OB99W

Neither of these test results gives me a lot of confidence. The 150 psi your mechanic found is low -- 185 psi is more typical for the engine. The numbers the Subaru dealer reported are possible for compression readings, especially if there's some carbon buildup, but for leak-down they make no sense. Leak-down testing results are given in a percentage, not in psi. Compression and leak-down tests can give an indication of ring condition and how well valves seal to their seats, but they don't reveal the performance of valve stem seals. If the two-month-old PCV valve was more gummed than the much older one which you previously replaced, it might have more to do with the change you made in oil than anything else.

If you reset the ECU (clear the codes) and run a couple of drive cycles, do codes 51 and 33 return?

Congrats. Subarus don't have particularly noisy ''lower'' ends (do we still call them ''lower'' or ''bottom'' on a boxer ?). However, if the engine was run with two cylinders not firing for any significant length of time, the oil might be fuel diluted. Change the oil and filter as soon as possible, and hope that results in less noise.

Thanks for the compliment, although I make no claims to such powers . However, one doesn't have to be a ''wizard'' to understand that when a timing belt breaks, the camshafts stop turning --the ECU no longer receives pulses from the cam sensor, thus setting a code for that sensor. That sequence of events isn't unique to Subarus. As to whether the mechanic is culpable, I can only repeat what others have said concerning parts that should usually be replaced along with the timing belt, especially on a high-mileage inteference-design engine.

A broken timing belt on a 1996 2.2 would not have been a big issue, because it wasn't until '97 that the 2.2 was changed to an interference design. However, I believe the car in question is an LSi, which came standard with the DOHC 2.5. That explains the previous need for HGs and why a broken timing belt is, ummm, ''unfortunate''.

PA allows inspection to proceed with two monitors not ready on 1996-2000 cars. See http://www.drivecleanpa.state.pa.us/drivecleanpa/docs/readiness.pdf

As I said, snapped timing belt. The story continues at http://www.ultimatesubaru.org/forum/showthread.php?t=102972

Who would have guessed? See post #3 of http://www.ultimatesubaru.org/forum/showthread.php?t=102756

Have the monitors on the car tested as ready before this? The 1996 (and some early-production 1997) Subarus had a problem where the monitors are set to ''not ready'' (incomplete) each time the ignition is turned off. It's known, and such cars are exempt from the requirement. See appendix D of http://www.epa.gov/OMS/regs/im/obd/r01015.pdf Also see http://www.ultimatesubaru.org/forum/showthread.php?t=100936 and links within it.

It's possible for a defective engine coolant temperature sensor to incorrectly report to the ECU that the engine hasn't warmed up when it actually has, and leave it running open-loop. That will definitely kill gas mileage. What's the idle RPM?

The igniter wasn't integrated with the coil until more recently. I believe it's mounted fairly high up on the firewall. It looks like the picture in this link: http://www2.remartautoparts.com/Items/LX-682-STP02. If you need an igniter, get the number from yours and pick up one or two from a junkyard if possible ('95-'98 models should be correct, but match the number), rather than pay the high price of a new one. Yes, the ECU can fail in that manner, but it's not common. The igniter or wiring/connections to it are more common problems. If you have access to a volt/ohmmeter and know how to use it, I can post some EndWrench links to help with diagnosis.

Sometimes oil leaks are due to excessive blowby (exhaust getting past worn piston rings), especially if the PCV is clogged, so check that. If the smoke is bluish-white, that's oil burning. If enough oil is burning, it can effectively reduce octane rating enough to cause knock, which in turn can cause the ECU to retard ignition timing, and retarded timing will reduce performance and gas mileage.

I'm going to assume that the emissions numbers posted were for an IM-240 test. NY doesn't require IM-240, so I wasn't that familiar with typical results. I looked into it, and it appears the Colorado limits are very generous -- even if the car passed, the emissions are high. It seems that if the engine is properly warmed and running closed-loop, more-typical results would be about: HC 0.1 gms/mile CO 1.5 gms/mile NOx 0.6 gms/mile Those are quite a bit lower than the readings posted. It seems that the first thing to do is verify that the ECU is going closed-loop -- if not, high emissions are expected. If it's staying open-loop, it might be that the coolant temperature sensor is defective. If the ECU is going closed-loop, then verify that the front A/F (O2) sensor is working correctly. As long as that's okay, the probability is that the catalyst isn't up to the job. Since the original motor had two bad cylinders, it might have dumped raw fuel into the cat and damaged it.

You're welcome, Mike. The EndWrench article describes the way to get readiness as quickly as possible, but it can certainly be done under less-ideal conditions. You might even find all the monitors ready when you next check. If you do a highway run, mainly try to minimize throttle movement. Anyone have a 4-wheel chassis dynamometer at their disposal?

Yes, ''interference'' refers to valves hitting pistons or other valves when the engine is mistimed, due to either wrong timing belt installation or breakage.

The ECU would have tested for EGR operation if there was one on the original engine, and noticed if it was missing -- since you didn't report a CEL and code, I suspected there was no EGR on your original engine. Subaru has used EGR off and on in various configurations, and I admit that I haven't memorized each of them. Earlier models (my own 1999 Legacy Outback for example) had strictly vacuum-operated EGR, with no electrical connection. I'd like to reconsider the emissions numbers. Even though the HC and CO are well within the limits you posted, they may still be high. Could you provide the units (ppm, %, etc.) for the HC, CO, etc.? EDIT: Usually AWD vehicles are exempt from chassis dynamometer testing, but if they run a 4-wheel dyno the test could be an IM-240. If so, the emissions might be in grams/mile. Is that the case?

