OB99W

The original problem is that water and gasoline aren't very miscible (mutually soluble). Isopropanol and ethanol are miscible with both water and gas, and therefore help prevent water from accumulating in the fuel system separately from the gas. In theory at least, the ethanol that's used in E10 is "dry" enough (isn't very hydrated) so that it can certainly pick up additional moisture that might be present in the fuel tank, etc., and additional alcohol shouldn't be needed. Also, Subaru doesn't recommend any more than 10% alcohol content in the fuel.

As has been said, the trans is designed to not upshift to top gear until it has warmed sufficiently. Not that there couldn't be a problem developing, but what you just said may be the key to what's happening; if the car is getting colder than usual, it will take longer to reach the temp at which fourth gear is allowed. In addition, using a remote starter to "warm up" the car doesn't do much for things beyond the engine -- the trans may benefit slightly, but the diffs, etc. are going to still be cold and have thickened lube. Also, the trans fluid itself expands and contracts with temperature changes, and when cold it's both somewhat more viscous and lower in volume, which can affect operation. So, if the cold acceleration is sluggish, that could be normal. Whether the torque converter is locked or not could be the difference you feel. However, if you see a large engine RPM flare when trying to accelerate but get no significant gain in road speed, that would indicate slippage. EDIT -- See this thread: http://www.ultimatesubaru.org/forum/showthread.php?t=70806

You might try a trip through a car wash, with a friend or family member in each seat near a door as observer. There are adjustments for glass positioning should you find any obvious entry points, although if there is definite wear the weatherstripping may have to be replaced.

Clearances do open somewhat as an engine accumulates miles, but it's probably normal; 10W-40 at 24 deg F could take a couple of seconds for the pressure to be sufficient for the sender to turn the light off. Of course, if the light ever comes on other than briefly at cold start, that's another story.

Not surprising; there are few, if any, free ($300?!) lunches. Subaru engineers may not get everything right, but missing on HP potential by 25% would be quite a stretch. Don't forget, by definition, "venom" = "poison".

I'm a true disbeliever (and don't have one of the cars it's supposed to work on), but for curiosity I took a look at the thing. The claim is 25% increase in HP over the 1,000-5,000 RPM range. Interesting... ...but not too likely. Energy produced is directly related to fuel burned. Since all the module connects to besides power and ground is TPS and MAP or MAF (according to the info at the website), the most the module could do is increase the fuel/air ratio. The module's application is for a '90-'91 manual trans Legacy. Unless those models normally run unusually lean in the 1k-5k RPM range, increasing F/A isn't going to help much. Going excessively rich won't improve HP, but it will kill your gas mileage among other things. Burning more fuel and air in the proper ratio can certainly provide more power, but that requires being able to get the mixture in and out of the engine faster. That could require freer flowing intake and exhaust, larger cylinder volume, or cramming more charge into the cylinders under pressure (that's what super-/turbo-chargers do). The control module can't do any of that. Anyone have a different "read" on this?

Whether it's a problem depends on just how long "a few seconds" is. So that brings up another question, "How many seconds?.

Might have some ideas, but a bit more info would help. Model/engine? Mileage? OEM or aftermarket filter? Oil viscosity? Last time oil changed and level checked? Outdoor temperature? Parking on level ground or hill?

Congratulations; glad we could help.

Being a person who knows something about such things (RF transmission and reception, not skulls as amplifiers), I could propose a couple of theories. However, I've grown wiser in my old age.

I'll have to agree with the majority; my own '99 doesn't take over 5 minutes for the heater to produce useful heat, even if the ambient temperature is quite low. Even at 60 MPH, 20 miles is 20 minutes, so yes, that does seem excessive. Could be a stuck-open (or missing?) thermostat, or a heater core that's partially obstructed, or maybe a cooling system with an air bubble that needs "burping". Are there any other symptoms? How does the engine perform during those 20 minutes, etc.?

Thanks for posting the pics. In my first reply in this thread, the post I linked to has further links to three End Wrench articles, and some from Motor Magazine. They're all useful, but perhaps the one most germane might be this one: http://endwrench.com/images/pdfs/TBeltEWWin05.pdf By the way, if the tensioner wasn't compressed correctly, it may not be working as it's supposed to now.

I don't doubt that you're sure the belt was installed correctly. However, I can tell you from experience that when an engine that was running well before a t-belt change isn't any longer afterwards, almost always it turns out to be due to the belt being off by a notch (EDIT: "off by a tooth") somewhere. Maybe it was initially correct and jumped after you started it, perhaps due to a problem with an idler or the tensioner. Having said that, sometimes an engine with HLAs will be a little noisy and rough for a short while if they've bled down due to moving the cam sprockets. You could also check for dislodged ignition secondary wires. It would be useful to know what the code is. There is no distributor on the engine, and there are no knobs to adjust; timing is set by the ECU (computer) in conjunction with sensor data. You can't be "off by a whole rotation on the crank gear"; if the crank mark is correctly aligned, its timing is correct.

