OB99W

I got that. So you pulled out all of the illumination "bulbs"? There are quite a few of them, and doing that would be fairly time consuming. If you didn't remove all of them, since they're in parallel, even one left in place would conduct so that it would appear that there was power applied to both leads when the "low" side wasn't being pulled towards ground. I'm sorry if you found my suggestion that you read up on Ohm's Law impertinent. It was made to be helpful to someone who appeared to not be well versed about it. I'm certainly not going to get into a "credentials" contest with you -- I've purposely avoided ever posting mine, and it's not because they'd look inadequate. It seems you're equipped to handle the job at hand. You could be a lot closer to 100% certainty if you posted the result of the measurements I suggested, but it's certainly easier to try swapping out the module. They were just trying to help, as I am. The difference is, I'm apparently slightly more knowledgable.

Good (assumedly rhetorical ) questions. Unfortunately, I suspect the lack of industry "reponse" is a means of ducking "responsibility". Since the issue doesn't affect current production vehicles, if anyone chooses the "wrong" ATF for an older car and has a trans problem related to that choice, there's really no accountability. "Sorry, but our fluid meets specifications, your transmission failed due to its age/mileage."

Welcome to the forum. Perhaps That's not weird at all. Power is applied to one side of all of the lamps. If the other side isn't grounded, you'll read approximately the same voltage there, since there's minimal voltage drop across the filaments if little current is drawn by the voltmeter. (Decent modern voltmeters have a very high input resistance.) If you want to better understand this, Google "Ohm's Law". Yes, effectively. The dimmer modules (Subaru calls them "Illumination Control Unit", or "Illumi Con Unit") do sometimes fail, especially if someone ties into the dimmer line incorrectly, as when improperly installing an aftermarket radio. However, there can certainly be other circuit faults that can have the same effect. You could just try replacing the module, if the cost (new or used) isn't too much of an issue. On the other hand, if you'd like help with properly diagnosing the problem, here's what you could do: 1) Find the Illumi Con Unit (dimmer module) 2) Connect the neg lead of your voltmeter to a good chassis ground 3) Back-probe each of the pins at the module connector, and make note of the voltage, with the stalk control turned fully up and fully down. Note the connector pin numbers, wire colors (including stripe), etc. 4) Let us know what you get in step #3

If you don't find the problem by doing what Fairtax4me suggested (checking SBF-2, etc.) in the previous post: Check for 12 volts to ground at the OBD2 connector, pin #1 -- it's in the upper row, first contact at the left. (You might need a fine meter probe tip to make the connection.) Let us know what you read there.

My post in this thread concerning Dexron-VI was in reply to 2001outbackwag. It was sort of a continuation of the thread he previously posted, where his question would probably have been more on topic. See: http://www.ultimatesubaru.org/forum/showthread.php?t=117922 You'll notice that I'm well aware of Dexron licensing and related issues. I suggested sticking to "III" (meaning MD-3, DM-3, or whatever is equivalent) as opposed to Dexron-VI, which may not be a good choice.

Welcome to the forum. The ABS warning light is controlled by the ABS ECU. The AT Oil Temp warning light is controlled by the TCU. Both of those control units can misbehave if the voltage supply drops sufficiently. Your Subaru specialist might be on the right track. Loss of alternator field current on Subarus typically causes the battery and brake warning lights to illuminate, but their not coming on doesn't guarantee that there isn't a battery or charging system problem. One thing you could try, to see if you can 'provoke' the problem, is to heavily load the electrical system. That can be done by simultaneously turning on several things that draw substantial current -- headlight high beams, rear defogger, heated seats (if you have them), etc. By the way, was it unusually cold last night in WA? (I've heard that the weather there has been a bit strange lately.) Marginal electrical connections can become worse due to contraction of metal contacts with cold.

Citgo says, in the "APPLICATIONS" section of http://www.docs.citgo.com/msds_pi/10165.pdf : "DEXRON-VI is not suitable for use in non-GM vehicles which called for DEXRON-II or DEXRON-III type fluids." Although other Dexron-VI suppliers don't agree, I'd stick with III to be safe.

A possible cause of radio interference and flickering lights is worn brushes and/or slip rings in the alternator. Sometimes you can verify that by tapping on the alternator while listening to the radio, and seeing if the interference changes in level.

Google "alloy wheel cleaner" (without the quotes). No cleaner is going to fix the corrosion near the hub and bead area, which is usually due to road salt. The rims could be stripped, wire brushed and/or sanded where needed, and re-clear-coated. Google "wheel clear coat". However, you might first want to verify that the bead area isn't so corroded that the wheels don't hold air. Since it appears that tires are mounted, I'd suggest inflating, waiting a few days, and rechecking the pressure (at about the same temperature). If leaking, you can mount tires using sealant if the surface isn't too rough/porous, but if bad enough even that won't help, and you'll be constantly refilling the tires. If the wheels hold air, and you don't care if the wheel finish looks uniform, you could do spot repair of the corroded areas.

