OB99W

Me too. One of my least favorites (at minimum, from the stench aspect) is 2-part polyester filler, commonly known to many by the brand name "Bondo". See MSDS http://www.plasticareinc.com/Merchant2/MSDS_Bondo/bondo%20body%20filler%20264.pdf Note that repeated overexposure to the stuff might mean that you can't have kids ("reproductive organ damage").

Now that my brain is more awake... I'm leaning towards there being insufficient air rather than excessive fuel. The engine's positive reaction to having the brake booster port on the intake left open (effectively a huge vacuum "leak") suggests that idle air is otherwise way too low. Even though in three places in this thread it was stated that the dealer replaced the IAC("motor"/V), and that it didn't help, a malfunctioning IACV would seem to be the most logical culprit. Furthermore, the $250 dealer price quoted seems insufficient for the valve -- that amount is more typical for aftermarket pricing on the internet. I might be wrong, but I suspect that the entire valve assembly did not get replaced. I was going to describe how to clean the IACV, but I decided to do an online search first, and found a thread that includes a post by porcupine73, explaining the procedure. See: http://www.subaruoutback.org/forums/66-problems-maintenance/12442-2000-legacy-w-rough-idle-occasionally-stalls-help.html Be sure to read both pages, especially the last one, which gives the resolution for that particular owner's problem.

Yes, the results were a little unexpected, which is why I specifically asked 2000 Legacy to verify that my understanding of his findings were correct. Any significant amount of unmetered ("false") air typically causes the engine to run too lean, and if not stall, at least bog down. The fact that it ran better with the additional air, and bogged when that air was cut off is, ummm, "interesting". It sort of implies that either idle air is inadequate, or idle fuel is too generous for some reason. I'm going to sleep and see if my mind will work on it overnight.

All the diagrams I've seen show that the fuse just acts as a switch, grounding a pin on the TCU when it's in place. What exactly happens after that is dependent on the TCU electronics.

Threebond 1215 is not an anaerobic sealant -- it is silicone based, oxime-cure type (RTV). See: http://www.threebond.co.jp/en/product/series/sealants/1200list.html RTVs can definitely deteriorate with extended storage. They last longer in the original sealed container, especially if the temperature isn't excessive. If you happen to buy it shortly after the dealer gets delivery of fresh stock, and treat it well, the shelf life will probably be considerably longer than the 12 month "spec". However, Josh said his was from 2005, and with RTV that old, proper curing is hit-or-miss. If anyone thinks that RTV doesn't have a limited shelf lfe, I suggest you read the FAQs at websites of companies that make the stuff. If you've never experienced RTV that didn't cure, consider yourself lucky. I've gotten "bitten" more than once by RTV (and epoxy) that didn't set up. In one case, the RTV was well within the allowed shelf life, so either it was exposed to high temps for too long before I got it, or it was just a badly formulated batch. Since cleaning up uncured material is such a pain, I now make it a point to put a small amount on a piece of paper and make sure it cures correctly before using the rest on a job.

I believe so, but I'll admit to not having ever verified that with a scope. If anyone knows otherwise, please speak up. I definitely have to modify something I said. While the earlier 4EATs did usually have 90/10 front/rear split, vehicles with VTD/VDC (might be starting in 2001) have more like 45/55. Using the FWD fuse on so-equipped cars would prevent the dynamic control from functioning.

As has already been stated, there's really no advantage and it may possibly lead to failure of the duty solenoid (the duty cycle does indeed go to 100% when the FWD fuse is inserted). The typical split of power is 90/10 front/rear (for the auto trans, which we're discussing) when the AWD system isn't detecting slippage. Hmmm, WA2NTK -- 73 OM, CUL

I sometimes have a tendency to incorrectly assume that forum members have certain knowledge. When checking engine intake vacuum, the best port is the one closest to the center of the intake manifold, and a "T" should always be used so that running conditions aren't changed. That is, except in the case of things like the brake booster which aren't related to engine operation, vacuum hoses shouldn't be left dangling. Before I comment further, just to make sure I understand what was said in points 7 through 9: The engine ran better (smoother, higher vacuum) with the brake booster hose disconnected from the intake and that port left open, then temporarily faltered (vacuum initially dropped) when the port at the intake was plugged. Is that correct?

