Checking out the ECU isn't a bad idea. While I haven't seen a problem with Subaru ECUs, I've definitely seen leaky caps on Toyota and Mitsubishi ECUs.
I can't read that DOC file, but there might be something to this bad ignition relay idea. The ignition relay has two sets of contacts. The second set of contacts normally powers only four things: the purge solenoid, the IAC valve, the pressure exchange solenoid, and the boost control solenoid. If that document was written with non-turbos in mind, it would obviously only mention the first two. But I would imagine if the ignition relay were causing your problems, you'd get codes 44 and 45 as well.
Nice detective work, though, log1call!
If you do want to take a look anyway, there's a fifth thing that set of contacts powers -- the Select Monitor. You can check to see if that circuit is getting power by looking for +12v on the yellow-with-red-stripe wire going to the Select Monitor connector. Or even try supplying power via that wire.
At a glance, those PDFs don't seem to be about the EJ22T at all. "04-dtc criteria" seems to be about EJ257 models, and Diagnostics seems to be about EJ20G models. Be careful about generalizing.
Running good only in D-Check mode - What's going on!?!
Moderators: Helpinators, Moderators
OBD2 is OBD2. The other document is generic for the older subaru OBD, they don't differentiate between the makes of ecu.
It's hard not to generalise. Until I see something that leads me to believe that American models, 97 2.2T engines or hitachi ecus are somehow special I will continue to approch them the same way I always have... applying the basic principles of internal combustion, to the equally basic principles of OBD.
If you have some documentation(or evidence from your work interperting the code), that suggests that this particular car/ecu has some exclusive system in place, then please tell me where I can have a read of it.
It's hard not to generalise. Until I see something that leads me to believe that American models, 97 2.2T engines or hitachi ecus are somehow special I will continue to approch them the same way I always have... applying the basic principles of internal combustion, to the equally basic principles of OBD.
If you have some documentation(or evidence from your work interperting the code), that suggests that this particular car/ecu has some exclusive system in place, then please tell me where I can have a read of it.
Interesting they mention codes 24 and 35 for a faulty relay...
In the midsts of my swap, I've had an issue with the self shutoff wire not working correctly- when wired in with the diode in place, the car wouldn't shut off. It continued to run even with the key out, until power to that pin was cut. I was unable to solve the problem after looking at the wire maps between the two cars.
As my "solution" I had to wire in a switch to cut power to that pin on the main relay - is it possible that over time I've damaged the relay? I have a few spares I could test with. Kind of surprising.
I wish there were more documentation on D-check mode. When driving, the CEL would flash code 24 intermittantly, and code 35 constantly. Such as that I'd see 35 maybe 3 times in a row, followed by 24.
I pulled out my engine harness and everything looks fine. I cringe at the thought of having to strip my harness down to bare wires and rechecking everything again. At that point I should just consider doing another swap with newer electronics - this is driving me crazy!
Thanks again to both of you for your input - I think there's a lot to learn here!
In the midsts of my swap, I've had an issue with the self shutoff wire not working correctly- when wired in with the diode in place, the car wouldn't shut off. It continued to run even with the key out, until power to that pin was cut. I was unable to solve the problem after looking at the wire maps between the two cars.
As my "solution" I had to wire in a switch to cut power to that pin on the main relay - is it possible that over time I've damaged the relay? I have a few spares I could test with. Kind of surprising.
I wish there were more documentation on D-check mode. When driving, the CEL would flash code 24 intermittantly, and code 35 constantly. Such as that I'd see 35 maybe 3 times in a row, followed by 24.
I pulled out my engine harness and everything looks fine. I cringe at the thought of having to strip my harness down to bare wires and rechecking everything again. At that point I should just consider doing another swap with newer electronics - this is driving me crazy!
Thanks again to both of you for your input - I think there's a lot to learn here!
SUBARUEHS Racing
-
vrg3
- Vikash
- Posts: 12517
- Joined: Sun Nov 03, 2002 2:13 am
- Location: USA, OH, Cleveland (sometimes visiting DC though)
- Contact:
Hmm, it sounds like you didn't have the diode going to the right wire. You basically want to replicate what the stock Legacy wiring has. You can see that diode in the FSM wiring diagram; the anode goes to ignition-switched power and the cathode goes to the not-grounded end of the ignition relay's coil. That end of the relay coil is also connected to the self-shutoff control pin of the ECU.
In any case, no, you didn't damage the relay by putting that switch in. The relay does wear out with time though; the contacts develop pits as little bits of arcing current cause damage. From what I hear, pre-Legacy Subarus had ignition relays with platinum contacts which were more resistant to that kind of damage.
It couldn't hurt to try another relay if you have spares. But, like I said, if the relay causes trouble codes for the IAC valve and CPC solenoid, it should also cause trouble codes for the boost control solenoid and for pressure sensing.
