OB99W

That's right, you look for rust, or perhaps just a bubbling of the paint finish. My own OB, when it was just a few years old, suddenly began dripping ATF. Fearing the worst (why would a low-mileage 4EAT be leaking ? ), I was relieved to find that the spin-on filter had some seepage due to the apparent action of road salt. A new OEM one isn't cheap, but I wouldn't gamble the difference in price over something that "fits" versus possible transmission repair or replacement.

The math doesn't lie. Drawing 40mA over a two week period shouldn't drain a properly sized and charged battery to the point of not being able to start the engine. As I said, either the battery isn't up to the task (for whatever reason), or the current draw is more than 40mA. Neither I nor Cougar said the battery was the initial problem, but it could become the victim. The fact that you're killing batteries in six months on average says that either they're not up to the job, or something is stressing the heck out of them. For now, I'll assume the batteries themselves aren't that bad. So, as Cougar said, there could be a problem with the alternator. Checking for too much ripple (AC) is a start. I'd also verify that the charging voltage doesn't get too high -- you can "cook" a battery. Also, if a car battery is run down too far (by leaving on accessories while parked, for example) before recharging (especially if left in that condition for some time), it will "sulfate", significantly lowering the charge capacity. A battery that discharges over time doesn't necessarily have to do that via an external load. Batteries normally slowly self-discharge, but that can be sped up dramatically by conductive "sludge" at the bottom of the cells. The sludge is material that gets knocked out of the battery plates; that's doubly bad -- besides causing an internal discharge path, it lowers the amount of active material. That sort of thing can happen when the battery is subject to excessive vibration. In other words, not properly clamping down a battery can cause the problem. High temperature is also a battery killer, even if not caused by the charging voltage being beyond what it should be. If for some reason the underhood area that the battery is mounted gets too warm, that can be enough for problems. By the way, I typically get about a dozen years from batteries in my own vehicles. They're not "no/low maintenance" or "high performance" types, since I've found that both of those usually have shorter lives. If you keep the electrolyte at the correct level (which is one reason I like individual cell caps), don't over/undercharge, don't overly discharge, keep properly mounted, don't overheat (don't live in the deep South ), etc., you should be able to get at least 6-7 years on average.

Unfortunately, there's nothing good that causes lack of compression. If your perception of increased cranking speed is accurate (likely due to lessened resistance from the engine), then that's the thing to investigate. Exactly! With good compression, checking further for spark, etc., makes sense. However, if there's a compression problem, the rest doesn't matter -- pull the timing belt covers and see what's up, then.

I'd otherwise buy a starter problem, but when I asked eagleb, the response was "well, the crank pulley is turning". As to an uneven cranking rhythm, no one seems to have asked yet -- however, it could have open valves, or "closed" (but bent) ones that aren't sealing, allowing the cranking pace to be relatively even but faster than if there was full compression. Can I assume you dealt with a bad exhaust valve on that cylinder?

Yes, it's difficult to diagnose from thousands of miles away. If only someone were closer... You may be right. Let's see, Corvallis and Beaverton. Hmmm... Since eagleb has a broken hand and is on painkilling meds, how about a trip of about 1&1/4 hours, east a bit on 20, then north on 5 (at least that's what a Google map seems to indicate), to be sure? Perhaps a person who might benefit from that could pick up the tab for the gas, etc.

It seems that the character between the "AU" and "827" is a "0" (zero), not an "O" (Oh).http://www.nskf-bearings.com/Product_number/products_general_catalogue/bearings_a.htm EDIT: An AU0827-4/L588 appears to cross to a Subaru part #28016FC001. That's a rear axle ball bearing, apparently one of those that don't live too long in Imprezas, Foresters, etc. It's usually replaced with an upgraded roller bearing kit for longer life.

