DaveT

This reminds me of an experience from long ago... I had a 1976 4WD wagon. It got hit, right front tire pushed back to the door pillar. It was before I started doing all repairs, and insurance was involved so I had it fixed at a local collision place, that had a good reputation. Something was never quite right afterwords. Subtle, and intermittent. Some time later, I decided it needed new strut cartridges, so I bought a pair, and went about installing them. Yes, back then, the struts could actually be disassembled and rebuilt, no big deal. So I was dropping the cartridge into the right strut, and it stopped short of going all the way in. The strut tube was bent. Just enough. New strut from a junkyard, cartridge went right in, mysterious problem went away. Learned about hidden damage that day.

I've been running and repairing Subarus since the mid 80s. Mostly EA82s. I've seen and repaired a fair number of head gasket failures. All began as a slow tiny stream of bubbles in the recovery tank. A couple were not slow, due to more extreme overheats while low on coolant. It could be a slow as 1 per second, even slower. All progressed to faster bigger streams. Some degraded more quickly than others. In the ones that degraded very slowly, there were traces of black soot at the waterline in the recovery tank. New head gaskets fixed the problem. Yes, if there is some trapped air in the coolant system, there will be bubbles while the temperature goes from room to operating, but that is not a continuous steady stream, it stops. If there is only a bit of air from incomplete fill, it will work out over a few drive cycles. If there is a tiny leak, there may be no trace but the air never getting out, and the coolant level slowly dropping. These don't cause a steady stream of bubbles after the engine is at operating temp. I've also experienced and read of others, that once you get an overheat due to low coolant, the head gasket/s are damaged, and the bubble leak can be slow, or fast, or anywhere in between. I know of no other source of "air" at pressure in the engine than that from the compression / combustion cycles in the cylinders that can force it's way into the coolant system.

+1 to WoodsWagon.

I haven't seen an o2 sensor cause crummy drivability. I've seen them cause a trouble code, burn out a catalytic, poor fuel economy. Unrelated, but at 100 miles you are due for timing belt and idlers.

+1 check the coolant IN the radiator. It is typical of the leaking headgasket to push coolant into the overflow tank. Since new air is added constantly while running, the coolant does not get sucked back into the radiator during cool down when the engine is off. Running the engine while low on coolant makes the headgaskets fail more, if the temp gauge shows above normal temp.

I used an aftermarket cat on our 2001. Ran fine, lasted long enough. Unfortunately, long ago enough that I have no idea what brand. But most likely, not the cheapest - I don't buy solely by price. Cats failing don't usually cause crappy running, unless the failure mode is that it is clogged up, which usually indicates something else is wrong, overloading the cat. A way you can test for flow - loosen the nuts that hold the Y pipe to the exhaust ports. [beware if these have never been touched, and everything is all rusted up] [mine get antisiezed and are easily removed for such checks] Drop the Y pipe enough to have a 1/2" gap between the block and the flanges. Get a vacuum cleaner, with a hose, and arrange things so you can connect it to the outlet of the exhaust pipe. The idea is so that you can connect and disconnect quickly. Listen for the pitch of the vacuum cleaner's motor. It should only change a slight bit higher, if at all, between disconnected and connected. If the cat is clogged, the pitch will rise noticeably.

If the headgaskets have never been replaced, it's possible the oil channel that goes from the block through the head to the cam tower is leaking. All of my original EA82s developed that leak a long time ago. Other common leaks - camshaft seals, and the o rings on the little pieces they are pressed into. Rocker cover seals. Front or rear main can leak, but I haven't seen that to be common. Oil pump o rings and seal. You have to clean everything well, and I've heard of fluorescent dye you can use to trace it. If your engine is all original, it probably needs all of them, or will soon.

This should be moved to the correct forum.. I accidentally went to this forum for the older cars, and found this thread. A lot of new axles are crap, so maybe the clicking is from one. Also, if the nuts are not tight. Test the coolant temperature sensor. It can fail without causing a code. Is this a CA car or all other states? For non CA, the code 35 is only checking the coil of the egr solenoid. If it is open circuit, or removed, you will get that code.

