salewit

Wow you know that may just be the best advice I've gotten on any forum anywhere! Thanks so much. It does concern me that the unit doesn't test correctly, but like I said, the car is running like silk, and my emmission stats were about 85% lower than the maximum.

Here we go again. 92 Legacy. Some might remember me from a couple of years ago when my car repeatedly failed smog. The problem was finally fixed, and I went in this year with confidence that the thing would pass easily. It didn't. It actually passed all the emmissions tests, but failed on a technicality. The CEL light was on. I had no idea I would get failed on this. This was illuminated from the problem 2 years ago and I just never bothered to clear it. Or so I thought. I took the car home, disconnected the battery for half a day. Reconnected it, and the CEL is still lit. I checked the code and its 31. Investigated the TPS and I'm getting weird stuff. The idle switch (Pins 1&2) works perfectly. But test 2 which is to backpin #2 and #4 of the connector with ignition on and the connector connected, is not behaving. I get virtually 0 volts (.03) closed, and about .05 wide open. I triple checked my back pins (with an ohmeter) and they're definitely making contact. I tried checking from #4 to ground and I got the same thing. I pulled the connector and then checked #4 to ground and I get about 3.4 volts! It's almost like the TPS is shorting to ground. But then why do I only have 3.4 volts on pin 5? Everything says it should be 5v. Pin 1 to ground by the way *IS* 5.0 volts. The car has been running smooth as silk, and my emmission numbers are excellent. Just need to get the CEL light out. And I assume smog will check to make sure the light comes on with ignition on and off when car starts. So I can't just remove the bulb. Help!

Well Gnuman you nailed it! I popped the hood and wiggled the connectors behind the battery, then went under the dash and wiggled some connectors going into the ECU, and the problem is gone! I knew something was different the second I turned the key as the CEL came on and stayed on. Tomorrow I'll clear the codes and see if the CEL stays off. I was also thinking of unplugging some of the connectors and carefully checking for bad pins and possibly spray them with an electrical cleaner before plugging back in. But it's running just as smooth as silk and I can't thank you enough for the advice. I just wish I had tried this before spending close to $500 on all the above stuff I didn't need. I've been driving my manual with my foot half-way down on the clutch the past 6 months to avoid this bucking. I'll have to learn how to drive again. Thanks again!

Yep... on the list. It's right on. In fact this is the first thing I did as I had a belt that jumped a couple of teeth when I put this engine in a year or so ago. It worked fine and smooth for more than a year after I fixed that problem. No I'm not sure the replacement MAF is good. Where do I get a good one to check it against without buying one for $200+?

No it's on the list only I called it a PVC instead of a PCV. I've heard it referred both ways. Gnuman: Interesting. The battery was loose on the deck and just floating around. I recently (months after the stumbling started) got a tie down kit. I'll look at that in the morning and give you a call.

92 Legacy. Symptoms: Bucking at light acceleration and coasting. Hot or cold. Ok only on heavy acceleration*. Has spasm attacks during idle. Can be idling smooth as silk and then all of the sudden starts trembling, then back to smooth. CEL light flashes on during moderate acceleration then shuts off. Codes are 12, 13, 14, 15, 16, 17 and 31. Cleared memory and same exact codes came back. AFTER this started, replaced plugs, wires, coil, ignition control unit, MAF sensor, cam sensor, TPS, PVC, fuel filter, air filter, oxygen sensor (twice), temp sender (in rear). Tested fuel pressure, fuel pressure regulator, compression, vacuum leaks, valve timing, ignition timing (but couldn't find mark on pulley). I can't find anything else to change or check and I'm stumped. *This week I had a new symptom. During a heavy hill climb, it acted like it was running out of gas... or the engine shut down.... pumped the gas pedal and nothing. Minutes later it recovered and went back to stumbling and bucking but at least it had power again. The only thing I haven't looked at was the ECU, but I can't find it. My book says it's behind the glove compartment or on the floor under the glove compartment. It's not there (hey maybe that's the problem!). Any help would be really appreciated. I'm about ready to park it next to the junker I stripped the MAF out of.

I've got a (still) stumbling 92 Legacy. A few days ago I noticed the CEL come on for just a second and go out. Every day is worse than the last. Stumbling, shuddering, surging. Now all speeds, hot or cold. Today I read in the CEL codes and got 12, 13, 14, 15, 16, 17 and 31. I cleared the memory (confirmed it), drove around the block and saw the the CEL flash on and again checked the memory and got the same exact thing. 12-17 plus 31. I checked the idle control for about 10 minutes running between idle and full throttle both by checking resistance and signal voltage. I moved the throttle slowly, but everything is behaving perfectly. One thing... my manual says signal voltage should be 0.5-1.0 volts with throttle closed and about 4.5 wide open. Mine reacts the exact opposite, so I'm guess that's a type. Closed 4.5 volts down to almost zero full open. Any ideas? Changed air & fuel filter, PVC, plugs, cleaned MAF sensor, checked coolant temp sensor, replaces cam sensor (had a spare). O2 sensor was changed about 6 months ago before problems.