Out of all the worlds in the universe, I'm sure that there's one where everything is mirror-image to things on Earth.

Make sure you are using the correct marks; see the links in this post: http://www.ultimatesubaru.org/forum/showpost.php?p=557934 There are both notches and arrows. In each case the notch is the correct mark to use.

So, you found humor in that, huh? Okay! (It's just that I've learned to never assume anything. )

This may be obvious, but if you decide to try the above or anything similar, fully deflate the tire and remove the valve core beforehand.

I've used bolt/screw extractors with success, but I'm careful; If gentle turning by hand doesn't budge things, I stop. As suggested above, a broken-off hardened piece of steel will be much more difficult to remove than the original bolt. Careful drilling and chip removal usually does the trick.

One method of checking for that is to wait until dark, start the engine, open the hood and look for sparks. You could mist the secondary wires (and the coil) with water using a spray bottle; good insulation will tolerate that, leaky will usually become obvious when wet.

Left-hand drill bits and reversible drill; see: http://www.madelectrical.com/workshop/broken-bolts.shtml There's lots more info on the Web; do a Google search on "broken bolt removal" or "broken bolt extraction".

That's what I thought might have occurred, and why I asked how you did the measurement. The quick story is Mark is right. The longer story has to do with how a voltmeter works. Let's assume that the radio (or some other load) is connected, but the fuse is blown. If you ground the meter's negative lead, and place the positive lead on the fuse terminal that connects to the battery, you should read 12V on the meter. If you move the positive lead to the other side of the fuse (the one that goes to the radio), if the fuse is blown you should see zero volts. However, if you put the meter leads across the blown fuse, you'll see 12V because one lead of the meter is then connected to the battery, and the other lead gets a "ground" connection made through the radio. You might think that the "hot" wire of the radio wouldn't make a very good ground, but it doesn't have to. That's because any halfway-decent voltmeter has a pretty high input resistance, and therefore isn't going to cause much of a voltage drop across the radio; the voltage measured will appear to be quite close to 12V. Okay, that probably wasn't as clear as it could have been, but it's getting late . Anyone who wants a better understanding could do a Google search on "Ohm's Law" and other electrical concepts. Anyway, I'm glad you found the problem.

By now I'm probably boring everyone else, so this will be my last attempt. We know that the early Phase-I/DOHC 2.5L engines developed internal leaks when the HGs went. We also know that the later Phase-II/SOHC ones typically leak externally when their HGs go. I'm trying to get a sense of the situation for the 2.5 engines which are still DOHC, but use the later block (rear thrust and whatever other changes were made). Do such engines only exhibit the internal leak, the external one, or are they possibly more prone to both (or neither )? This is hardly a critical issue to me, so if it's an unknown, we can certainly drop the topic. That was my attempt at avoiding the term "hybrid".

Thanks for the additional info. I don't have any "specifics" either, and some of this discussion seems to have become one of semantics, so I think I'll stop trying to get any further details. Of course, if anyone knows for sure ...

Thanks, Emily (ccrinc) and WAWalker (the SOA info is especially interesting). This seems to support the contention that there are "later" blocks (so-called "8-bolt", really 6-bolt 2-stud, with rear thrust) paired with "earlier" heads (DOHC), as my own MY99 OB came set up that way from the factory. The pistons would have to be chosen to not interfere, provide correct compression ratio, etc. for any particular configuration. Apparently 97-99 was a transitional period in the EJ25, Phase-I versus Phase-II. Here's what "hybrid" may mean in the context we're discussing, from: http://www.answers.com/hybrid&r=67 **hy·brid - n. Something of mixed origin or composition. --also-- **Refers to a myriad of products that combine two or more different technologies. Whatever "hybrid" specifically means to SOA, thanks to everyone who helped clarify that there are indeed EJ25 versions that combine DOHC heads with the block later used on the SOHC engine. So now I'm back to my previous question -- are there any statistics (or opinions ) as to how the HGs hold up on the "mixed heritage" engines?

Even assuming that this is an ignition-related problem (and the simultaneous cylinder 1-2 misfire would seem to suggest that), the coil pack of course isn't the only thing that can cause it. If you haven't seen these before, they might help with troubleshooting:http://endwrench.com/images/pdfs/IgnitionCoil.pdf http://endwrench.com/images/pdfs/DirectIgnition.pdf http://endwrench.com/images/pdfs/Correction.pdf http://endwrench.com/images/pdfs/IgnitionCoilSum04.pdf The info in the PDFs above is a bit general; if you do resistance checks on the coil, what you get may not meet the spec in the articles. Since apparently your coil for cylinders 3-4 is okay, measure that first and use it as your "standard". If the measurements for 1-2 differ, then suspect a problem. Although an ohmmeter often won't show a problem (it often happens only under high-voltage conditions), it may be worth the time to do a leakage-to-ground test on the coil secondary windings in addition to the primaries. There have been variations of coils packs over the years; for example, some have the igniter separate, while in others it's integrated.