OB99W

Congrats on finding and fixing the problem. It wouldn't be typical for the lamps' solder to get hot enough to melt, no matter how long the parking lights were on. Some corrosion could increase contact resistance, which can cause abnormal heat. However, I'd suggest you verify that the lamp type numbers are correct, in case at some time they were replaced with incorrect (higher candlepower) ones. Also, some "cheap" lamps don't meet spec, even if supposedly the correct type.

Hmmm, someone had a 2001 ... http://www.ultimatesubaru.org/forum/showthread.php?t=48309

Some of the weird behavior you've noticed (excessively bright/dim lamps under certain conditions, etc.) is probably due to "back-feeding" of current. Since the fuse is taking 15 minutes to blow, obviously the current being drawn is not much over the rated 10 amps. You're on target with the lamp circuits you've looked at so far. However, "I do not have a trailer hitch or connector." should be reconsidered. It's not uncommon for trailer connectors to get damaged or corroded, and they do exist (but can be hard to locate). See http://www.etrailer.com/question-6780.html .

The fact that the pressure readings are holding after shutting the engine down is good -- it means the injectors aren't leaking. I wouldn't have expected much change in the idle with a small vacuum leak caused by the hose being left open, but we shouldn't gloss over this. You should verify that there is full vacuum at that hose, just in case something is blocking it. However, assuming vacuum is getting there and the gauge you used is in good calibration, it looks like there is a problem with the FPR. The pressure you read is considerably too high -- since your engine vacuum at idle is somewhat lower than normal (17'' versus 19-20"), I'd have expected the pressure to have been slightly above the typical 32 psi -- perhaps 33-34. The pressure with vacuum to the FPR removed, which should be about 43 psi, at 50 psi is also too high -- obviously, engine vacuum has no effect on that. Besides the FPR itself, another thing that can cause the fuel pressure to be excessive is a blocked return line (that's the one that runs from the outlet of the FPR back to the gas tank). However, it's much less likely than the FPR itself being bad. You could depressurize the fuel lines, disconnect the return line from the FPR, remove the gas filler cap, and apply some low pressure/velocity air to the return line (under about 5 psi). That should easily blow any gas in the line back into the tank, and you might be able to hear the air escaping into the tank by listening at the filler. If the air can't push the fuel through the return line, inspect the line. Or, you could take a chance, not check the return line, and just replace the FPR. If you do so, clearing the ECU so that it can start "clean" and relearn settings would be a good idea. It just might have done that. EDIT: If you don't have a source of air for checking the fuel return line, a hand-operated tire pump with gauge could be fairly easily adapted for the purpose. When connected, keep it higher than the return line, in order to minimize the chance that gas will wind up in the pump. Not very expensive: http://www.google.com/products?hl=en&safe=off&q=tire%20pump%20with%20gauge&um=1&ie=UTF-8&sa=N&tab=wf

Since I didn't see an answer to that question -- all Subaru 2.5 liter engines are interference, both DOHC and SOHC.

I don't know specifics of the EJ20G, so the following is based on USDM EJs. The sender near the alternator is for oil pressure. The temperature sensors are on the coolant pipe. There are two -- the one with a single wire is for the gauge, the one that the ECU reads has two wires.

Yes, if you want to get that fuel pressure gauge from a local Walmart, you have to place the order via the net and then elect to have it sent to the store for pickup. It's good that you'll be able to use one in the class.

It seems like you don't have a service manual. Maybe this will help: http://www.freepdfdownload.net/1997-subaru-legacy-service-manual.html

If the fuses are all good, and the dimmer control itself is okay, then the failure might be the Illumination Control Module ("Illumi Con Unit"). Have you recently wired in any accessories?

The inhibitor switch is on the transmission. Make sure you use a new gasket if you swap the IACV.

Once more, just to be sure we're understanding each other. The cables should be adjusted only just tightly enough to remove any slack, but not so much that the cams get rotated. As has already been mentioned, see the attached (click on the thumbnail).