Yeah they were done post crudiness. Sure there's junkyards all over, but I didn't think they'd ever take a part back. Especially an electrical part. But I'll check it out.

Yes that's me again! Actually this is the first thing I checked when this problem started getting worse. The valve timing is right on. On the plus side I can now get to my timing belt in about 7 minutes flat! What's weird is when I changed all the basic "tune-up" things noted above, the car smoothed out considerably... for a few days... and then went back to its bad habits. Getting worse almost daily. But absolutely no problem with heavy acceleration or climbing hills. Doesn't miss a beat. I haven't had any new CEL messages since I cleared it during my smog fiasco last year. The smog test numbers from what I remember were excellent once the O2 sensor wires were connected properly. The car wasn't scoped. Manual transmission. Can't find my book.... what and where is the IAC valve? I'd love to try swapping out the MAF sensor, but that's a $300 gamble. Is there any way to test the sensor with a multimeter? The plugs I used I think were NGK gapped to .044". As far as getting it diagnosed at a shop... I know I'm being stubborn, but I just have such a bad taste in my mouth from the shop I took it to during my bad smog test that diagnosed it as a bad cat. It turned out to be the sensor connected improperly. This shop was very highly recommended... Valuestar and all that. Not saying that this is a smart move for me to be fumbling around here, but I suspect a lot of you know where I'm coming from on this. Kind of like going to a prostitute.

92 Legacy. Car bucks or surges in ALL conditions except heavy acceleration. Doesn't matter if engine is cold or hot. From idle to highway speed. Most noticeable when steady on the pedal at about 30MPH it'll surge then hesitate. Problem seems to be getting worse. At idle, my tach will sit at about 800RPM, then all of the sudden the engine will shudder for about 3 seconds and the RPM dips to 400-500, then other times it will race up to 1500 RPM. Hot or cold. I've done a lot of searching here and found LOTS of people with the same situation. The problem is everyone has a different "opinion" on what is wrong. I've read injectors, fuel pump, MAF sensor, temperature sensor, plugs, wires, coil, O2 sensor, crank sensor, etc. I don't have the money to just change things left and right and some things I changed a few months ago during a different problem (failed smog). I'm wondering if anyone can recommend a sequence of things to diagnose to help me find out what the problem is. I've recently changed: Plugs, PVC, air filter, fuel filter, oil and filter. Approximately 3-6 months ago I replaced: cam sensor, temperature sensor, O2 sensor and visually inspected inside of catalyic convertor, tested throttle position sensor with ohmeter. Thanks

What I do is take a straight pin and stick it in there between the wire and the connector and probe the pin. If you don't have a pin, a safety pin, or possibly an uncoiled paper clip might fit in there. The connectors on my 92 Legacy would not allow a probe the size of my multimeter in there. No way!

I hate to bring this old thing back to the top, but I just wanted to thank everyone that helped me out with this beast. The car finally passed smog! The culprit was apparently the mis-wired O2 sensor. Cat didn't need changing either. The numbers were all below average. The car is running unbelievably well. Thanks again for all the help!

OB: I don't think I let the engine get fully up to temperature, so I'm thinking the CTS is probably about right on. Alright I just put the old O2 sensor back in with the white from the harness connected to the black on the Bosch replacement and stuck a pin in the connector on this signal wire. The engine had fully cooled at this point. The cat was cold. I started it up, and put my voltmeter on it right away and it was jumping all over the place from .1 - .9 volts. From what I understand this is what is supposed to happen after two minutes or so. I never got the "open loop" mode of just .1-.2V. I have one muffler strap I missed, so when it cools down, I'm going to fix that strap, test it again cold, and maybe go for a spin. Then what? Smog test?

Well I think I'm going to run out and get a new one. I measures the resistance between the heater wires and got 4 ohms. My book says it should be 30. What's $40 after spending hundreds?

Hmmmm you know I think I do have this O2 sensor installed wrong. I disconnected the connector, and measured 12V on the black and red with the engine cooled. So the question is did I fry the sensor or should I try connecting it properly? How would I know if it's good? Also, could this have ruined my cats? If so, which one if not both?

