Mikevan10

Sorry for the delay (I know you guys are on the edges of your seats), but here is some freeze frame data: TROUBLE CODE P0446 ENGINE RPM 800 CALC LOAD 5.1% MAP 9.7" Hg COOLANT 138 F LT FTRM1 2.3% ST FTRM1 0.0% VEH SPEED 0 MPH FUEL SYS 1 OPEN FUEL SYS 2 N/A That's it. Drove the car on another 100 plus mile trip, which again included highway cruising and city driving. OBDII I/M readiness check still says "Not all supported on-board readiness codes are complete"! And it goes on to show "Catlyst Mon" and "Evap System Mon" as "inc". Everthing else shows as ok or N/A. There is now well over 200 miles on the car and many engine starts since the codes were cleared and it still is not ready??? I cannot believe this is normal, expectable or even tolerable. This is supposed to be a "READINESS" test, not a pass/fail test so even if my catalyst system, for example, was malfunctioning, I would expect OBDII to say the test was complete even though it may give me a DTC. Right???

I don't know about West Chester, but in my neck of the PA woods (that's redundant isn't it), it's all about OBDII. You could be belching anything out of your exhaust pipe but as long as OBDII (Big Brother) doesn't know it you are golden as far as an "emmissions" sticker goes. Great system, aint it?

I have not tried swapping out the valve lately. But this code first appeared about two years ago and I replaced the vent/purge valve at that time to no avail. Code has existed since then. So the fact that the monitors have not "set" in now well over 100miles of driving in various conditions is normal? Really makes getting a car through inspection a challenge doesn't it? Mike V

OK, I'm back (sorry). Since we last spoke, I have learned that the particular "model" (i.d. number) of ECU that is in my car (AD050 in case anyone's interested) was ONLY used on 1997 2.2 Legacys with manual trans and that were California spec!. That really bites since it severely limits the availability of used examples of this part. This kind of practice unnerves me. Anyway, I could not find one locally so I went ahead and mail ordered one from so no name junk yard in Jersey. Should be less than 60 bucks with shipping so I figured I roll the dice. New ones are on the high side of $500.00 and I don't even know 100% sure that I need one! Here's some more new info that may be interesting: My wife drove the car about 100 miles, round trip, last night. This included highway and city driving. When she left there were several OBDII parameters that were "not ready/incomplete" since I had cleared codes recently. Well, when she returned last night I ran out with the trusty Actron CP9145 scanner and read the codes. The "good" news is that there is still only one fault code (P0446 of course) but the troubling news is that there are still two parameters that are coming up incomplete when you do a I/M Readiness check. These are "Catalyst" and "EVAP Sys". Everything else shows ok or N/A (and some of these did show "inc" prior to the drive). What say you now? Thanks for your time and consideration. Mike V

Now, why wouldn't Red's ECL be on if his motor is acting up so badly? Obviously he has at least one cylinder MISFIRING. And I gotta belive a dodgy ECT sensor would be a great reason to light the CEL. Sounds like there may be deeper problems here.

johnceggleston said, "this is just my observation, but based on the fact that some of the engines". I am not sure exactly what you mean, but if you are suggesting that Subaru used the same ECU in several models and/or years, that does not square with what I was told a few hours ago. I ran over to a local job that specializes in Subaru repairs. They also have a Subaru salvage yard of sorts and they sell used Soobies. I asked them if they had an ECU for a 1997 Legacy, 2.2, California car. The guy referred to some documentation and then told me I needed to get the numbers off of my ECU since they used more than one version in vehicles meeting my description!

Thanks again for everyone's comments. I need all the support I can get! Here's how it all looks from here (in no particular order): 1. Is it really practical to repair an ECU without board level drawings, etc? 2. The scanner I am using does capture "freeze frame" data. Last I looked at it, it indicated that it picked up the "fault" right away because the coland temp was close to ambient. Idles speed, very low load, vehicle speed=0, etc. All stuff you would expect to see upon just starting the engine. I will write the exact info down and report back just in case it may reveal anything. 3. I will begin my search for a locally available ECU out of a junked vehicle today. 4. As far as "swapping them out", I am afraid that no one I know would want to risk it (assuming I knew someone who's car used the same ECU). 5. As far as the cost of trying to solve the problem by putting in a new ECU, I am afraid that this points to the unfortunate condition that most folks don't have the knowledge/interest in trying to find the root cause of problems like this. They take the car to the dealer who then shotguns in new parts until the problem goes away. Then he hands the poor slob owner a big fat bill. What's he to do? Oh yes, OBDII is such a wonderful thing (I read it in Endwrench - gotta be true). More later Thanks again all! Mike V.

