The Dreaded Misfire Saga Continues

[subsince77]

Ok, got it. Idles at just a hair under 15, drops a little when you push the gas, then recovers. Increases a bit when you let off then recovers. Hangs right at 15 when holding the throttle steady. I have it hooked up and mounted on edge under the wiper. It seems pretty solid. I think it will ride there on the highway. So, tomorrow when I get time I'll run up the valley and see if I can get it to misfire. MisFire Test # 100094586799847403284895090 is under way.

[OB99W]

Ok, got it. Idles at just a hair under 15, drops a little when you push the gas, then recovers. Increases a bit when you let off then recovers. Hangs right at 15 when holding the throttle steady.[...]

From your description, the responses to throttle seem about right. I'd normally expect 20-22 inches with a warmed engine in good condition at sea level, but since about one inch is "lost" per 1,000 feet of elevation, the 15" reading is probably okay for where you live.

[subsince77]

From your description, the responses to throttle seem about right. I'd nomally expect 20-22 inches with a warmed engine in good condition at sea level, but since about one inch is "lost" per 1,000 feet of elevation, the 15" reading is probably okay for where you live.

 

It was pretty much cold too, had been sitting for 2.5 hours after running 15 minute, after sitting all day. I'll get some more accurate readings tomorrow. But, yeah, my house is at 8050'.

[subsince77]

You guys are geniuses. It turns out that merely hooking up a vacuum gauge cures the whole misfire problem instantly. I can't get this thing to misfire for nothin'! I'll keep trying. Tomorrow has to be the day.

[McDave]

You guys are geniuses.

Thank you. Thank you very much.

[subsince77]

Thank you. Thank you very much.

Now, as long as i never need my windshield wipers...

[OB99W]

You guys are geniuses.

Of course we are.

 

 

It turns out that merely hooking up a vacuum gauge cures the whole misfire problem instantly. I can't get this thing to misfire for nothin'! [...]

Watched pots never boil (although at 8050 feet, a lower temperature by several degrees will do when you aren't watching).

 

See http://en.wikipedia.org/wiki/Observer_effect_%28physics%29

[frag]

If someone somewhat intellectually challenged can be permitted to join, for a moment, this learned assembly of geniuses, I would propose that if your misfire never returns while the vac gauge is connected, maybe the problem has to do with the vaccuum connexion you used to connect the gauge.

A little too obvious maybe, but I'm trying not to exceed my range.

[OB99W]

If someone somewhat intellectually challenged can be permitted to join, for a moment, this learned assembly of geniuses,

Self-deprecation doesn't suit you.

 

 

I would propose that if your misfire never returns while the vac gauge is connected, maybe the problem has to do with the vaccuum connexion you used to connect the gauge.

A little too obvious maybe, but I'm trying not to exceed my range.

If the misfire never occurs while the gauge is attached, the problem might have something to do with the particular port or hose, or it could be that something else was disturbed in the process of connecting the gauge. However, if the misfire should return when the vacuum hose is reconnected as it was originally, we have something more specific to consider. Personally, I'm waiting for additional data from subsince77 before coming to any conclusions.

[frag]

Self-deprecation doesn't suit you.

 

 

If you say so. But I hope humour suits us both.

[OB99W]

If you say so. But I hope humour suits us both.

I hope so as well -- the "wink" was certainly meant to convey that my comment wasn't serious.

I am serious when I offer suggestions or opinions concerning Subaru problems, however.

[subsince77]

I had the same thought about whether the connection of the gauge could have changed something. Only two things to report. First, I drove for about 30 minutes on the highway, nothing, then came into town and parked. Drove away after about 10 minutes, nothing. Had to go through the one slightly congested area in downtown where there happened to be several tourists and a couple of bicycles crossing the street at the same time. I took my eyes off for half a minute, looked back down and the light was on - I MISSED IT! Didn't someone say a watched pot... Since then unable to duplicate problem, then had to drive our pickup for the last two days. So, back at it tomorrow.

