OB99W

There should be two "ground probes", right next to each other, since both are used to clear the memory. (Diagnosis only requires one.) But just to be clear -- they're usually thinner gauge than others in the harness, and have a metal pin ("probe") on the free end. No, a black wire in B82 isn't for SRS diagnosis. And you're welcome. I want to make absolutely certain that you've found the correct connector. Please determine what color the wires going to pins #1 and #2 of B82 are (see the diagram for pin numbers), and get back to us. If they're correct, I'll post the procedure. Sorry if it seems that this is taking a long time to get to the point, but since you're rightly concerned about making an error, I certainly don't want to assume anything from about 3000 miles away (if "East Bay" is in CA). Some people have done damage to their cars by working with an incorrect connector and/or pin number -- let's not do that.

A tire calculator, like http://www.net-comber.com/tirecalc.html , can tell you what effect changing the tire size will have.

It seems like you didn't find the diagnosis connector -- it's not part of the SRS (Airbag) system, although it's wired to it. The 6-pin B82 connector is typically black in color. As I mentioned previously, you should find nothing connected to it (no mating connection). Look for a couple of loose green connectors (don't connect them) -- B82 should be nearby. Check near the bulkhead (firewall), it's possibly taped to the harness there. I also want to be very sure to give you accurate information. I believe that your "guess" about which wire to use is wrong. Could you verify the wire colors and locations (pin numbers) in the yellow connector? Please make sure to use lots of light, so that the colors are accurate. (It's easy to confuse yellow and white, for example, with insufficient light.)

It's much safer to use a diagnostic connector than one that's part of the SRS. See if you can find the one shown in http://www.ultimatesubaru.org/forum/showpost.php?p=668884 The connector and the "diagnosis terminals" ("ground probes") are usually under the dash to the right of the steering column, above the accelerator pedal. If they haven't been used previously, or were returned to the factory position, they may be taped to the harness, and appear to be a "lump" there. If you locate the connector (which should currrently have nothing plugged into it), let us know, and I or someone else can tell you how to use it to retrieve the SRS codes and/or clear the memory.

Sometimes it's just surface dirt that causes an electrical leakage path. You could try spraying down the coil pack with electrical cleaner (the kind that's just solvent, and doesn't leave anything behind once it evaporates). Dielectric grease can be used on the inside surfaces of the rubber boots to better seal them to the coil. If that doesn't cure it, new wires (get only OEM Subaru ones) and coil pack if needed are probably the right move.

Based on your experience Mr. Carb, it would seem Legacy2000 should consider investigating the possibility. I snipped a bit of Mr. Carb's fuel pump picture and circled the piece of protruding O-ring. Compare to the original picture to better visualize.

If the engine is healthy and warm, at an altitude not much above sea level the intake vacuum should be about 19 inches (or even a bit higher) at idle. Start the engine and allow it to reach warm operating temperature. Connect the vacuum gauge as you did before (throttle body), and take a reference measurement. Yes, disconnect the brake booster hose at the intake manifold (which if you do so with the engine running might cause it to stall), and close off the port at the intake (with tape or rubber plug, etc.). Start/restart the engine, and take another vacuum reading. If there's an improvement, the brake booster is leaking. If there's no change, that eliminates a possible leakage point. They do sometimes fail, so the data it's sending to the ECU should be checked. Can DaveT help? I don't know the Actron line well enough (especially without a model number) to know whether it would be appropriate. Typically a "code reader" doesn't do much more than reveal stored OBD-II DTC codes, while a "scan tool" can read real-time ("live", as it occurs) data. I personally use an Innova Equus scan tool. You should of course verify that any scan tool you're considering will retrieve the necessary data. Exactly. Real-time data is nearly instantaneous. The time to trigger and store DTCs, freeze-frame data and the like depends on how long it takes for the problem to occur. The drive cycle requirement for a particular code will determine whether it will be seen as history or pending. (See answer to #4 for some more.) Depending on the model, it should reveal stored codes, pending codes (ones where the problem is detected, but an insufficient number of cycles have been completed yet for the CEL to be lit), freeze-frame data (what sensors were indicating when a DTC is triggered), I/M monitor status, etc. For (a) and (, a decent scan tool should read them in real time. There is no direct indicator of vacuum leaks, although they can be inferred to a degree from other data. We could be looking for other sensor data, depending on where the diagnostic path leads. (If we knew for sure where we were going, we'd probably be there already. ) To quote myself from post #32 of this thread: "The P0133 O2 sensor code is usually a reliable indicator of a failing sensor. The only other thing that could likely trigger that code is an exhaust system leak before (upstream of) the sensor. If there isn't such a leak, the sensor is probably bad. While it may not be related to the current problem, if the front (upstream) sensor hasn't been replaced within some reasonable number of miles, it probably should." I still feel the same way. The ECU runs "open loop" (using default values and not looking at some sensors) until the engine and exhaust system have reached a certain minimum temperature. So, if the car has drivability problems that occur when the engine is cold, they shouldn't be related to the condition of the O2 sensor. However, O2 sensor data is obviously used once things warm enough, so a bad one could contribute to problems after warmup. You're welcome. I hope the above clarifies some things.

