idosubaru

The relay changed the symptoms? What else is in line with that relay that could have blown? And what caused that relay to blow? If two (or more) things blew that’s suspicious. Might it be a good idea to make sure the alt isn’t overcharging before installing a new ECU?

Faikure localations close to combustion chambers are experiencing temp swings of hundreds of degrees. The materials at the localized failure points wont reduce the full 20 degrees either. A lower thermostat sounds diminutive to me. I don’t feel it’s worth pursuing so I don’t care personally but for clarity - GD and Gloyale are suggesting dissimilar reactions to lower Tstat? It will influence ECU or it makes no difference at all?

First you need good diagnosis some of which is easy and you’re skipping over. ***If the car won’t turn over then this has nothing to do with the ECU.*** The starter needs 12 volts and a signal for the solenoid to turn the engine over - and a security system that’s hapoy - that’s it. So it’s just missing 12 volts ***at the starter*** or the solenoid signal if everything is “dim”. 12 volts at the battery is meaningless if the starter or rest of the car doesn’t see that 12 volts. 1. Check voltage of battery and report here 2. Check voltage at the starter. Takes 30 seconds, really easy. Negative lead on negative battery terminal and positive lead on main starter bolt. Report that voltage here. 3. Check main relay in engine compartment and all relays. 4. May need to check into security system issues if it’s been flooded then it may lock the car. All the older security modules used to be located by the drivers side left foot area and was likely flooded. 5. Once the car is getting 12 volts and the dash lights are coming on and engine is turning over - turn the key off then ON (but not to start) and listen carefully immediately afterwards. You should hear some light noise from the fuel pump priming for two seconds. If not then your pump isn’t getting power.

one can google various combinations of "thrown rods EJ25 bearings failure" etc - it's widely known and common. on EJ25D's it's related to the random overheats. car with 30,000 miles overheats and in the 1990's no one was thinking headgaskets, it wasn't known back then and the internet wasn't yet sparkling with insight. in the early 2000's mechanics were calling me about it because they thought I knew a little bit about Subaru's. Typical scenarios - car overheats, limp it home but it's fine the next day so "bad luck" and keep driving. overheats a month later and they limp it home. mechanic finds nothing and changes coolant. wash-rinse-repeat with various combinations of radiator, thermostat, radiator cap, water pump replacements - and the car has been limped and overheated multiple times when all along it was the head gasket failing. that overheating compromises the oil which degrades the bearings that fail sometime later. they can sustain a fair amount of overheating so it's not like they all do that, but it's certainly not uncommon either. they're all rusting away out here but 10 years ago you could buy blown EJ25 vehicles weekly, if not daily, on craigslist.

bolt a carburetor on it and forget the wiring. lol

nope. no wheel can lock. i'm not sure why he would say that, it doesn't make any sense. when it's binding - it's a massive twisting happening from the wheel to the center diff so all that load is being carried by each component from the center diff, rear driveshaft, rear diff, and axle. the weakest link will fail, but not clue how long or what that is in an MT.

i haven't seen one. too complicated of a compilation. in terms of major engine specific issues (not maintenance, timing belts, pulleys, alternators, batteries, starters which will all strand you if they fail...). In a sense if you just pay attention to the long block - that's what matters and gets expensive. Intake manifolds are rarely probelmatic and cheap and easy to repair or maintain, so that's benign in my eyes. So personally i pay attention to the long blocks for reliablity, easy of preventative maintenace and cost: EA engines are great but old and slow. TOD is their big thing. Get one that's never been overheated or oil compromised and they make 200,000 miles easily on typical maintenance items. ER are great but hard to find parts for and not many people are familiar with them. TOD as well and clean the IAC, and address corrosive CTS terminals. EG33 are great but hard to find parts for and not many people are familiar with them. 96 and earlier EJ22's and EJ18's are non interference and routinely make 200,000 miles without blinking on regular maintenance. Knock sensor is most common failure i used to replace these routinely preventatively, CTS probably second most common 1997+ and all EJ25's and newer gen engines are interference so if timing belt breaks they usually bend valves. Timing belt reliability should include excellent on-time maintenance and OEM parts or an OEM supplier (currently Aisin does for some platforms, but not all). EJ25's - headgasket issues, knock sensors as well. 1996-1999 EJ25D's are randomly symptomatic and get worse over time. 1999 Phase II's leak externally 00-04 is usually coolant, 05+ is usually oil, but any of those years can do both and they can burn oil internally as well. EZ engines have higher mileage headgasket issues - usually onset later in life and have the same general predisposition as EJ25D's. No timing belt maintenanace. Later models starting sometime later in the 2000's start having smaller 7mm oil pumps and worse lubrication issues - like oil control rings in early FB engines.

I'm unsure about impreza's - 2004's and earlier you can swap ECU's all day long like batteries and not think about it. 2005 I think starts the change, but sometimes impreza's lag and 2005 was an odd year in a few days. a little googling or call to subaru should answer whether yours needs programmed or not.

cant be incorrect, that’s impossible as its just opinion. yes if you’re talking about buying and comparing an engine right now then any older engine could be problematic for lots of reasons. I’m comparing the engines only, ignoring age and markets. EJ25Ds ate headgaskets, rod bearings and tossed rods through block cases while still under warranty. EA82s were not comparable to that in terms of early expensive failures needing updated and revised parts. though EAs are wonky, gutless and less than ideal quarter century old EA82s are a different topic. if ill maintained cheap craigslist specials, or the current market, or performance, are your thing you’ll find EA engines problematic.