Hmm. You know what? I may be reaching here, but I wonder if clearing codes requires self-shutoff control to be working correctly. What if the ECU clears the trouble code information only after the engine shuts off? I do know that when you use the Select Monitor (or my standalone scantool) to clear codes you have to shut the engine off and wait for self-shutoff to complete before you can power the ECU up again.
Huh. Wanna give the diode another try?
In any case, no, you didn't damage the relay by putting that switch in. The relay does wear out with time though; the contacts develop pits as little bits of arcing current cause damage. From what I hear, pre-Legacy Subarus had ignition relays with platinum contacts which were more resistant to that kind of damage.
It couldn't hurt to try another relay if you have spares. But, like I said, if the relay causes trouble codes for the IAC valve and CPC solenoid, it should also cause trouble codes for the boost control solenoid and for pressure sensing.
Hmm. You know what? I may be reaching here, but I wonder if clearing codes requires self-shutoff control to be working correctly. What if the ECU clears the trouble code information only after the engine shuts off? I do know that when you use the Select Monitor (or my standalone scantool) to clear codes you have to shut the engine off and wait for self-shutoff to complete before you can power the ECU up again.
Huh. Wanna give the diode another try?
"Just reading vrg3's convoluted, information-packed posts made me feel better all over again." -- subyluvr2212
Not sure if it applies to this particular model but.... in some models at least,(well all as far as I know), that self cut you are talking about is only meant to work if the engine is cold.
Not sure if it will be related to your other trouble but it's important to recognise all symptoms to make a reasoned diagnosis.
Not sure if it will be related to your other trouble but it's important to recognise all symptoms to make a reasoned diagnosis.
I do know that codes have cleared on this car - I used to have TPS issues, and once I corrected them, the code for that went away.vrg3 wrote:Huh. Wanna give the diode another try?
I fought long and hard with the Self Shutoff wire and posted my battle with it. I honestly don't know what else I could do to it to make it work correctly. I've got copies of my 95FSM I ordered from the dealer if you'd like to take a look.
http://bbs.legacycentral.org/viewtopic. ... 473#288473
Maybe I should leave the battery disconnected for a while (10+ hours) and that will help unless the codes get stored in some NVRAM of some kind...
SUBARUEHS Racing
Ok. At the risk of making things confusing(and because the other posts are old), I'll comment on your other post about the diode issue here.
First off, I don't know anything about that diode being fitted to the car but I know a bit about diodes in general. The diode will not be to let an extra two volts through to the main relay as you suggested... diodes don't work like that. If the diode is going to let power through it will let the full twelve volts through(well all but point six of a volt anyway).
Second thought... The ecu has some power to one or two pins all the time, when you turn on the key it feeds another pin with power which tells the ecu to trun on the main relay, which it does by feeding out power. In your car, when you turn off the key the ecu will check engine temperature and if it's not above a certain temperature(don't ask my why a low temperature, I'd have thought if it was a hight temp but that's not what the book says), it keeps feeding the power to the main relay, keeping it running for a minute.
As I understand it you have put a diode between the ignition switch and the wire coming from the ecu, Right?
Ok, so if the diode is one way it will feed full power straight to the main relay as soon as you turn the ignition on. This over-rides the ecu's control but won't matter as when you turn off the key the ecu will continue to feed power to the main relay if it needs to, but... the diode won't really be doing anything useful.
If you connect the diode the other way around(which is what I think you must have done), the diode won't pass power through to the main relay from the key, but it will allow the ecu's power feed to the main relay to be fed to the ignition side of the key. Feeding power out of the ecu, through the diode, back to the ignition(which was what turned the ecu on to start with and which told the ecu to feed the main relay power), will operate as a "latch" and once switched on won't turn off.
So, one way the diode is serving no usefull purpose, and the other way it is causing a problem.
I have read a lot of subaru wiring diagrams but I don't ever remember seeing this diode. Do you have a diagram I can have a look at? Perhaps I will be abe to figure something out if I know which car it is fitted to and have a look at that cars manual and diagram.
First off, I don't know anything about that diode being fitted to the car but I know a bit about diodes in general. The diode will not be to let an extra two volts through to the main relay as you suggested... diodes don't work like that. If the diode is going to let power through it will let the full twelve volts through(well all but point six of a volt anyway).
Second thought... The ecu has some power to one or two pins all the time, when you turn on the key it feeds another pin with power which tells the ecu to trun on the main relay, which it does by feeding out power. In your car, when you turn off the key the ecu will check engine temperature and if it's not above a certain temperature(don't ask my why a low temperature, I'd have thought if it was a hight temp but that's not what the book says), it keeps feeding the power to the main relay, keeping it running for a minute.
As I understand it you have put a diode between the ignition switch and the wire coming from the ecu, Right?