Holding the steering at full lock generates some pretty high pressure, so if something was ready to fail, that's when it often will. Clean things off and determine whether you have a seal leak at the rack, or a loose or bad line to it. You might be visiting the Pick-N-Pull again or getting a rebuilt rack, etc.

I can't seem to find a complete list of the Subaru DTCs online (not requiring lookup), but the following two links cover a lot: http://www.troublecodes.net/Subaru/ http://www.obd-codes.com/trouble_codes/subaru/subaru-obdii-obd2-codes.php

It wouldn't hurt to measure fuel pressure and delivery, although it seems a bit counterintuitive for a failing pump to show more symptoms at lower speeds (when the fuel demand is less) than at higher loads.

Just to be sure: http://endwrench.com/pdf/engine/FtTimingBeltReplaceF00.pdf

Well, at least it was easy to find, once you knew where to look.

Thanks for taking the voltage readings. The result is good news/bad news. The good is that the voltages are correct, eliminating questions over the neutral position switch causing problems. The "bad" is that we need to look elsewhere. (I'm apparently getting obsessed with this -- I just reread the entire thread, and took notes this time!) Back to basics -- declutching is resulting in a stall. It's been determined that this model has no clutch switch other than the one used for starter interlock, meaning that the ECU isn't informed of the clutch position. So, what could be happening? I see some possibilities: In typical street driving, people generally get off the throttle just preceding declutching. If the engine isn't running correctly, the transition towards idle conditions might not be smooth. Also, the clutch assembly has some mass, which when disengaged no longer aids the flywheel. The ECU could detect more misfires, etc., and might not be able to compensate for them as readily. As has been mentioned previously, the intake vacuum reading (16-17 inches, varying) is too low for your location (Colchester apparently has an elevation of about 500 feet above sea level, which would only drop vacuum from typical by about 1/2"). Your reading seems to be a couple inches below where it should be. There are several things that can cause low vacuum. Since leakage is a primary one, it's probably good to start there. A bad diaphragm in the brake booster is one potential cause. You can check this in various ways, but the most direct is to disconnect the hose at the intake manifold, and plug the port. If there was a leak in the booster or hose, the vacuum will come up and the idle will get smoother. Vacuum leaks can often be heard as a hissing sound. A piece of hose used as a stethoscope, with one end held to an ear and the other near potential leak sites (intake gasket, etc.), can help pinpoint them. Also, some intake cleaner sprayed at suspect areas with the engine idling will cause a change in idle if you hit a leak. I've assumed (perhaps without merit ) that things previously checked and/or replaced are not the cause of the problem. One thing that I didn't see mentioned in this thread is the intake air temperature sensor. Has it been checked or replaced? A bad IATS can report to the ECU that the air temp is higher than it is, affecting A/F ratio, especially on a cold engine. A scan tool (or computer with appropriate software) used in real time (live) and looking to see if sensor data makes sense (air temp not saying, for example, 110 degrees when the car's cold at 60 degrees). If possible, verify that all such data (ECTS, even though it was replaced) is "in the ball park". If you do run a scan and have questions about the results, let us know.