Since you said you can hear the injector pulsing, it's obviously getting turned on and off electrically. Therefore, I'm doubtful that a oscilloscope will reveal anything useful. A voltmeter or noid light would probably tell you enough. As to swapping injectors, pick one that's convenient. However, to answer your question, each injector is pulsed at a rate equal to half the RPM of the engine (since it's a 4-cycle) -- to get pulses per second, divide RPM by 120 (60 seconds per minute, 2 revolutions per pulse). Choose an appropriate scope sweep speed based on that, or just switch the horizontal timebase until you see a reasonable number of cycles. At idle, it will be pretty slow (example: 600 RPM divided by 120 yields only 5 Hz pulse rate), and might require a scope with either long persistence or storage to easily see. Could you elaborate on "vacuum is good"? What was the reading at idle and higher speeds, up to about 2,000 rpm? How steady is the reading? What's your elevation? As mentioned previously in this thread, valve contribution to misfire can be due to either not sealing correctly or not opening enough. (That assumes there's not a timing issue, which is less likely since you aren't seeing a problem with cylinder #2.) If there's a leaky valve, a drop in vacuum will occur each time the problematic cylinder is operated -- the amount of drop is dependent on the degree of leakage. Some vacuum gauges have a very small orifice, and aren't a good indicator because they tend to average/smooth the measurement. Especially if the vacuum reading is inconclusive, a cylinder compression and/or leakdown test could be helpful. EDIT: See for injector info: http://hp.auto.ru/Subaru%20Legacy/1997/Service%20Manual/ENGINE%20SECTION/FUEL%20INJECTION%20SYSTEM/

Josh, you're probably not going to like this, but the "official" shelf life for most RTV is 6-12 months. That doesn't mean it can't remain usable for a longer period -- I suspect that short period takes into account other time spent in stock, on dealer's shelves, storage at temps outside the recommended range, etc. -- but 5 years is asking an awful lot of it. Oxime-cure RTV (Threebond 1215, Permatex Ultra Grey, and other "sensor safe", non-corrosive ones) are dependent on moisture in the air to cure (as are the acetoxy-cure RTVs that give off a vinegar-like [acetic acid] smell). RTVs designed for things like engine assembly have a cure that's slow enough to allow parts to be mated before surface "skinning" occurs. However, you should typically expect surface "drying" within an hour or so. Up to a point, additional heat and humidity will accelerate things, but if the normal rubbery cure isn't reached within a couple of hours, the material probably has just gone bad. If that's the case, not only might the initial seal be questionable, but the oil resistance will likely be compromised as well.

John, many EndWrench articles seem to no longer be accessible. One for the DOHC 2.5L that still can: http://endwrench.com/pdf/engine/FtTimingBeltReplaceF00.pdf

Hot weather in combination with high enough engine load can lead to pinging even if the fuel octane rating would otherwise be sufficient. It would be nice if knock sensor data would just cause the timing to be retarded, but sometimes that's not the answer. Especially if the cooling system isn't quite up to par (but not bad enough to cause overheating), that can be the tipping point. Since there's apparently a "Danville" in several states, I couldn't check to see whether particularly high temps are a recent situation. Is the temp gauge reading where it usually does? Have you checked for any obstruction of airflow through the radiator? (A bunch of dead insects can make a difference.)

Yes, and code 14 which the OP reported is specific to improperly locked connectors. See my post #19 in this thread.

If you don't know the maintenance history, it would probably be a good idea to replace general "tune up" items. That would include spark plugs (get originals), ignition wires (suggest Subaru only), fuel filter, air filter and PCV valve (clean the tubing as well). If you don't know that the timing belt was changed on schedule, it and probably the idlers, perhaps tensioner, and water pump should be considered. The engine is an interference type, and if the belt breaks there will be internal engine damage.