OK, I went ahead and tried what Cougar suggested: I attached (rear connected) a ground jumper to the purge valve solenoid pin (B134 pin 10) and, after clearing the fault codes, ran the engine. As ususal, after starting the engine the second time, the CEL came back on. Upon reading codes, I had two of them: P0446 (as usual) and now also PO443 (which also somehow realtes to the EVAP system). So I cleared the codes and retsted. Now I am back to just getting P0446 after starting the engine twice. So, my friends, is the consensus (conclusion?) that my ECU is defective??? Any other possibilities? I do understand Ohms law and that voltmeters are high impedance devices and ohmeters only flow low currents, etc. But the checks I did make all indicate the harness is fine. I have also visually inspected virtually all of this circuit and everything looks to be in fine condition. The solenoid pulls the purge valve in with a positive "click" when I ground B135 pin 10. Suggestions? Running out of hope, Mike V.

Has anyone ever used a "Mitchell" CD ROM service manual? I jsut got one (via Ebay) thinking I was getting a Subaru factory service manual and boy was I mistaken. I only had time to play around with the thing for a few minutes last night (was tricky just to "install" the darn thing on my PC) and it is not a vry user friendly fortmat. And I have serious doubts about it being any more detailed and model specific that a Haynes manual. Anyone used one of these to good effect? Mike V.

Today's update: Ran the tests that OB99W suggested and here's the results: Voltmeter connected between ECU connector B134 pin 10 (back probe, connector still hooked up to ECU) and ground. Ignition off: 0 volts Ignition on: 12.5 volts Engine running: 13.7 volts I let the engine idle for several minutes and reading did not change. Revving engine had no effect (voltage may have increased a little due to higher alternator output at increased engine speeds). Engine was up to operating temperature. Apparantly, the ECU was not gorunding/completing the circuit. Any ideas what to try next? Also, where can I procure a Subaru Factory Service Manual? Thanks again, Mike V.

Thanks for chiming in '99! I will run the test you suggest tonight and I'll report back. I have been wondering under what conditions the good ol' computer tells the purge valve to open. Sounds like you understand it to be open pretty much any time the engine is actually running? Since P0446 seems to ONLY be related to the electrical circuit (no one has opined otherwise nor has anything I have read specifically indicated otherwise), I think we have, at this point, boiled it down to one of two things: 1. Bad connection between the ECU and the B134 connector. or 2. Problem internal to the ECU. Cougar has been man enough to go in and do internal repairs on the ECU. Now that's persistance!

Thanks C3! Yeah, I found the "Main Relay" up to the left of the driver's feet and I did notice the round realy in the same vicinity which I now know must be the fuel pump realy. Good info. Thanks again. Need to get one of them factory manuals. Hopefully I have one on the way to me. Mike V.

I really don't believe I was fully screwed on the knock sensor I bought from Liberty Subaru. When I initially inquired about the part, the Liberty parts man said price was $97.00. By the time we got around to finishing up with the shipping details, he asked me where I heard about his company. He said that if I told him I read about it in Road & Track he'd give me the discount price of about $75.00. Also, I checked price at "1st Subaru Parts" or whatever it is out there in WA state and they wanted about $75.00. Are you sure you are talking about an OEM sensor for $50.00 or less? Not that I know if a brand X knock sensor would be any problem - may be perfectly good as far as I know in which case Liberty did screw me.

Thanks Coug! From your comments I was able to loacte the relevant relay and verify I had the correct one by identifying the wire colors. Anyway, I tested this relay and it checks out fine: contacts are open when coil is deenergized and both pairs of contacts close (less than 1 ohm resistance) when the coil is energized. And the contacts faithfully open and close when coil is energized and deenergized. So my evap. sys. vent control malfuntion (P0446) must be elsewhere. At this point, it is looking like the problem is in the ECU itself! Very frustrating. Thanks again, Mike V.

Like Kimo said, it's an easy job and basic tools are all thats needed. A little manual dexterity will help. I just bought a genuine Subaru knock sensor from Liberty Subaru for about 75bucks. Probably did not need it. Probably just another CEL/OBDII B.S. situation... Mike V

Enjoy it while you can. Your days are numbered, I am sure. This CEL scam is too good a money maker. Is your CEL on????

No, I don't believe I am looking for a fan relay. Looking for the one that controls power to the Evap. sys. vent/purge control valve. Any other suggestions? Thanks, Mike V

UPDATE: I did the Step 1 check on the P0446 troubleshooting procedure from the factory manual (thanks to Ferret) and I get battery voltage at the appropriate pin (ECU connected, measuring probe inserted from rear of connector) when I turn the ignition on. HOWEVER, when I turn ignition OFF, battery voltage REMAINS at this pin for several minutes and then drops out. I was able to repeat this test several times. I believe that the power should drop out IMMEDIATELY upon turning ignition off, right??? So now I am suspecting a problem with the "Main Relay" that controls the power to this circuit. Trouble is, my Haynes manual does not tell me where this relay is. Can anyone help? Ferret? Nipper? Anyone? Very frustrated but at least I have found what seems to be an anomaly. Thanks, Mike V.