It shouldn't be too hard to check the front sensors and wiring to them. Just be sure to disconnect the battery before doing any probing that could set off the system.

You're welcome. Let's get the voltage test results, since they might be easier to interpret, before digging in any further.

Something seems to be amiss with the neutral switch, but the ohmmeter readings aren't making it obvious. Also, a reading in neutral as well as in gear would be helpful. Let's try a slightly different approach, using voltage rather than resistance. 1) Reconnect B135 to the ECU. 2) Reconnect the battery. 3) Set your meter to an appropriate voltage range to measure 12 volts or less (20 volts?). 4) Turn the ignition "ON", but don't start the engine. 5) Backprobe B135 pin #26, and record the voltage with respect to ground in both neutral and the gears -- two runs is sufficient. Let us know what you find, although I'm suspicious that things won't be normal.

The following is based on the '95 FSM. A long flash is on for 1.2 seconds. A short flash is 0.3 seconds. Continuous flashes of 0.6 seconds is the "normal" (nothing wrong) code. I'm interpreting what you're saying as "no long flashes, three short flashes". If so, that would be code 03. Code 03 indicates: Front sensor harness circuit is open --or-- Front sensor unit circuit is open Probably the first thing to check for is corroded connections or damaged wiring for both the left and right sensors. With the SRS, you can't clear codes until the problem is resolved.

If there's enough room, one trick that sometimes works with flare nut wrenches is to clamp vise grips around the wrench end to help keep it from spreading. And yes, it also helps to buy good ones.

This thread should give you the general idea: http://www.ultimatesubaru.org/forum/showthread.php?t=108631

Yes, measure from B135 pin #26 (wire harness) to ground. Any chassis ground will do. You're welcome. I think we're nearing resolution of your problem, but lets wait to see what you measure before I comment further.

Correct (and welcome to the forum, madcrazynomad). Besides heated front seats already mentioned, the AWP includes de-icers for the windshield wipers and the exterior mirrors are heated to de-fog/de-ice them.

The attachment shows the B135 connector. The page you've referenced shows the pinout from the perspective of the mating connector on the ECU, hence the seeming inversion. It's likely you were on the correct pin, but just to be certain, there should be a black wire with green stripe at that pin. From the readings you got, it seems that you back-probed pin #26 of B135 without disconnecting it from the ECU. (That could account for the changing readings, since it would be dependent on the electronics in the ECU.) Assuming I'm correct, please disconnect B135 from the ECU first, and then make the resistance checks again. That's interesting. Are you certain that there's no wire at pin #24 of B135? (The FSM indicates a green with black stripe there, coming from the VSS.)

The best place to check (best because the entire neutral position switch circuit can be tested, not because it's the easiest to get to) is at the ECU. Disconnect the car battery (to minimize a possible accidental short, etc.), and connect an ohmmeter between ECU connector B135 pin #26 and ground. The reading should change between high and low resistance as the shifter is moved each time between neutral and other gear positions. Move the shifter several times to verify that the switch contact is always reliable. See attachment for B135 pinout. Let us know the resulting readings. Don't assume the switch is bad if you don't get expected results, as there could be wiring or connector problems.

2000 Legacy, did you ever verify the correct operation of the neutral switch? While this may not be the source of your problems, it can absolutely cause the symptoms you are experiencing.

Check what's already been mentioned. If you don't find the cause that way, see: http://www.endwrench.com/current/02SpringInsiderInfo.pdf http://endwrench.com/pdf/fuel/MoreP0440TipsInfoW00.pdf http://www.endwrench.com/pdf/fuel/FtEvapEmissionTestW00.pdf