Do you have 12 volts at pin #2 (yellow wire) of the ignition coil pack connector when the ignition key is ON?

One of the best ways to check ATF color is to put a drop of it on a white paper towel or napkin.

If it doesn't do it at idle speed, try again with left foot firmly on brake, right on accelerator, holding the engine revs a bit higher ("fast idle", about 1200 RPM). Doing that, there might be a bit of shift shock, but it shouldn't clunk if nothing's loose.

Neither Dexron II nor III is licensed anymore by GM. The currently available equivalent for Dexron-III/Mercon is usually sold as MD-3 (or sometimes DM-3). EDIT: You might find this older thread interesting: http://www.ultimatesubaru.org/forum/showthread.php?t=84146

Welcome to the forum. See http://www.ultimatesubaru.org/forum/showthread.php?p=843909 .

With your foot on the brake, does it "clunk" if you shift back and forth between reverse-neutral-drive?

Since it appears that the expense of the part is important to you, I'd suggest trying to find someone with an ohmmeter to help you verify the original sensor is bad before replacing it.

You could check for corrosion at the wheel sensor connector, or damaged wiring. With an ohmmeter, you could check the resistance of the working rear sensor versus the non-working one.

Yes, John, see post #6 of this thread and the link to the NASIOC thread at the end (see post #40 of that thread). It shows where to find the connector (for a particular arrangement), how to read the codes, and what they mean. I'm sure bstone was well-intentioned in trying to "make it simple", but sometimes that leaves out important info. The pages shown in the NASIOC thread has step-by-step instructions, and complete info. Subaru has moved the location of the diagnosis connector a few times, and admittedly even the factory manual can be wrong in that respect. Once you find the connector, the rest of the factory info can be applied, and it's pretty straightforward. Yes, that's correct to retrieve the ABS code(s). You left out part of what code 27 signifies. Yes, it refers to the rear left ABS sensor, but also importantly "Abnormal ABS sensor (Open circuit or input voltage too high)". That means either the sensor itself is electrically bad, or there's a bad connection to it, somewhere between it and the ABS ECU. The suggestions to clean the sensor or tone ring could be useful if the problem was a bad signal (code 28), but not in this case. If you know how to use an ohmmeter, or have a friend who does, you could do some further checking. Otherwise, replacing the sensor is probably the way to go. Being in NW Ohio, you may find it difficult to remove the original sensor, due to rust.

Well, that's wrong. I've linked to the factory info -- why not try it that way?

What procedure are you using?

Thanks for posting the pictures. It does appear that the fit is the same, at least for the particular manufacturers. Since some companies apparently claim differences other than longevity, and some forum members have reported such differences, it begs the question as to why. The Cloyes belts apparently fit well in your application, so for them it seems either the 304 or 307 is acceptable.

The switch that you measured 4.7 volts on is the neutral position switch. It connects to the ECU, and allows it to determine whether you're in neutral or in gear. The voltage should drop to less than a volt when in gear. The other switch, with the 12 volts on it, is the back-up light switch, and apparently is the problem area.

See my post #8 in: http://www.ultimatesubaru.org/forum/showthread.php?t=117581 If they haven't been used before, the connector and probes might just look like a lump on the harness. They don't really come as loose as in the illustration, either. EDIT: The diagnosis connector is usually near the two green test connectors, which typically are hanging loose -- see if you can find them first. For alternative location of the connector, see post #40 of: http://forums.nasioc.com/forums/showthread.php?t=192583&highlight=abs+light&page=2

If the fuel pressure is okay (checked in the manner previously described), then the return line is fine. If the pressure is too high with the new FPR, then the line is likely at least partially obstructed. (Obviously, if air passed through the line to the tank, it couldn't be completely blocked.) Physical inspection might reveal damage. Unfortunately, it's probably not something you'd be inclined to do yourself. To see what's involved, look at the FSM section "Engine", file "valve clearance.pdf". Before the 2003 models, Subaru specified a 105,000 mile interval for checking/adjusting the valves on engines with mechanical lifters (hydraulic lash adjusters, or HLAs, don't require adjustment). You might find these interesting: http://www.mastertechmag.com/mtwp/2009/April/image/pdf/IS_200904_subarutrainvalve.pdf --or-- http://www.furi-dorifto.com/tech/Subaru_dox/Valve07MayEW.pdf

The 2003 Legacy has a pressure sensor, and more importantly, a combination MAF/IAT sensor. Proper functioning of any MAF is dependent on air flow past it being relatively smooth. The factory intake is designed to accomplish that. By replacing it with something else, at various engine speeds you'll likely encounter resonance/turbulance/pulsation that will upset the MAF's ability to properly read the air flow. Even if you manage to gain something with your modification, there's a good chance that there will be a performance and/or drivability issue of some sort in exchange.