Ok, so if the diode is one way it will feed full power straight to the main relay as soon as you turn the ignition on. This over-rides the ecu's control but won't matter as when you turn off the key the ecu will continue to feed power to the main relay if it needs to, but... the diode won't really be doing anything useful.
If you connect the diode the other way around(which is what I think you must have done), the diode won't pass power through to the main relay from the key, but it will allow the ecu's power feed to the main relay to be fed to the ignition side of the key. Feeding power out of the ecu, through the diode, back to the ignition(which was what turned the ecu on to start with and which told the ecu to feed the main relay power), will operate as a "latch" and once switched on won't turn off.
So, one way the diode is serving no usefull purpose, and the other way it is causing a problem.
I have read a lot of subaru wiring diagrams but I don't ever remember seeing this diode. Do you have a diagram I can have a look at? Perhaps I will be abe to figure something out if I know which car it is fitted to and have a look at that cars manual and diagram.
-
vrg3
- Vikash
- Posts: 12517
- Joined: Sun Nov 03, 2002 2:13 am
- Location: USA, OH, Cleveland (sometimes visiting DC though)
- Contact:
If you have successfully cleared codes before, then I don't know what to say, except that maybe we should still just try to solve whatever electrical gremlins we can and hope that makes the other ones easier to solve.
I'll address the diode problem in the other thread.
I'll address the diode problem in the other thread.
"Just reading vrg3's convoluted, information-packed posts made me feel better all over again." -- subyluvr2212
FIXED!
So Friday I mentioned that I pulled the harness and all looked well. After I put it back together Saturday nite, the car wouldn't start. It cranked and cranked, but no good. I went to my Cam sensor.
It was loose! The plug hasn't survived the test of time very well, and while I did the best to repair it during my swap, it's apparently lost it's ability to stay completely fastened.
Anyways- I got the plug tightened up and it fired. Hesitation was gone! So apparently the cam sensor had enough signal to start, but not to run right. Once I wiggled it completely loose, all bets were off. Once I fixed it, it's running perfect, and both my codes have gone away! I drive down the road with 0 codes.
D -check must set to a default timing map, which allowed me to run without problems. Still surprised to see those two codes present when I had issues with my cam sensor. Completely unrelated? Thanks for all the help. I was able to take the car to the auto-x without any issues.
I owe you guys big time for being so helpful and patient!
So Friday I mentioned that I pulled the harness and all looked well. After I put it back together Saturday nite, the car wouldn't start. It cranked and cranked, but no good. I went to my Cam sensor.
It was loose! The plug hasn't survived the test of time very well, and while I did the best to repair it during my swap, it's apparently lost it's ability to stay completely fastened.
Anyways- I got the plug tightened up and it fired. Hesitation was gone! So apparently the cam sensor had enough signal to start, but not to run right. Once I wiggled it completely loose, all bets were off. Once I fixed it, it's running perfect, and both my codes have gone away! I drive down the road with 0 codes.
D -check must set to a default timing map, which allowed me to run without problems. Still surprised to see those two codes present when I had issues with my cam sensor. Completely unrelated? Thanks for all the help. I was able to take the car to the auto-x without any issues.
I owe you guys big time for being so helpful and patient!
SUBARUEHS Racing
Ha, good old mechanics stand-by.... keep driving it, if it gets worse we will be able to see what it is.
That's good to hear that it's fixed. Now If I could just figure what's with those wiring diagrams!!!
Vrg3, not trying to be smart, or pull rank, but....
I have always been under the understanding that a sensor's reading being out of range could cause codes to be set, well some codes anyway. It is pretty well universally accepted as far as I know amongst mechanics that it can happen. Training manuals say it can happen. Old generic diagnostic books I have going back fifteen years and more say it can happen. I have seen it happen.
If you look carefully at the criteria pdfs I think it implies it without saying directly that it is so. There are sensor code checks that say check vacuum hoses, check cam timing... to name a couple of common ones. Not all the codes are caused by electrical faults.
I'm going to have to get my hands on a turbo now and check this D mode throttle cut thing. There is alays something new to learn.
Anyway, interesting case and good work on finding the problem.
That's good to hear that it's fixed. Now If I could just figure what's with those wiring diagrams!!!
Vrg3, not trying to be smart, or pull rank, but....
I have always been under the understanding that a sensor's reading being out of range could cause codes to be set, well some codes anyway. It is pretty well universally accepted as far as I know amongst mechanics that it can happen. Training manuals say it can happen. Old generic diagnostic books I have going back fifteen years and more say it can happen. I have seen it happen.
If you look carefully at the criteria pdfs I think it implies it without saying directly that it is so. There are sensor code checks that say check vacuum hoses, check cam timing... to name a couple of common ones. Not all the codes are caused by electrical faults.
I'm going to have to get my hands on a turbo now and check this D mode throttle cut thing. There is alays something new to learn.
Anyway, interesting case and good work on finding the problem.