To determine if the original engine had an EGR valve, look at the emissions label pasted on the underside of the hood. What wiring? Since the emissions numbers other than NOx are ''in the ballpark'', it's unlikely that the cat isn't functioning well enough. Anything you can do to cool combustion temps will lower NOx. As I said previously, be sure there are no intake vacuum leaks that might lean the mixture. Make sure the cooling system is up to par. Don't use more than 50% ethylene glycol in your coolant, since water does a better job when it comes to getting rid of heat. Even an OEM Subaru thermostat versus an aftermarket one can make a difference when things are that close. Why was the original engine replaced? (That might give us a clue as to what's going on now.)

Welcome to the forum. Not to put too much of a damper on your plans, but are you sure you want to go the JDM 2.5L DOHC route? $900 is a lot to spend on an unknown. First, the DOHC 2.5 head gasket failure is common, so there's a chance that your import engine will suffer from the same problem unless it's had the latest HGs already installed. The claim of only 40k miles is also dubious, given that those engines are now around 12 years old. Such an engine, even if the HGs are good now, should probably at least see a new timing belt and related parts -- the engine is an interference design, and if the mileage is really higher and the belt goes, you'll have a lot more than HGs to be concerned about. It might just be better to reconsider working on your current engine, assuming besides the HGs it's okay (as in hasn't been severely overheated, etc.), even if you have to pay someone to do the work. Or, you might get something domestically. I'm not especially fond of it, but some folks install an earlier (non-interference) 2.2L in place of a 2.5L.

Those are likely (but certainly not the only) culprits. See http://www.catalyticconverter.org/news/news_page.cfm?Key=catalytic_converter-&News=120 That's right, leaks well beyond the downstream (rear) O2 sensor won't cause P0420 to be generated. It's usually not too bad with '97 and later models. After the engine is warm, a 10-15 minute run on the highway at constant speed will usually do the trick. See http://endwrench.com/images/pdfs/OBDInfo.pdf .

As with many problems, there are a few possible causes of excessive NOx. If we assume that the cat is okay, then it pretty much leaves things that cause the combustion temperature to be too high. Generally, that would suggest either the mixture is too lean, or the EGR isn't feeding enough exhaust back to the intake. If you can post other emissions numbers (CO, HC, CO2, O2), I can give a better idea of what to concentrate on. Otherwise, you could start with checking for intake vacuum leaks and/or cleaning the EGR valve and passages and see if that helps.

You're welcome. Let's hope that it's resolved long-term. Although in your case the problem apparently was internal to the transmission, this sort of trouble is often caused by external wiring/connection problems, since there's more exposure to the elements and things that can cause physical damage.

Here's a link to a picture of one style of Subaru tach: http://www.lovehorsepower.com/images/Subaru/tach.jpg. It doesn't have the smaller increments, but like those that do, the first mark after that for the zero would indicate 500 RPM -- the remaining increments are of 250 RPM. Therefore the next (not as heavy) mark would be 750 RPM, and the ''1'' would of course be 1000. While 600 RPM is slightly lower than typical for idle, it isn't too far off. However, if what you're describing is close to the 500 mark (about 1/3 of the way to the ''1''), that would be somewhat lower than expected. Being a half-mile high will affect operation to some degree. Simple things like a dirty air filter should be checked for if not already done. Taking an intake vacuum reading could help to determine what's happening.

Just how slow is ''super slow''? They normally idle around 700 RPM. The first mark above zero on the tachometer is at 500 rpm, with each small increment being 100 RPM, so 700 RPM would be 2 increments above the one for 500. Actually, that's not a given. The engine management system should detect the additional load of the A/C and compensate for it. I have some ideas, but knowing the idle RPM and answers to the following questions would help direct a diagnosis. Is the idle smooth or rough? If you're in a mountainous area, what's the elevation?

True, that's what is done. The injector is cut to disable the cylinder, rather than killing spark, because the other way around un-ignited fuel would just get pumped out at every cylinder #1 exhaust cycle. See: http://endwrench.com/current/spring04pdfs/InsiderInfo.pdf. It's true that the wasted-spark pair arrangement means firing cylinder #2 would also provide spark to #1. However, at the time that #2 is being lit, #1 is at the top of its exhaust stroke, and of course if the injector has been cut there's nothing to ignite. But that doesn't mean that the ECU couldn't cut ignition at the time #1 is near TDC of its power stroke (since doing so wouldn't affect #2 then, which would be at the top of its exhaust stroke) -- it's just more practical to cut fuel. Getting back to ron98's problem, the first thing to verify is that #2 is getting a good spark. If so, then the ignitor and coil are probably okay, and as Cougar suggested, a damaged sprocket reluctor is a possible cause of the problem. Also, sometimes a weak enough signal from a position sensor can primarily affect firing of one particular cylinder.

The best way to deal with the problem is probably ''air exchange'' -- that is, ventilate. Open all the doors of the car before entering it for as long as it takes for the smell to mostly dissipate. Unfortunately, once the VOCs are no longer annoying (because they've mostly already been exuded), the plastics in your car will likely become brittle, possibly cracking and otherwise falling apart. Maybe the manufacturers can be convinced to to add a charcoal canister to the car interior, and purge it by connection to the EVAP system. Might even improve gas mileage. ;)