OB99W

Carl, you've got a Private Message that might help.

It seems that we'll have to respectfully agree to disagree with each other for now. I'd already downloaded everything I could find online on the topic, including the EndWrench items you thoughtfully provided links for. Unfortunately, not one gives descriptions of the 7 shift maps that the '99 Phase 2 trans uses. Even the "4EAatPh2Win04.pdf" article is incomplete, since it neglects to mention the torque converter lockup inhibition that occurs in addition to the 4th gear lockout when cold. Also, the same article lists "Faulty shifting when cold" as a consequence of ATF temperature sensor trouble (see "Self Diagnosis" table). It isn't specific as to what fault(s), unfortunately. I have read elsewhere (although I can't find anything quotable at the moment) that the delayed/raised shifts when cold are implemented to warm the engine and emissions-control systems (including the catalytic converter) faster in order to cut emissions. If indeed that is the case, I have no problem with it; I'd just like to be sure my trans isn't thinking things are colder than they really are.

You're welcome, and sorry if the first part was too basic; it was for general consumption, not directed towards anyone in particular. I kind of cringe when terminology is misused, and sometimes tend to over-react. Exam? Next week, but relax, it's open-book.

First things first . Are you saying the AT temp light is coming on? If so, when and how? Trans hot, or cold? Lighting steadily, or flashing? If flashing, is at a sequence of longs and shorts (a trouble code)? If a code, what is it? More info, please.

One approach: http://www.inct.net/~autotips/musty.htm As already suggested, be sure to verify that the drain isn't clogged. If using a spray, when applying it via the exterior air intake (near the windshield wipers), be sure the a/c is not in the "recirculation" position, or it won't pull the product through the system.

First, I agree with others who have said that the fuel issue is probably bogus; your engine isn't likely to "complain" about the gas in only one out of four cylinders . Having said that, I'm wondering about what standard the dealer is using to determine the volatility. The usual is RVP (Reid Vapor Pressure), but the number you quote (3.5) doesn't make sense on that scale. RVP typically averages about 8 psi, and winter grade of gas is often over 12. Gas of 3.5 RVP probably wouldn't work too well in any of the cylinders, possibly even in summer. You might want to ask the dealer what standard and method the "3.5" is based on. By the way, premium gas often is less volatile than lower grades, so using it to minimize misfire due to low volatility is going the wrong direction.

I don't think the sensor could be open or shorted, since the TCU is supposed to flash the AT Oil Temp dash light if things are that out of bound. It hasn't ever done that, and the light comes on when starting, so the bulb isn't out. Thanks for the suggestion concerning resetting the ECU/TCU, but I've had the battery disconnected for long enough to accomplish that (for other reasons), with no change in trans operation. So, I'm back to asking if anyone has specs for the cold/hot resistance of the temp sensor in a '99 4EAT. Anyone, please?

It seems that later '99 4EATs got new shift maps, including one called "Cold ATF". See: http://endwrench.com/images/pdfs/994EATInfo.pdf It would seem that the delayed upshift when cold was planned, perhaps to more rapidly warm the trans and engine, or avoid lugging a cold engine by keeping the revs up. However, mine seems to stay in that mode too long, and revert to it too readily. The behavior it exhibits has been there since new, continuing through filter and fluid change, and if my TCU is "crazy", it was "born" that way . Thanks for the suggestion Jacek, but the trans shifts are a lot more "correct" when things are fully warm, so throttle position sensing wouldn't seem to be the problem. I'm much more inclined to think that the trans temp sensor is "crazy", or that Subaru went overboard on the '99 4EAT "Cold ATF" map. But I'm open to other suggestions.

A bit of investigating at the Endwrench site has led me to the information that the 1999 phase 2 4EATs have 7 shift maps, one of which is specifically used when the ATF is cold. As Nipper previously suggested, it does appear that the temperature sensor is attached to the lower valve body (next to duty solenoid and that would seem to be a location that wouldn't rapidly change temp. I'm beginning to wonder if the temp sensor might be out of spec, and telling the TCU that the trans is colder than it really is. That could explain both the overly-delayed upshift and the particularly poor gas mileage I get when it's cold. Anyone know what the correct resistance readings are for the trans temp sensor, cold and hot? Any help would be appreciated.

Thanks Nipper; that's what I would have expected, so I wonder what explains the behavior of my trans. After a reasonably long trip, even a short stay in cold air (certainly not long enough for the valve body or vicinity to cool down much) is enough for the trans to again take a long time for the first shift, just as if it had been left out in the cold overnight. It acts like the sensor were hanging free on the outside of the trans, not making good thermal contact with something massive. Unless someone can tell me precisely where the sensing is done, I'll have to look into it myself after the million other things I need to do are completed. EDIT: Of course, this problem may have nothing to do with the temp sensing, since it concerns the shift out of 1st (not inhibition of torque converter lockup or of 4th, both of which I'm aware are normal when the trans is cold). Unless Subaru had some mis-programmed TCUs...

