend down near to the bottom of the case? And is that where the remaining oil is leaking out? As others have said, these diffs don't usually leak much oil, and can go for years without attention. But yours may be 'the exception that proves the rule'.

I would change, not just the engine oil, but the transmission and rear-diff oil as well. Cheap & easy to do.

I've used the 'Starter-Bump' method, for the crank-pulley bolt. Works first-time, every-time.

Just noticed that you posted this twice. Here's what I wrote in the other post: A bit of math: 1/4" tread-depth is 1/2" wear on diameter. And that is pi x 1/2" = about 1.5" on circumference. Definitely a mismatch!

A bit of math: 1/4" tread-depth is 1/2" wear on diameter. And that is pi x 1/2" = about 1.5" on circumference Definitely a mismatch!

The release lever may have cracked. It happens sometimes. If so, it means removing either the engine or gearbox for access.

The individual grid lines are arranged electrically in parallel. That means that each grid line has 12V across it. One line can break, and the others can still work. If nothing works, then the break is likely in the feed to all of the grid lines. Check for voltage in the wider vertical feed lines on the left and right. If you have an aftermarket film on the glass, you may still be able to check for voltage by piercing through the film with the sharp end of an electrical probe.

If the boots are the originals, why not replace them while you've got the axle out? Especially the inner-boot, since that is the one that often fails (due to heat from the nearby exhaust).

What happens if you leave your remote at home? i.e. out of range of the car?

Or even simpler: Is the gas cap screwed on properly?

Could be a faulty plug. You could try switching that #1 plug with another, and see what happens.

It is interesting that this happens when both of: - Low fuel-tank-level; and - Going uphill. If you could confirm these two juxtapositions, then it would lead you closer to the solution.

I check my car's fuel-consumption by averaging over 5 fills. I call this a 5-fill moving-average, and I graph this on a spreadsheet. I ignore the fuel-consumption from a single fill. I cannot control how full the tank gets, especially how much fuel gets up into the filler tube. This 5-fill method is best for monitoring long-term changes in fuel consumption, but does also help in the short term too, just not after a single fill.

A faulty Knock Sensor might give you similar behavior. Although I would expect to also see a Knock Sensor CEL come up. I've had good experience with aftermarket Knock Sensors from EBay. And some are not too expensive.

I am getting ready to replace the head-gaskets on my non-turbo EJ251. I am going to use new turbo-gaskets (#11044AA642), which I think are slightly thicker than the old OEM gaskets. How should I torque the new head-gaskets? - Should I use the 'non-turbo' torques (since this is my engine)? - Or should I use the 'turbo' torques (to match the turbo-gaskets)?

My 2002 Forester has a rheostat-type control, that dims the panel-lights. There's a 'ring' on the left-side control-stalk. Turn it, and the panel-lights can be dimmed, and turned completely off (I think).

I doubt that the coeff of thermal expansion changes much between different brands of antifreeze. Could be a head-gasket leak. Or it might just be an air-bubble in the coolant, expanding when hot, and filling the overflow bottle. Get that air-bubble out, and you may be good to go.

Year? Model?

This is a good question. On my EJ251, the head-gasket leaks are typically on the outer-bottom edge. So I could rephrase the question: Why aren't the outer-bolts torqued slightly higher?

You need Subaru part #11821AA450 ($14). How do I know? I had exactly the same experience as you. The plastic tee broke when I was trying to get to the PCV. I epoxy-glued the old tee together (so that I could drive to the dealer). 9-months and 17,000km later, I still haven't installed the new tee. My epoxy work seems to be holding just fine.