OB99W

The ECU's determination of catalyst efficiency is based on a comparison of output from the front and rear O2 sensors. Under normal operating conditions, the front sensor rapidly switches from somewhat over zero volts to somewhat under 1.0 volt and back again, while the rear sensor switches much less frequently, and over a narrower range of voltage. That's because when the catalyst is doing its job, oxygen in the exhaust is used by the catalyst. A sluggish front O2 sensor can cause the ratio of gases in the exhaust to change enough so that the catalyst can't operate as designed. In that case, the rear O2 sensor could see fluctuations in oxygen level that it shouldn't. The ECU can inaccurately interpret those variations as a failure of the catalyst. Even if we assume that the cat should be replaced, at a minimum the front O2 sensor should also be changed to provide the new cat with a better chance of survival. It therefore makes sense to try changing the sensor first, since it's a lot easier to replace and less expensive than a cat, even if the cat is eventually changed anyway. Of course, if an OBD-II scan tool is available, the best approach is to look at some real-time data from the O2 sensors, see what the long and short term fuel trim numbers are, etc. That info would minimize the chance of a wrong interpretation of the cause of the code.

Welcome to the forum. Fuse SBF-2 would be a good place to start checking, since it provides power to the ECU, fuel pump relay, main relay and the OBD-II connector.

If the plugs and ignition wires haven't been replaced in a while, it wouldn't hurt to change them. Experience shows that using NGK plugs and Subaru OEM wires tends to be the most reliable way to go. If the O2 sensors are really originals, at 219k miles they're probably a little tired. If you have access to some test gear, monitoring the real-time operation of the O2 sensors and some other data could help narrow the diagnosis. If you have no way of looking at OBD-II data (other than reading trouble codes), I'd start with changing the front O2 sensor, which has primary effect on how the ECU sets mixture once the engine warms sufficiently. It's important that an OEM sensor be used. If you decide to change the rear O2 sensor, it's less critical and a good quality direct replacement would probably be adequate.

The OEM unit for a 00-03 Legacy is reasonably priced at about $20 list. I don't know for sure, but I suspect the Delphi replacement goes for at least as much. It was nice of your friend to allow you to try those parts, and to determine that they're not the cause of a problem. (Perhaps buying non-OEM has some advantages.) It's still important to verify that there's strong spark at the plug end of the wires on 3&4. You can check that by pulling each wire at the plug, and seeing how far the spark will jump from that end to a nearby ground (the block). Of course, make sure you don't get shocked, and don't allow the spark to jump to any other wires. As you said, your gauge might be a little optimistic (typical for your engine would be somewhat lower), but at least the readings are well balanced. There could be some carbon buildup. However, it's also possible that the valve (cam) timing is a bit advanced, due to the timing belt being off a bit. If you haven't recently verified that all the timing marks align, it might be worthwhile checking. The OBD-II trouble codes are only as reliable in pointing to problems as the ECU's programming. If something is wrong that the programmer didn't anticipate (or isn't easy to detect), the trouble code(s) generated may not accurately reflect the nature of the problem. The reason I keep coming back to the timing issue is because the ECU ''assumes'' crank versus cam timing is correct, and directs systems to function based on that. Both fuel delivery and ignition timing/dwell are very much dependent on the pulses the ECU gets from the crank and cam sensors, which if mistimed will throw everything off. The P0340, possibly higher than normal compression, and the apparent fuel (and possible ignition) issues might be linked via the timing belt alignment. (And yes, I know it's no fun to pull things apart to check.)

Thanks. Please tell Tim Maguire what happened. While it was a Bill Cooke location, things were mostly done right (although as typical at dealerships, not inexpensively!).

In CNY? Cortland, 'Cuse, Ithaca?