Well, I guess "they gave" me "a junk tranny" too, because mine has behaved very nearly as badly as the one that Steves72 has, and it came right from the factory attached to my then-new 99 OB ! I've never investigated where the trans temp sensing is done, but on mine it acts as if it senses air temp and not the fluid. On a cold day, even after the engine and trans are quite warm, if I park it for just a few minutes, when I start it again it's right back to the reluctance to upshift, etc. Mine also doesn't seem to know when the engine isn't loaded, as upshifts after climbing even a moderate hill and then driving on flat terrain are slow as well. The fluid is at the right level, etc., and I doubt replacing it would do anything good, since the problem has been there since I took delivery of the car. Of course, this is "normal" behavior, according to the dealers. The trans is hardly the only thing on my OB that performs poorly, but talking to the dealers and SoA is as productive as talking to the proverbial brick wall. I gave up long ago. I've repaired the things that were critical myself, and I'm ignoring the rest as best I can. It's one thing to buy a car used, not know its history, and then expect the dealer/manufacturer to stand behind the product; it's quite another to pay $26,000 for something new and get treated like dirt. I envy those who own a Subaru that has given long and relatively trouble-free service, and those who have found good dealers or had SoA be cooperative when there was a problem.

Ahhh, "Star Wars" technology; so it's actually like "DeathStar" ?!

As you probably know, "Knocks and pings" is a farly general description (are we talking "detonation", or preignition"?), and can have several causes. If the problem came on suddenly at 13k miles, then you need to investigate what changed mechanically/electronically at that mileage. An excessively lean air/fuel mixture can cause knock, and could come on gradually if one or more of the engine management systems has become contaminated. The knock could even relate to something such as carbon in the combustion chambers. Carbon buildup can cause "knocks" in two ways. The first is that it can increase the effective compression ratio, and retarding of the timing through knock-sensor data can help. However, a second and probably more likely possibility is that carbon will cause combustion chamber "hot spots", where the glowing carbon prematurely ignites the mixture. Retarding ignition timing doesn't tend to make a difference if that's happening, so even a properly-functioning knock sensor won't help. If the dealer has verified that engine timing, mixture, running temperature, EGR, etc. are okay, and you're using the correct grade of fuel, then maybe something as simple as a can of combustion chamber cleaner might help. Admittedly, at 13,000 miles carbon isn't usually a problem, but certain conditions can cause it even at low mileage. A bit more info as to when the knocking occurs (engine hot/cold, accelerating/cruising, level/uphill, etc.) might permit better advice.

I seem to remember reading about a fairly recent "campaign" by Subaru for MAF problems on the 99s. This may not have been a "recall", so letters might not have gone out to owners. I realize that you've resolved the problem for now, but perhaps you could call a dealer or Subaru of America and ask them about this. If you do get confirmation of a campaign, please let us know here. Thanks.

Oh man, does that open a can of worms ! First, we have to differentiate between "current" and "voltage". Then, we can discuss how the injectors are switched on/off. To begin, here's a bit of "Basic Electricity 101": 1) "Voltage" is the difference in electrical potential between two points. Even if not connected to anything else, a charged car battery has a little over 12 volts of potential difference between its two terminals. However, no current flows if nothing is connected between the battery terminals. 2) "Current" requires the flow of electrons in a circuit. If you connect a light bulb from the positive to the negative terminal of the battery, there is then a complete circuit. The bulb lights because the current flowing through it heats the filament, which brings us to "resistance". 3) "Resistance" is the characteristic that determines the relationship between voltage and current. A fellow named Georg Ohm figured that out, and we measure resistance in "ohms". See: http://en.wikipedia.org/wiki/Ohm%27s_Law Now, to get back to the injector operation: All the injectors have one connection to +12V at all times that the ignition is on. They are activated, when appropriate, by making a ground connection to the second lead, completing the circuit and causing current to flow. That ground connection is made by an electronic switch in the ECU. While more-sophisticated test equipment can be used to verify that the proper voltage and switching are available at the injector, a "noid" light can simplify things. If it doesn't light, and +12V has been confirmed at the injector terminal, that leaves only the ECU itself (or a bad connection to it) as the cause. I hope that helps, and doesn't muddy things further .

Once my '99 OB was out of warranty, the dealers never saw me again for service. My experience with 2 dealers during the warranty period was that they either were incompetent or just didn't care. My experience with SoA is that they exercise little or no control over their dealers. "Deny" seems to be SoA's approach to all problems. When (and I do mean "when", not "if") the OB has a major failure, be it trans/HGs/etc, I may not even fix it. Rather than put any significant time and money in, I'll likely dump it as-is and buy another brand. While I'm sure that some Honda, Toyota, etc. owners could tell their own horror stories, in my circle of friends and family, vehicles from those manufacturers have proven to be much more reliable than my Subaru, given equivalent care. It seems that SoA's corporate philosophy is that they have enough customers so that losing a certain percentage isn't important. Maybe when the number they've alienated reaches a critical mass, the sales figures will begin to reflect it.

"Time delay" I could comprehend, but I want to know how they so accurately determine road mileage. Remote reception of odometer data?