I didn't check. All I did was check between the BLACK and ground. Check this out: http://www.sparkplugs.com/sparkplug411.asp?kw=Universal+O2+Sensor+Installation+Instructions&mfid=0 and click on the connection chart on the left side. I've got a Bosch. And here: http://www.ultimatesubaru.org/forum/showthread.php?t=43580&highlight=oxygen+sensor+connections The Bosch chart here: http://www.boschautoparts.com/NR/rdonlyres/3BB8A8FA-C707-4333-9098-727C01AF4B7D/0/OxygenSensorInstall.pdf is totally useless. But here's the schematic from my book which confuses me more... it looks like the WHT is the signal... which would mean the O2 sensor has been hooked up wrong!

Ok I spent the entire morning on this beast and checked a bunch of things with my digital voltmeter. I checked the throttle position sensor. and had 0 ohms when closed, and it popped to 1 ohm just as soon as I opened it and it stayed at 1 until wide open. Now on the variable rate, I got a little surprise, and I'm assuming my book (Haynes) is incorrect. It says "The voltage should vary from 0.5 to 1.0 volt at closed thorttle to 4.5-5.0 volts wide open". I got the complete opposite. I got 4.6 closed and smoothly went down to 1.0 at full open throttle. The coolant sensor was behaving properly. I had 5v on the harness, and the ohms ranged from 2480 cold to 460 when the engine warmed up and I stopped. According to my book 2480 = 75F and 460 about 158F. I checked for 5v on the harness for the MAF, but that was all I was able to check. My book says I'll need a scan tool to check it. On the oxygen sensor, I got bad numbers. I had set a pin in the connector last night, checked continuity between the pin and the connector to make sure I had a connections. Then today with the engine cold, and my meter on the 2 volt scale, I checked voltage on the signal line of the sensor. This is the BLACK line, correct? Everything I'm reading is saying black is the signal line... connected to black on the replacement sensor, and the red and white are the heater lines. Anyway, I checked cold and got .001 volts. I started the engine and my voltmeter didn't budge. After about 5 minutes it sill hadn't budged. I stopped the car, disconnected the connector and checked continuitity with the pin, and there was none. So I set the pin, made SURE it was touching the connector, reconnected and ran the engine and I got .011 volts on the signal line. I checked the pin again and it was good. Plugged it back in and again .011 - .013 volts at operating temperature. I will check it again and even check the Bosch connectors when it cools down. But so far it looks like this sensor is not working. If it IS the sensor, all I'm seeing are replacements. I haven't found any source that has OEM sensors. How important is this? I would like to find one with the connector already on it though. That funky Bosch system I don't like at all. Sam

I don't know anyone that would have a scope unfortunately. I'm guessing they don't rent those anywhere. Well the big problem is that I now have an expired tag on my plate. Leaving the car on the street I already got two tickets for an expired tag. I got an extension from DMV but that expired in June. Driving it around risks my getting pulled over and getting more tickets. So the vehicle is in our driveway and it's taking a chance to drive it around. I may check on getting another extension, but I was told that they give you one shot and that's it. I wonder if the zero O2 reading was accurate from the garage? What I don't understand (and I don't really need to), is how they analyze the cat. Aren't they just picking up these readings from the tailpipe? If so, why couldn't the smog place flag this as a bad cat? It took about a quart every 3 months and the car was used by my wife who is a sales rep and used the car heavily at the time. She also thought it was due to leakage more than burning, but that's what it is. Do you think I just need to take this out on a hot day and on a demanding road with the AC on and just try and blow out all this crap? Can it be that simple? I already did the Seafoam thing early on in the process.

Okey doke... I reinstalled the cats, got a fuel gauge and hooked it up between the fuel filter and the fuel rail. Even 27psi at idle. Removed vacuum hose on the regulator and it jumped up to 36. Pinched return line for a second and it jumped up real high. Shut engine off and checked gauge after 5 minutes and it is dead on 30psi. According to my book, the "hold" after 5 minutes test can detect leaking fuel lines, bad regulator, leaking fuel injectors or a fuel pump. The specs in my book call for 26-32 normally, 32-38 with vac line disconnected from regulator, and 30 hold press. All 3 numbers I got are exactly where they should be, so I think it's safe to rule this one out. I started the engine again and let it warm up for awhile. I then shut it down, propped open the throttle full, pulled the plugs, pulled the injector connectors, and pulled the harness on the coil. I cranked 7 turns on each one. All 4 cylinders jumped immediately on the gauge. #1 was 180, #2 was 185, #3 was 195 and #4 was 180. Now I'm more depressed than ever. No matter what I check, everything is checking out. I took a picture of the plugs, and they are not as fouled as I remembered them. It's hard to see in the picture, but basically the plugs are sooty black except for the tip. Any possibility with all the things I've changed, I fixed the problem, but just never got it hot enough for smog check? Hmmm but then the repair shop also got high numbers too and you'd think they'd know enough to get things warmed up. Oh and I re-read your previous post. We've been using this car for about a year until I was of course forced to get the smog test. In that year, we had to put in a quart of oil from time to time, but I certainly wouldn't classify it as an oil burner. An oil leaker maybe....