Nah, you’re out of luck. But the engine should be fine. How long have you let it sit since being flooded? Is this from months ago or recently. Take the ECU apart and let it dry out, preferably lay it in a bag of rice or desiccant What year? I can mail you an ECU for $30 if it’s a 1996-1999 legacy/outback Www.car-part.com I’d its 2005 and up you’ll need the dealer to marry your new ECU to the vehicle, keys and FOBs. Roughly $100 for that service

i already answered that question directly earlier, reread my post.

I’ve only ever matched them up, don’t know.

I don't think that matters at all - just call them and ask. insurance is strictly a financial product, not a material physical thing, so to that end it doesn't much matter. they are charging you small premiums at the risk of possibly a large pay out some day. you're paying them to take a risk, they don't really care what the item is, what it's called, or how you define it - that's meaningless. they just need an idea of risk assesment. then they can calculate how much that risk is worth to them. as long as they have some capacity to assign a risk and dollar value in their business model, they should be good to go. a classic car insurer should have plenty of data and experience insuring cars that they'd gladly insure a Subaru. he replied while i was typing - i could see that happening as well if it's such a small segment that it's not worth their time. they're more than likely run into fraud that way if you could insure any old worthless 1990's taurus. you could also ask your current insurance company if you could increase your amount and see what options they may have to not "go by the book".

one could look into it. i'd wonder how many miles did it have when he bought it and were the headgaskets or engine replaced before then, etc? if he bought it new, how does he know it wouldn't have made it that long without the tstat? around here the coolant is cycling between -10 and 180 degrees (or whatever it is) and so with a 160 degree tstat it would still be cycling between -10 and 180 degrees. that's a 190 degree temp swing compared to a 170 degree temp swing, a rather meager difference. and, it is localized materials temp gradients that matter, not coolant temperatures. so to use your 'bending clip' example - materials can bend nearly indefinitiely in their elastic range. they fail when they exceed elastic limits (which is presumed in your example) and enter plastic deformation. so correspondingly we would need to know what that "elasticity" of the temperature swings are to determine if the range difference matters. so, maybe, but i think the engineering perspective is more nuanced and less convincing than armchair engineering, as is usually the case.

oh right you already checked power - so yep be interesting to see what you get from the ground pin to it's associated pin on the ECU. i've seen rodent damage more than once - but tracing will lead you to that if that's the case. one was in an apartment complex with zillions of chipmunks and another was a vehicle with 3 kids and food particles everywhere around the car seats.

No. how many EJ25's has he personally done this too and then followed for a decade and 100,000+ miles? My guess is it's probably an anecdotal amount, if any at all. he probably works on dodge neons, or knows a friend who works at a shop.... someone could say they always use a certain brand of coolant, or always assemble the engine when it's above 50 degrees ambient temps, or they always use crest toothpaste, or always use XYZ brand oil- and never have headgasket failures. just because it's "true" (for them) or "it happened" or "they saw it" doesn't hold any quantitative significance.

The ECU fires the injector so, I'm weak at electrical, but from here I think you check wiring from injector to ECU and wherever the other side pulls 12 volts from. You pulled the engine - I haven't seen this happen on newer subaru's or new gen models but make sure none of the main harness connector pins are bent, damaged, corroded. If this engine ran 10 years with no misfires and has a misfire after the engine pull then it seems likely something happened during that work. There are accounts of H6 ECU injector driving circuits failing, so that's a definite possibility though you're headed in the right direction by diagnosing. Used ECU for $50 - $150 and then $100-ish for the dealer to program it and all your keys/FOB's at the same time. New ECU's i think are $500+ General Disorder, GLoyale, Numbchux, Fairtax are members here that would have something good to say about that. This stuff is generally beyond by proficiency.

Nice work on the diagnosis. It sounds like you’ve verified it’s not the injector but would it be worth swapping injectors just in case? injector not firing is unfamiliar territory, be interesting to see what others say to check next.

Great call - separated crank pulley.

I’ve seen comments of people replacing the compressor bearing but may depend which compressor you have if it’s accessible or not.

I hesitate to derail this pinout discussion but you mentioned a toggle switch and it seems like you got this car from a previous owner that dId some other custom wiring, you mentioned grounds. I’d be checking that toggle wiring and other edited wiring. No one here would be shocked, or hasn’t seen, custom wiring be an issue. I’d also be verifying the year/part of the vehicle, engine, ECU and other components. Maybe this thing was swapped or wrong parts were installed before you got it?

Knock sensor - most common EJ sensor failure CTS - second most common sensor failure Plugs/wires - should be NGK or OEM Lazy front O2 sensor - almost a quarter century old now Leaky injector(s) Dragging brakes from sticky slides. And of course weather/winter gas can only decrease mileage, not increase it.

Thanks FT. I’ve placed the atf suction side in fresh fluid and the output hose dumping into an empty catch pan and cranked the car over to move fluid but it’s cumbersome - is there any merit to that or no benefit over drain and refills I normally do?

the issue i mentioned is 00-04 related, so not yours. 1. look for bubbles after you've significantly heard the noise 2. change the fluid? 3. replace orings/hoses close to the pump - they don't have to leak, they simply suck in air. Subaru power steering failure is rare, it's common to replace one and still have symptoms because people swear it's gotta be the pump. That said - used pumps are cheap since they never fail there's zero demand. www.car-part.com, ebay, etc.

Technically there is no such thing as a failed transmission, that’s a huge generalization that’s all but meaningless. A particular part has to fail - front diff, gear, synchro, a bearing.... What are they saying failed? And how did they diagnose that on a car that doesn’t run?