Right before it jumps a few teeth or breaks. Now, where did I put the keys to the TARDIS? http://en.wikipedia.org/wiki/TARDIS Otherwise, sooner would be better than later. See: http://endwrench.com/images/pdfs/2.2SingleOverWin01.pdf http://endwrench.com/images/pdfs/TBeltEWWin05.pdf Also: http://www.gates.com/brochure.cfm?brochure=2256&location_id=3487 Gates has downloadable literature, and you can check a particular vehicle online.

There's another method besides the "spray" approach that often works. A piece of automotive vacuum hose (not the one from your Hoover ) or similar small-diameter tubing of appropriate length is used; one open end is held to an ear, the other end moved around as close as possible to the area of the suspected leak. A "rushing" or "whistling" sound is heard when approaching the leak, louder the nearer the tube is to the leak. Of course, if you can borrow a stethoscope (minus any solid diaphragm and perhaps even a belled end), that would probably be easier to use.

If there's any question as to fluid age/condition, it's probably a good idea to completely replace it; that's what a "flush" does. In my opinion, don't have any additional chemical cleaning done, as that often leads to problems.

Thanks for the link. Note the part near the end of page 13, "The timing chain on the EZ-3.0 is designed to last the life of the engine. Proper engine oil maintenance is necessary to ensure its longevity." (I added the emphasis.) While good oil is obviously important for all of the engine components, timing chains really don't like incorrect/contaminated/fuel-diluted oil.

Timing chains, like other engine parts, do wear. Since they have to be lubricated (unlike a belt), the rate of wear partly relates to the condition of the engine oil. If the oil isn't of sufficient quality or changed frequently enough, the chain (just like other engine parts) will wear faster than it should. In addition, the sprockets the chain runs on (sometimes called "gears") also wear. The wear has the effect of "stretching" the chain, with ever-increasing timing errors being the result. The fibers in timing belts don't tend to stretch much, keeping timing more accurate during the life of the belt. Of course there is some wear of the teeth and eventual cracking of the belt's rubber compound, but hopefully the belt would be changed before that causes a problem. Timing chains tend to be noisier than belts. Some manufacturers use plastic-toothed sprockets to reduce noise. Unfortunately, they don't seem to be as reliable as metal ones. Also, due to the need for lubrication and other factors, a chain can't be enclosed in simple unsealed/ungasketed covers. In addition, the tight turns from sprocket-to-sprocket necessary on some engines (especially certain DOHC) are more easily accommodated by a belt than a chain. And as "1 Lucky Texan" suggested, the mass of a chain versus a belt would almost always be a design consideration.

If possible, driving another Subaru of similar vintage to compare shift characteristics could be useful; it could either resolve your concerns or confirm your suspicions. Mercon is apparently very similar to Dexron III, so the dual-usage fluid is what's commonly available. As Nipper said, the 1-2 shift does typically seem a bit harsher on 4EATs than one might expect. On the other hand, if anyone has ever added some Ford F-type fluid (which isn't friction-modified like Dexron/Mercon and is incompatible), that in itself can result in hard shifts. EDIT: Problems with duty solenoid "A" or the related resistor can cause hard shifts due to higher-than-intended line pressure during shifting, although it likely wouldn't affect just 1-2. As to the fluid level, it depends what you mean by "cold". Is that with the engine running? Drainback from the torque converter after the engine is off for some time might explain the higher level; that typically shouldn't happen to any great degree within a day or so. By the way, be sure to remove and wipe the dipstick carefully, allow several seconds for fluid to run back before reinserting the dipstick, and then remove it again and check the level. Otherwise, I find that fluid sticking to the dipstick tube transfers to the stick and makes it difficult to get an accurate reading. While you're at it, you might smell the fluid on the dipstick and compare that to fresh fluid. The odor is often as good or better an indicator of bad fluid as the color. In the long run, a fluid flush (complete change with fresh fluid, not using chemical cleaners) should probably be considered. Be aware that if the trans has some "issues" and an additive has already been used, replacing the fluid without an additive can sometimes reveal any problems, making it seem that the fluid change made the trans worse .

Here's how to check the level: http://endwrench.com/pdf/drivetrain/02Winter4EATDiagServ.pdf Probably uses Dexron-II or -III, but here's how to get a manual and be sure: http://www.subaru.com/common/faq/index.jsp#6 I hope that helps.

If nothing else, a vacuum gauge won't leave your hand smelling like burnt dinosaur . And yes, I have in the past stuck my hand in car exhaust streams. Although it would be a lot easier to see on a gauge, does the "power by palpation" test lead you to believe that the cylinders are firing evenly, or do you notice a periodic drop in exhaust pressure? If it's uneven, I'd consider a single-cylinder "misfire" scenario, which could have several different causes. If it's relatively constant (considering that with a 4-banger, the power cycle doesn't come around very often at idle ), I'd start with the simple. Maybe all you've got is a vacuum leak; check for disconnected or split vacuum hoses, or a bad power brake booster. I still think the vacuum gauge is an under-appreciated and under-utilized tool. EDIT: A couple of more links to interpreting gauge readings: http://www.ifsja.org/tech/motors/vacuum.shtml http://www.v8sho.com/SHO/IntakeManifoldVacuumTest.htm