Burned oil can contaminate O2 sensors and/or the catalyst. A too-rich mixture, which could have several different causes, can prevent the converters from doing their job. Following are some links to P0420 info, most not Subaru-specific but interesting anyway: http://www.motor.com/magazine/pdfs/102006_09.pdf http://www.wellsmfgcorp.com/pdf/counterp_v7_i3_2003.pdf http://www.wellsmfgcorp.com/pdf/counterp_v7_i4_2003.pdf http://www.catalyticconverter.org/news/news_page.cfm?Key=codes-po420-po421-po430-po431&News=120

The static pressure is too high. Typically at 90 degrees F it should be between about 75 to 100 PSI. (A rough ''rule of thumb'' is static pressure in PSI equal to temp in degrees F.) The can of R134a that was added probably wasn't needed. Based on your first post, with the system operating the low side pressure was too high, and the high side was too low. I agree with your thoughts that the compressor is bad or there's a control valve problem. R134a isn't an ozone-depleting gas -- however, since it is one that can contribute to climate change, it would be good to take the car to a pro who can recover the excess refrigerant and do a proper diagnosis.

You're welcome, Cougar. The wire coloring info you provided agrees with what I have. I'm beginning to wonder if certain assumptions have been made that could be leading us astray in diagnosing the problem. I reread rvac99's first post in this thread, concerning the appearance of the plugs and the smell of the exhaust, but no details were given. It's possible to misread similar symptoms. For example, plugs that are sooty and exhaust that smells of unburned fuel would usually point to a rich mixture. However, if the plugs are wet and the exhaust smells of raw fuel, one might also assume that the mixture was too rich, but it would more likely indicate that reliable combustion wasn't occuring -- along with the cam sensor code, that could indicate a timing-related problem. A quick check of compression could help determine if timing issues are likely. If compression is low and if all the belt covers haven't been pulled and the crank/cam timing checked, it would probably be one more thing to consider looking at (especially if investigating #3/#4 ignition hasn't lead anywhere).

Ever have your head under the hood when someone ''accidentally'' hits the horn?

It's probably best not to clear the ECU memory until you have any stored trouble codes read, since they could help determine what the problem is. Typically major auto parts stores like AutoZone, Advance, etc., will do that at no charge. If you have it done, try to get the code(s) (Pxxxx), not the code definition(s), and let us know about any.

Unlike earlier models, there is only one engine temp sensor on the 2000. It's a 3-wire, and incorporates both the ''thermometer'' (gauge) and ECU functions in the single unit. (The white/green wire is for the gauge, but there are two others.) A problem with either function naturally requires changing of the CTS. EDIT: I attached a picture of the TPS connector end to minimize any possible confusion.

You're welcome.

Those cans with a built-in gauge won't replace a manifold gauge set, but they're not as useless as it might seem. If you know the refrigerant being used, and the ambient temperature, it's possible to determine the approximate level of charge by reading a single gauge when the engine is off (that is, compresssor not running and time allowed for the system to equalize). If that's in the ballpark, and with the system running the low side pressure is also about right, it's likely the system is close to correct charge and operation. Having a high side gauge is nice, but not always absolutely necessary.

I don't have specs for Subaru horns, but typical car horns have output around 108 db SPL (I believe the spec is for the level taken at one meter distance from the horn). Truck horns tend to be about 10 dB louder, so one outputting 118 dB is probably meant for a truck. Up close, that level can actually damage your hearing, especially if sustained for any significant length of time. Ten decibels might not seem like much, but our ears and therefore the dB scale aren't linear, they're logarithmic. That 10 dB equates to 10 times the sound power. A 20 db difference would be 10 times 10, or a 100-fold power increase. By the way, since sound level is relative, it's always a comparison to something, and a reference is either given or assumed by convention.

Compared to what? Do you mean in dB SPL (Sound Pressure Level)?

Something that's common to cylinders #3 and #4 is that they are paired in the ''wasted spark'' ignition. If there's a problem with the rear section of the coil pack/igniter, you'll get a weak or no spark on those two cylinders. It's also possible for a bad ignition wire for either #3 or #4 to affect both cylinders. Furthermore, the ECU could be failing to provide a pulse for 3&4. So that begs a question: Do you have good spark at the plug end of #3 and #4 wires? It also brings up another question: If cylinders #3 and #4 are working so badly that you find no difference in the way the engine runs when you disconnect spark from them, why isn't the ECU indicating P0303 and P0304 misfire codes? Are you sure that P0340 is the only code stored?