The brake light indicates more than whether the parking brake is engaged. A simple thing to check is the brake fluid level, which if low can trigger the warning light. Erratic gas gauge indications on many Subarus is "business as usual", and typically doesn't indicate a general electrical system problem. The speedometer problem, given the VSS code, is likely related to the VSS. Since you have it on order, I'd suggest installing the new one when it arrives.

You're welcome. Congrats!

If those are the only things out, it seems you have a bad connection to the black/red-stripe wire at pin #2 of connector F42 (on the interior fuse box). See the attachment for the connector diagram -- note that this is as viewed facing the connector (the pins will be "flipped" as you look at the box). Try wiggling that connector while the blower fan is supposed to be running, and see what happens. If that doesn't reveal anything, determine if there's 12 volts with respect to ground at the wire I described.

There shouldn't be any slack when you're done with the adjustment. The idea is to back off until there's slack, then adjust until that slack is just removed, and retighten the nuts.

Anything else not working? Fog lights? Radio?

Yes, the tension of the accelerator and cruise control cables is adjustable. 1) Loosen lock nuts at bracket -- see the attachment 2) Back nuts off so that there's a bit of slack in the cables 3) Readjust as needed -- just remove slack, don't cause cams to rotate --the accelerator cable (1) will seem "tighter" than the cruise (2), because the cruise has a lighter spring on it 4) Retighten nuts, making sure to not change the adjustment(s) in the process This may not be what's causing the code, but it's easy to check.

Check the throttle and cruise control cables, etc., and make sure nothing is sticking or misadjusted. It's not uncommon when an engine is replaced for those cables to be too tight and hold the throttle open a bit.

It seems Subaru is trying to address the HG issue, since a smaller bore could result in a more rigid block (less chance for the cylinders to "walk"). The chain drive theoretically could require less maintenance than a timing belt, assuming the chain guides, etc., are better than they were on some of the H6 engines.

The least I've seen 0-100 psi fuel pressure gauges go for (on eBay, anyway) is about $16 including shipping charges. That's just for the gauge -- you'd then need at least some hose, hose clamps, and a "T". The Equus unit I linked to has all that, plus the nice feature of a built-in way of safely and easily depressurizing after you're done testing. (Sorry if that sounds like an ad. ) By the way, I found a pretty good diagram of a generic fuel injection system, although it omits showing the vacuum line to the FPR. Of course, the Subaru has two fuel rails "in series", rather than the single one shown: http://www.2carpros.com/how_does_it_work/images/fuel_injection_system.jpg I can appreciate your wanting to move this along, and I admire your dedication.

The Holts product is designed to harden when coolant containing it hits air. It shouldn't clog the heater core, unless there's air in the cooling system. The problem is sometimes people use the Holts, but don't verify that the coolant is topped off and remains that way. Agreed -- I've always said that Subaru needed to "step up to the plate" on the HG issue, and any others where it's obvious that the problem is due to an original parts or design failure. Since they're not doing that, the conditioner is at least a partial option.

No, the "conditioner" isn't a fix, but it may extend the useful life of the HGs for a reasonable time. My point was that being "worried" serves little purpose. HG failures on the phase 2 engines don't tend to be catastrophic -- a small leak would likely be plugged by the conditioner, and if you begin to see any significant coolant (or oil) losses, then the gaskets should be replaced. Just don't ignore a leak and allow the engine to overheat.

I suggest you Google "Subaru Cooling System Conditioner", if you're not familiar with the product. Then, use it. If no leak occurs, there's nothing to worry about. If a leak develops after using the conditioner, talk to Subaru of America and see if they can help you.

Getting "codes for the EGR valve" doesn't always indicate the valve itself is at fault. Specifically which codes (Pxxxx) are being triggered?