Welcome to the forum. Please define "a while". Does that mean "when the engine has warmed"? If you watch the temperature gauge as the engine warms, what's the position when the problem goes away? About how many miles and how long does that typically take?

You should probably concentrate on why the engine is running so poorly, and not specifically on the P0420 code. If emissions are bad enough before the catalytic converter, things can't be cleaned up sufficiently in the cat, and you'll get a P0420. A "lazy" or failing front oxygen sensor can cause all kinds of grief. However, if the system isn't remaining closed loop once the engine is warm (with the possible exception of WOT), something else is wrong. One possibility is a bad ECTS (engine coolant temperature sensor). If your scanner can monitor that, see if what it says makes sense.

Think about the conditions under which the problem is most likely to show up, then take a look at these: http://www.furi-dorifto.com/tech/Subaru_dox/ChaseEWAug06.pdf http://www.aa1car.com/library/vibrations.htm http://www.aa1car.com/library/tires2.htm

In addition, with a turbo, MAF trouble can set a trans code -- with non-turbo, the MAP can.

Yes, fixing the problem that's causing the warning light to come on will allow the light to go out. That's a good way to tell that the problem has been repaired. However, most of the codes are stored in memory. One reason is that a problem may be intermittent. The memory ensures that even if the light is out (because the problem isn't occurring currently), a code history can be retrieved so that you can diagnose the intermittent problem. Any codes can be cleared using the correct procedure, but only if the problem isn't happening at the time. Currently ocurring SRS codes can't be cleared to "see if they come back". Pin #2 (white wire) of the B82 diagnosis connector is used for clearing memory. It needs to be grounded, as per the manual pages that I previously gave a link to. But first the problem should be repaired so that the warning light goes out.

John, for complete info on '98 doors, download all the PDFs here: http://hp.auto.ru/Subaru%20Legacy/1998/Service%20Manual/BODY%20SECTION/DOORS%20AND%20WINDOWS/ For just door alignment, page 12 (PDF page 6) of this: http://hp.auto.ru/Subaru%20Legacy/1998/Service%20Manual/BODY%20SECTION/DOORS%20AND%20WINDOWS/MSA5TCD98L21235.pdf

You're quite welcome. I hope it turns out to be just connector related.

Code 14 indicates that at least one of the connectors to the airbags, front left and right sensors, or the control module are not properly connected. There are switches built into those connectors, and if they aren't properly latched the control module can sense it. Sometimes unlatching and reconnecting the connectors will resolve the problem. Before trying that, be sure to disconnect the battery, and wait for at least a minute so that the back-up capacitors can discharge. Make sure that the connectors are double-locked. Do you know where all the connectors are located? EDIT: The link below was unavailable for quite a while, but I just verified that it's back online. Download the PDF files (for '95, but should apply), and you'll have complete troubleshooting info for the SRS, including what the codes mean and how to check for the causes. http://hp.auto.ru/Subaru%20Legacy/1995/Service%20Manual/TROUBLESHOOTING%20SECTION/SUPPLEMENTAL%20RESTRAINT%20SYSTEM/

Terminal (in more ways than one, if there's no 'chute deployment!) velocity for a person in free-fall is about 130 MPH. I wonder what it is for a Legacy.

We're good -- let's start by getting the trouble code(s). Turn the ignition switch "ON" (engine not running). Plug a ground probe into B82 pin #1 (blue wire). The lamp should start flashing. The length and sequence of the flashing determines a code. Trouble codes will be in "longs" and "shorts", with the longs the first digit, and the shorts the second. Single digit codes have no longs. See my post #3 in http://www.ultimatesubaru.org/forum/showthread.php?t=114914 . Let the flashes cycle until they repeat, to make sure you get all possible codes. Once you retrieve the code(s), post what you get and we can tell you what they indicate.