Low tens of milliamps is common. Let's do the math. 24 hours/day times 14 days yields 336 hours. At .04A (40mA) discharge (not an unreasonable parasitic draw), that would deplete a bit less than 13.5 A-hours. While that's not insignificant, a typical small car battery's capacity is about 45-50 Ah, so at 40mA you shouldn't be depleting more than about 1/4 to 1/3 of it. That should certainly leave enough to start the car (unless it's very cold, when thickened oil, lowered output from the battery due to temperature, etc., might make the difference). I suspect the battery isn't up to par, or the parasitic load is sometimes more than 40mA.

No apologies are needed. I've been on some heavy-duty painkilling meds myself the last few days, so I can empathize. Good luck -- let us know what you determine, when you get the chance.

I was thinking starter drive problem myself, and was going to comment on that possibility. However, you said "didn't feel like the engine was fully turning over", so I assumed it was at least partially engaged; apparently so did others, hence the suggestions of a broken timing belt. Are the crank pulley and any driven accessories turning? If not, either the starter drive has failed, or the flywheel gear has missing teeth. If the crank pulley, etc., are turning, unfortunately you're back to the likelihood of a broken timing belt.

The TPS doesn't work the same way on all models. Some of them have the low/high voltage relationships inverted. See page 5 of http://endwrench.com/images/pdfs/Fuel.pdf .

Sure. Even assuming the voltage reading is with the battery completely unloaded, it's a bit high. A more typical reading would be 12.65-12.70 volts for a battery in good condition with any "surface charge" removed. It doesn't indicate a problem, but probably that recent charging (possibly just via the alternator) has left a surface charge that wasn't bled off before testing. A 2.5L 4 cylinder engine with average compression and correct viscosity oil requires a battery rated at least 300 CCA (Cold Cranking Amperes) to be able to crank the engine for 30 seconds at zero degrees F without dropping the voltage below 10.5 volts (whew!). If yours has a capacity of 500 CCA or so (I don't know why you've listed two readings), that would appear to be quite adequate. I assume that's the voltage measured at the battery terminals during engine cranking. It's fine, but without knowing at what temperature, whether the engine was "cold" or already warmed, etc., it's not as revealing as possible. Without knowing engine RPM, not as useful info as it could be. Assuming about 2500 RPM, voltage is about right; current is dependent on load, including whether or not the battery was just somewhat depleted by starting the engine. Again, parameters missing -- what RPM, load? Your alternator apparently can output almost 50 amps while keeping the voltage above 13.8 volts, a reasonable performance. If this was a test of parasitic drain, it doesn't jive with your statement that the battery runs down when connected, but doesn't when disconnected. In fact, if the least significant digit can be trusted, it indicates that there's not even 10 milliamps of current drain. Of course, the drain could be intermittent...

Both fans could run if the A/C was involved: http://endwrench.com/images/pdfs/CoolingInfo.pdf The only other thing I can think of offhand is somehow having tied into the green test mode connector wiring.

To a great degree, the above determines the answer to the below. If you can run it at all (even for a short time) without apparent problems, it's likely a properly done repair (that may include more than just HGs) will add lots of life to the engine. One thing that helps determine potential longevity is whether engine maintenance has in general been (and will be) done on schedule. For example, when was the timing belt last replaced (will it need one within that "50 - 75K" period)?

To find software to reduce an image to the relevant portion, you could do a Google search on "image editing cropping", for example. Here's one way to accomplish the task: http://www.updatexp.com/image-cropping.html EDIT: I used MSPaint as explained in the link above, and saved it as a JPG file. I'll attach it to this post and see how it looks. Further EDIT: There's some loss of detail, but I think it serves the purpose, and the file is much smaller.

These may be of interest:http://endwrench.com/images/pdfs/OBDSpring04.pdf http://endwrench.com/images/pdfs/FeedNov06EW.pdf http://endwrench.com/images/pdfs/07FebFuelEW.pdf Good; getting any codes read would be useful.

My point was that care has to be taken when sanding the clear coat, since if you accidentally go through it, there's only the thin color ("base") coat beneath it. I've seen the result of sanding while assuming the clear coat had full factory thickness when it didn't, and it wasn't pretty. If you're the original owner, and know that no top coat has ever been removed with polishing compound, etc., then there's less of a concern. By the way, it shouldn't require 600 to remove haze. For the reasons I already mentioned, I'd use a finer grit.