The blown headgasket can be anything from a tiny pinhole, that makes a small bubble every few seconds - which is the only symptom. to very blown, where it could effect a compression test. I have also seen pinhole leaks in radiator s, or slightly loose hoses that let coolant seep out while running, with almost no trace, but allow air into the system on the cool down. This does not cause the bubbles that a pinhole to the combustion chamber does.

#1 test. Check for continuous bubbles in the coolant - radiator cap open, or in the recovery tank. Make sure the system is full, start with the engine cold. Let it idle at least until it starts to get warmed up. OR until you see continuous bubbles. IF you see continuous bubbles, the head gasket/s is/are blown. The most common head gasket failure in Subarus is compression into coolant system. Once there is enough air in the system, it can't cool, and causes overheat. Do the fan/s run when it start to get over normal temp? If your systems is not showing signs of always having air in the coolant, something else is wrong. IS the air path through the radiator clean & clear? Inspect the tiny fins that are between the tubes. Are they all still there and solidly attached to the tubes? Flushing does not test the flow of the coolant path. How I have checked that is block the lower hose. Fill with water. Remove block quickly. The water should just gush out fast like it was all coming from 1 solid tank.

try a 1-2 second shot of carb cleaner into the throttle body, then try cranking. If it fires momentarily, spark, timing, etc are ok. If the timing belt is original, it is overdue for replacement, but this is another topic.

The ones I've rebuilt, I sanded until gone. Just be sure to be even to keep the face flat. An overheat event could accelerate problems. If the engine was OEM / stock / original, it was pretty much overdue for those seals to be leaking. And pretty much all of the other rubber seals. I've not tried gasket maker for the oil pan.. If whatever type you try is rated for that, and follow all the directions, I guess it should work. Getting the surfaces truly clean is probably the key.

It is normal. Use a flat bar and sandpaper to resurface and remove it from the block. Use sandpaper stuck to a shwet of glass to resurface the heads. Search for " post apocalyptic head resurfacing " for details.

I could buy 2, if the idea of more than one helps them get building...

It is entirely possible that the rear ones are different. I am mostly going by what the fsm specifically warns about, as I have never tried to force one back in. For clarity, I'll add that by "no brake" above I was referring only to the parking brake, not the driving brakes.

Exactly. The spindle thing is threaded. So that the handbrake compensates for the pad wear. If it got crushed bent, it could cause the no brakeing. The fsm warns against forcing the piston. If the lever isn't right, maybe assembly error, or maybe crunched up spindle.

This looks like one: https://shop.performanceradiator.com/index.php?route=product/product&path=1_32557_32785_32786_32787&product_id=10727

It sounds very much like the handbrake adjuster screw / spindle has been damaged. It is inside the piston, so the entire caliper must be disassembled to inspect it. I am not sure about parts availability. Other than getting a used or rebuilt caliper.

Are you saying that you broke a c clamp on one on your car now? That would ruin the parking brake system in that caliper.

If the fans don't run when the car isn't moving, it will overheat. The thermoswitch on the radiator or it's wiring or relay, etc. Need to be checked.

Stumbled across this page, claiming an 88 XT6 radiator is the same as an EA82T radiator. Not sure if that will get different matches... http://bbs.legacycentral.org/viewtopic.php?f=25&t=49123&start=350

Nuts. I looked at the feedback. All 3 located in China. The feedback is overall very positive, BUT... Unfortunately, the feedback does not specifically tie to the listed items, so it's hard to tell which complaints are about which products.

It's odd how those 3 sellers look so similar. I found a website going into all kinds of pro & con of Al vs Copper. Looks like Al is better corrosion resistance. Copper is easier to repair, but typically, I've had radiators fail by the tiny fins corroding away from the tubes, which is not repairable, except by re-coring. I have to contact them to find out about shipping. I must say, that the OEM Al & plastic radiator in our 2001 Forester was still good when we sold it in 2015. Longer than any copper / brass radiator I ever had.

How do aluminum ones withstand salt compared to copper / brass ones.

I have yet to use Rock. Some seem to have good luck, some not so much. Some of the replacement part companies cheap out on everything. That is a risk of only looking at the price. The way I look at it - I am saving tons of money running these old cars. And doing the repairs. I can spend for the better / best parts, and still be way ahead. Reliability & longevity. Just keep a close eye on the radiator. My guess, is it just might not last as long as a beefy one.