I suggest you get an OEM O2 sensor. It would be useful to know what voltage(s) the shop read from the sensor that lead to the diagnosis. Other than if the catalyst is found to be physically damaged, the only other way you can tell for sure if a cat is bad is to measure the exhaust gas composition before and after the cat. Some shops will do that by pulling the O2 sensor and measuring both there and at the tailpipe, and then comparing the numbers. If they don't see an improvement at the tailpipe, they assume the cat is dead. However, as I've explained before, if the engine is running so rich that there isn't enough oxygen left in the exhaust, the cat can't work (a cat oxidizes, but not if there's no oxygen to work with!). If you have the patience to read another long thread (I think you'll find it worthwhile), see http://www.ultimatesubaru.org/forum/showthread.php?t=76747 -- the poster had been told his high Hcs and CO were due to a bad cat, but I was unconvinced and eventually it was determined that the O2 sensor was misconnected. Resolving that fixed the emissions without replacing the cat. Since you intend to replace the O2 sensor, do that and check the emissions again after a ride to possibly burn off some cat contamination. As to the injectors being ''clogged'', that usually means they aren't putting out enough fuel, or that the spray pattern is uneven. Since you're experiencing a rich mixture, it would be more likely that if anything is wrong with the injectors, they would be staying open too much. The Seafoam probably won't hurt anything, but I'd start with the O2 sensor first. I'd be interested in reading what they had to say. By the way, did they check the fuel pressure?

What do you mean by the gauge ''never moved''? Are you saying that it remained at the bottom of the range? If so, it's possible you have a bad connection at the sensor -- you may have to scrape the connector contacts if they corroded due to the leak. Did you get an OEM sensor as I suggested previously? EDIT: Is it possible there's a simpler explanation -- the gauge didn't move because the engine wasn't run long enough to warm it? If so, the CTS replacement may have actually resolved the erratic gauge operation. I've run across situations where there are multiple problems, but the car only gets worked on when the most recent one is bad enough so that it can't be ignored. If the driver isn't sufficiently observant, the problems are sometimes then all reported as having occured simultaneously. For example, is it possible that your wife didn't notice the temp gauge fluctuations, and the assumption has been that it's related to the other symptoms?

Thanks for getting back to us with ''the rest of the story''. Considering the possibilities when a timing belt jumps on an interference engine, you probably couldn't have asked for a much better outcome.

If you haven't done so already, you might want to recheck that all the mounting hardware is tight. It's also possible that the front end geometry changed enough so that something that was okay before might now be noisy. This is ''old news'', but the problem might be in the same area: http://www.endwrench.com/pdf/steering/FrontEndPopNoiseInfoSp99.pdf

Unfortunately, the CTS is one of the small parts on a Subaru that's a bit of a pain to change. What, you don't have a spare (car, that is)?

Based on his past threads and that there's no AT oil temp light, it seems that Rob98obsport has a manual trans. If that's correct, it would of course mean that the AWD is via a viscous coupling, not transfer clutch as in an automatic. Since there would be no ''C'' duty solenoid, there wouldn't be the FWD fuse holder either. If there isn't a tire size or inflation issue and the brakes are okay, then the viscous coupling should be looked into.

Since coolant has seeped through the sensor, in addition to the gauge portion it's probably compromised the ECU section as well. That in itself might account for the rich running condition. The next step is obviously to replace the CTS. If the connector contacts are contaminated, of course clean them. Many circuits in a vehicle have loads such as relay coils, lamps, motors, etc., that will measure low resistance. Even some semi-conductor circuits will do that. Their nature changes when powered, and they don't draw excessive current when operating normally. Of course, if there were an actual short to ground, a fuse would blow. As long as you haven't changed anything in those circuits, they're probably okay. (I'm assuming the grounds you added are just in parallel with ones that were already existing.) You're welcome -- hopefully the problem will get resolved soon. Let's see what happens once you replace the CTS.