There are several things that can cause that problem. Assuming it does this with most of your CDs, one common fault is that the output of the laser becomes reduced over time. It's rarely worth the cost of repair, so probably best to look for a replacement. If you do find one at a JY, make sure they have a reasonable return policy should the pulled one be bad.

I'll assume you haven't replaced the rotors with stainless steel ones. Therefore, the disks are likely smooth where the pad runs, but are probably rusty towards the hub and outer edge. That can be a cause of the type of noise you're describing; the edges of the pads can "bump" on the rust until enough pressure is applied. If you haven't done so, besides the more easily seen outer surface, it would be good to inspect the inner surface of each disk as well.

As long as there's no delamination, it might be possible to just compound the clear coat in those areas, and from then on keep a good coat of wax on it to minimize further damage. A light sanding and recoating would of course be an option. Since on clear-coated finishes the color coat usually isn't very thick, sanding (and sometimes even using clear-coat compound) has to be done with finesse. If you're not sure what's left of the finish in the bad areas, try to find a shop with a paint thickness gauge, which should help determine the best course of action. (Even some professional detailers have a gauge.)

If the DTC translates to the "system" is "rich", then it could be indicating P0172. That code is set when the ECU is having to push the fuel trim lean (past a point that's reasonable) to compensate. If that's the code, common causes are fuel pressure problems, bad CTS (coolant temperature sensor), or bad MAP (manifold absolute pressure) sensor.

Depends on the type of failure. Not peeling is good, but are we talking cracks/crazing/"spidering", chips, or just some hazing? How extensive (how large an area is affected)?

Right! It's all about acceleration. [physics discussion]: force = mass x acceleration -and- work = force x distance -so- work = (mass x acceleration) x distance Therefore, once you've got the mass moving at a constant speed on level ground (zero acceleration), work is zero no matter the mass and distance -- at least in theory. In actuality, additional mass carried inside a car (so that it doesn't add to air resistance) will cause a slight increase in bearing and tire friction, and very slightly increase fuel consumption when at constant speed. As Gary said, the energy goes into accelerating the extra weight. That's why carrying more than you have to in mountain driving isn't a good idea, since you fight gravity's acceleration of the additional mass when going uphill, and have to dissipate it as braking heat when slowing/stopping when going downhill. [/physics discussion]

It does have to do with fuel temps, and other factors. Gas at the stations is almost always stored in underground tanks, where it's cool. That gas expands in your car's tank, so effectively you get "more" gas, which will have some effect on apparent mileage. It's also likely that the viscosity of drivetrain lubricants will be lessened in the heat, resulting in lower frictional losses.

Assuming the engine management system (ECU, O2 and other sensors, etc.) is working correctly, all that will happen with a modern car by using "less dense" fuel will be that the fuel trims will change to compensate. It won't run any leaner, so no MPG improvement results. Jerryjetman, the mileage you're getting seems to be close to what might be expected. Terrain (hilly versus flat), traffic and number of stops, weather conditions, "winter" gas, individual driving habits, etc., obviously all play a part in how far a full tank gets you. You seem to have already done much of what could be. One thing here in the Northeast that sometimes is a problem is corrosion at the brakes; slight dragging that isn't otherwise noticeable can cost you up to about 10% in MPG. (You'd probably notice any brake problems that would affect the mileage by more than that.) I assume that Pittsburgh uses salt on the roads in winter -- if so, it might make sense to inspect the brakes and be sure that everything is free to move. For example, it's not uncommon for brake pad "ears" to stick in calipers.

Applying 12 volts directly to the starter's solenoid terminal using a remote switch or just wire will obviously crank the engine. However, if I don't know someone well enough to gauge their care and competence, then I don't suggest that approach. Why? Because doing that bypasses the car's safety switch, and you never know what gear the trans may have been left in.