Well I'll be working on this thing today and I could do another compression test. That test was done before the valve timing was corrected. The O2 sensor is not an OEM. But this sensor was changed to the same non OEM (Bosch with the funky wire connector system) a few years ago, and it passed smog with no problem. By the way, when disconnecting the cats and the O2 sensor which is brand new... maybe 15 minutes of running time on it, it is sooty black already. Let me ask this... there is no way of putting lets say a "Y" on the ECU to see what is feeding it the bad numbers to increase fuel? The service place I picked to diagnose my problem is a Valuestar repair shop which means it's highly rated for satisfaction. They've also been in business for a long time. Not that any of this means a whole lot, it does lead me to believe that if customers are happy with their repair work, they're getting things done right the first time. Just my assumption. But today I will put the cats back together, run the fuel compression test and then the engine compression test. I'll report back.

This is the service stations report and he's (or she) has got just a zero (0) in the O2% columns. The temperature gauge when fully warmed up, sits almost horizontal (3 o'clock or 9... I forget which way it goes). Maybe 15 minutes or so lower. And it takes a very short time to get there. Maybe a 1/4 mile. My smog place is 3 blocks away. I'd drive it around the block about 4 times and she's showing warmed up. Not sure what you're getting at, but when I was getting ready to pull the engine a few months ago, and I pulled the top radiator hose, it was powdery white. I'm told this is from an air bubble up top. I know Subaru's are hard to get the air out of the cooling system. Since then I've refilled and massaged the heck out of the hoses to try and get the air out, and I'm pretty confident (but not positive) that the air is gone now. Thanks everyone for sticking with me on this.... I'd be lost without you guys here.

Ok well I'm a moron. All day long I was trying to remember to pick up a pressure gauge (my fuel pressure gauge only goes up to 12psi... it's really a vac gauge). Anyway, I didn't get the gauge, got home, jacked up the car and removed both cats. It was only after I broke a couple of bolts in the process did I remember I needed to check the fuel pressure FIRST. Duh! Anyway, both cats are out, and they're both intact. Nothing broken inside them and no rattling. The honeycomb is black. Here's the best pic I can get (rear cat) Here's wha the receipt from the car place says: Do a 4 Gas Tailpipe printout at Idle and 2500 RPM... Idle 243HC, 2.79 CO%, 13.6 CO2%, 0 02% 2500 84 HC, 1.21 CO%, 13.1 CO2%, 0 02% "We did a cat efficiency test. The cat failed the test. The HC's when hot was at 755ppm, it should be below 500ppm. The CO2 was at 3.5%, it should be above 9%. The cat should be replaced. Should I put the cats back in and test the pressure or just replace them both (at least $400) and test the pressure?

So are you saying I should change the regulator before the catalytic converter? If I don't change the regulator and it's bad, will that damage the new cat?

Well the saga continues. After 3 years of high school automotive technical school, and 2 years of State University of NY for auto tech (in the 70's... fairly useless now), I bit the bullet and brought it into a local garage for diagnosis. My first time going to a garage in 30 years! I felt like I was going to a prostitute! Anyway, they say the catalytic converter failed every test they gave it. When I asked how much, they told me they'd have to farm it out and get a price. I told them I'd *try* to do it myself. The question I have is this. When I go to Autozone or whatever, they list and front and rear cat. My repair book also shows a front and rear cat. When I called the garage back asking if they knew which it was, they thought there was only one in the front (where the maniforld Y's and the O2 sensor is). Do I have to replace them both? Is there any chance whatsoever I can do this myself? I do have a stick welder if welding is absolutely necessary. Is there anything wrong with using cats from a local parts store? They're both under $100, yet the OEM's are like $500. Thanks for all the help here. I noticed someone in this thread mentioned the catalytic.

Ok.. changed the O2 sensor, went in for a smog again and she failed again. Idle HC 233 (120 Max), CO 3.36 (1.0 Max) 2500 HC 66 (140 Max), CO 1.84 (1.0 Max) I'm lost now. I don't know what to do or where to turn. I've replaced the PVC, Cam sensor, Coolant sensor, Oxygen sensor. The oil has been changed, there are no vacuum leaks. The car runs smooth as silk. Any further suggestions? Tried the "magic potion" on this last round as well. I'm convinced it is something ON the engine (since my last engine always passed smog).