I've had that sort of thing happen to me in the past, and not just on this site. I now import the post to which I'm responding into a word processor (actually, it's just Windows Notepad), and write my reply. I save that frequently as I'm writing, because power glitches have also "eaten" things before. When I'm done, I copy it to the online reply form. If it gets "lost", I still have my original. (I'll admit it took a few frustrating experiences to get me to do that.) Thanks, tables are good, graphs even better, but there's a limit... I'm not "tired", but I'll admit this is getting "just a bit" drawn out. I was thinking that your Adult Auto class might have a gauge available. If not, Equus has a decent gauge with accessories (no, I don't have stock in the company, I just like some of their products ). As much as I don't like Walmart, they have a good deal on the Equus 3640. If you have a local Walmart, you can order it online and have it sent to the store for pickup. http://www.walmart.com/ip/Equus-3640-Innova-Professional-Fuel-Injection-Pressure-Tester/14644670 For a vacuum pump, there's always an Equus 3660. But seriously, you can probably get away with not using one. I'll explain as I go on (and on, and on...). That isn't quite what I had in mind. Perhaps it would help if I explained how the FPR works: Fuel is pumped from the gas tank, through the filter in the engine compartment, and then into a line that delivers it to the left (driver) side fuel rail to which the injectors are mounted. It then crosses through a line over to the front of the right side fuel rail. The FPR "inlet" is mounted to the rear end of that right side rail, and whatever fuel is excess to maintaining the correct pressure in the rails exits the FPR "outlet" and is returned via a line to the tank. What the "correct" pressure is depends on engine load, determined by the magnitude of the intake vacuum applied to the FPR's diaphragm. I'll describe a procedure that doesn't use a hand-operated vacuum pump (although steps 3 and 4 in the manual are pretty much it). 1) Safely release pressure in the fuel lines. 2) Disconnect the fuel hose/line that goes to the engine from the fuel filter outlet. 3) "T" in the fuel pressure gauge between the fuel filter outlet and the hose feeding the engine. 4) Attach a vacuum gauge to the intake manifold as you've done in the past. 5) Start the engine. 6) Run until the idle is relatively stable, and intake vacuum is at least 17". 7) Note the fuel pressure. It should be about 32 psi. 8) Disconnect the vacuum hose at the FPR and plug the loose hose end. 9) Note the fuel pressure again. It should be about 43 psi. 10) Shut off the engine. 11) Don't immediately disconnect the gauge. Instead, get an idea of how rapidly the pressure drops, which will be an indication of whether any of the injectors are leaking, among other things. 12) After a couple of minutes or so, safely release the fuel pressure again. 13) Remove the pressure and vacuum gauges and reconnect lines as needed. The pressure measured with the FPR's vacuum line attached (step 7) may be slightly higher than normal if the engine vacuum is low. If only all this was that easy. For one example, let's say that (either due to a bad/obstructed FPR or a blocked fuel return line) the pressure going to the injectors is too high. At idle, the amount of fuel needed is considerably less than at higher engine loads. Only so much air is available, via the IACV, to mix with it. If it's really rich, the plugs could be fouling, leading to misfire. (We know, based on the LTFT numbers, that the ECU thinks it's too rich.) With the engine cold enough for the system to be running OL, the default fuel map (not to be confused with the MAP sensor) already calls for enrichment, and if the fuel pressure is too high and even more fuel is therefore injected, the ECU doesn't know that and couldn't do anything about it anyway. For another example, let's assume that the A/F sensor is very sluggish to respond. Then, it might have a hard time keeping up with the changes that occur with closed-throttle deceleration when going from CL to OL-Drive and back again. The response time should be a fraction of a second for decent control. In most cases, one second is too long, and triggers P0133. On the other hand, the further above idle speed/load, the greater the volume of fuel and air passing through the engine. Small leaks, excesses and deficits that impact idle get swamped out, and while everything might not be running exactly as planned, the end result is often hardly noticable. Sometimes (other than drivability problems at low speeds) a loss in gas mileage can be the only clue that something is amiss. We've been talking about changing the A/F sensor throughout much of this thread, but realizing that other things could be the cause of the problem, we did a lot more diagnostic testing and some procedures to rule those out. That sensor is old, has a lot of miles on it, has triggered codes, etc. It should be replaced. I would be delighted to have it resolve the problem. But (yeah, there's always one of those), if the mix is too rich because of a fuel pressure problem, after a short time it can affect the new A/F sensor. So, I'd suggest checking fuel pressure, if not before replacing the A/F sensor, then within some reasonable time afterward. And yes, I think we'd both like to see a happy ending for this sometime soon.