On a 1997 Legacy 2.2. I only have a Haynes manual and I do not see where they tell you exactly which relay is the "Main" relay. I am ASSUMING it is in the fuse/relay box under the hood on the driver's side, but even that may be wrong. Can someone pin point this for me please? Thanks! Mike V

OK. I will interpret the Step 1 instructions as saying to disconnect the ECU, turn the ignition on and measure the voltage between the Purge valve pin on the ECU connector and ground. Correct? This is not what I would call the "OUTPUT SIGNAL FROM THE ECM" in the instructions. Not sure why they would call it that. As far as "living with the CEL", I could not give a darn about it. This is ALL about passing state inspection and if the CEL light is on for ANY reason, I do not get a sticker and risk a $100.00 fine every time the car leaves the driveway. Thanks again, Mike V

Well, if I were buying I would buy a 1995 or older one due to the OBDII related state inspection system we have in PA (as, I believe, in several other states). Seriously. Mike V.

Ian, I am fighting a similar issue but my DTC is P0446. You may want to read the cuurent thread under that name. Please update us on what you do/find! I will do the same. Good luck! Mike V

Thanks much Ferret! I quickly read through the procedure and (unfortuneately) I believe I have already accomplished most of what they describe. You said "I'll bet you have a high resistance from the drain valve return back to the ECU" but I did measure the resistance from the drain valve plug connector to the ECU plug connector and it was only 0.7 ohms. I believe step 10AJ7 in the procedure says this needs to be below 1.0 ohms so (again unfortuneately) that looks ok. I am not sure I fully understand your comment "This fault states the ECU does NOT see that 12 volts before it 'energizes' ( provides a ground via a semiconductor ) so it sets the CEL.". Are you saying that if the resistance in the circuit between the drain valve and the ECU were too high there would be too much voltatge drop there and the ECU would then see too low a voltage? I will consider the procedure more carefully and perform any checks that I have not already done. But I am seriously afraid I will get to the very last step on page 193: "CONTACT WITH SOA"!!! Don't think I could afford that.... Again, thanks very much for your comments and consideration! Mike V

Thanks for reading my post Nipper. No, I do not smell any gas fumes at all. Actually, I wonder if the computer actually is monitoring the anything other than the electrical circuit. Do you think it is somehow monitoring the gas vapor side? Also, does anyone know under what conditions the computer tells the valve to open? I beleive my testing ruled out a loose connection except at the ECU connector plug itself but the pins there looked clean. Please, any suggetsions? I need to fix this in order to pass inspection! Thanks again! Mike V

OK, here'e the deal: 1997 2.2 Legacy CEL on. Fault code 446("Evaporative Emmission Contrl System Vent Malfunction"). This same code had come up about 2 years ago and I replaced the vent solenoid valve at that time. I do NOT know if that actually fixed this fault at that time because my CEL light stayed on but I probably had other "faults" which would have kept it lit. I did not have access to a scan tool and did not take it to Autozone to get the codes scanned at that time. So this week I got hold of a scan tool. Read the faults. There were four lsited and P0446 was one of them. Cleared codes. CEL off for first time in years!! Let idle for a few minutes and CEL stayed off. Shut down and restarted. CEL came on in a few seconds. Scan codes. Only one - P0446. Disconnected solenoid vent valve and measred resistance across its terminals: 26.4 ohms. Haynes manual says this should be anywher from 10 to 100 ohms so it appears (from that test) that the sol. valve is ok. Measured voltage at power side of solenoid valve connector: 13 volts when key is turned on. Did a point by point (tedious) check of the wiring between the "downstream" side of the solenoiod vent valve connector (White with blue stripe) and the ECU. Everything checked out fine. Total resistance between valve connector plug and pin in computer connector plug is 0.7 ohms. That sounds ok, right? Removed solenoid valve from vehicle and connected it to a 12 volt battery. Solenoid "pulled-in" just fine. Nice positive click. Tested by mouth and valve is closed when de-energized and open when energized. Exercised valve by connecting and disconnecting power about 20 times. Seemed fine. Reinstalled solenoid vent valve. Made sure all connections were secure. Cooencted scan tool and cleared codes. Started engine and CEL did not come on. Let it run for a while and fault code P0446 popped up in about 5 minutes!! Hair is getting pulled out. What could possibly be happening? Is the vent valve truly an ON-OFF deal or does the computer proportionally open the valve up? Sorry for the long post but I wanted to be thorough (like I was in my check-out procedure). Someone please!!! Thanks, Mike V