[McDave]

I had the same thought about whether the connection of the gauge could have changed something. Only two things to report. First, I drove for about 30 minutes on the highway, nothing, then came into town and parked. Drove away after about 10 minutes, nothing. Had to go through the one slightly congested area in downtown where there happened to be several tourists and a couple of bicycles crossing the street at the same time. I took my eyes off for half a minute, looked back down and the light was on - I MISSED IT! Didn't someone say a watched pot... Since then unable to duplicate problem, then had to drive our pickup for the last two days. So, back at it tomorrow.

My thinking is if you had a valvetrain problem, then it would have shown up sooner. Even though it didn't trigger the cel at the time, you probably would have noticed the gauge acting up before since you were keeping a close eye on it.

 

At this point I would take a very close look at all the electrical connections between the ecu and it's engine sensors (and injectors). Check for a pin that has pushed out the back of a connector, or a corroded connection, or one that has gotten hot and melted the plastic connector, or a bare wire that is touching metal when you move things around (Ford calls it a wiggle test, engine running).

 

Then consider changing the crank and cam sensors.

[OB99W]

I had the same thought about whether the connection of the gauge could have changed something. Only two things to report. First, I drove for about 30 minutes on the highway, nothing, then came into town and parked. Drove away after about 10 minutes, nothing.

If the gauge has not been indicating anything "suspicious" when you have been able to watch it, it's improbable that anything of significance is related to what CCR is responsible for. It's likely their rebuilt engine is just fine, mounting of the intake went okay, etc.

 

 

Had to go through the one slightly congested area in downtown where there happened to be several tourists and a couple of bicycles crossing the street at the same time. I took my eyes off for half a minute, looked back down and the light was on - I MISSED IT!

I wonder if this was another time of deceleration (as you were responding to those crossing the street); you had previously reported you thought that might be a condition that triggered the problem.

 

 

Didn't someone say a watched pot... Since then unable to duplicate problem, then had to drive our pickup for the last two days. So, back at it tomorrow.

Yes, I joked about watched pots, etc.

 

I'm still thinking about the possibility of EGR problems. Although it's not a definitive test, sometimes temporarily disconnecting and plugging the vacuum line at the EGR valve is enough to diagnose the possibility. Not all Subaru engines have EGR, but assuming yours does, the valve sticks out almost horizontally to the rear of the intake manifold cylinder #4 runner.

 

If you would humor me, try disconnecting the vacuum hose at the EGR valve itself, and plug the hose end with a nail or something similar that will provide a good seal. Then drive as usual and see if the misfire code(s) appear again. (It would still be possible to have a problem if the valve never seals properly due to carbon buildup, but might be telling if the valve is just sticking.) If you get a CEL while the EGR is disabled, please read the code(s) and report what you've gotten, since it might be other than the misfire ones under these conditions.

[OB99W]

[...]At this point I would take a very close look at all the electrical connections between the ecu and it's engine sensors (and injectors). Check for a pin that has pushed out the back of a connector, or a corroded connection, or one that has gotten hot and melted the plastic connector, or a bare wire that is touching metal when you move things around (Ford calls it a wiggle test, engine running).

 

Then consider changing the crank and cam sensors.

While what McDave is suggesting is reasonable, it was my understanding that cam and crank sensors, etc., came with the "new" engine, and were not transfered from the "old" one. If that's the case, it would be a significant coincidence if the problem were the same with both the old and current engine.

 

It's also my understanding that the intake manifold and what's mounted to it are primarily what got transferred from the old engine, and that's why I'm concentrating on things related to that. If I'm incorrect about what was reused, please clarify that.

[McDave]

While what McDave is suggesting is reasonable, it was my understanding that cam and crank sensors, etc., came with the "new" engine, and were not transfered from the "old" one. If that's the case, it would be a significant coincidence if the problem were the same with both the old and current engine.

 

It's also my understanding that the intake manifold and what's mounted to it are primarily what got transferred from the old engine, and that's why I'm concentrating on things related to that. If I'm incorrect about what was reused, please clarify that.

Good point. I had assumed the new motor didn't come with the C&C sensors and you had reused your old ones.

 

PS I like the egr theory too. Note, you may want to fill your tank with premium so the knock sensor doesn't retard the timing so much with the egr vacuum plugged off.

[Steves72]

Please do not read any personal diagnostic abilities into this post. I just wanted to relate a story that happened to me about 6 years ago.

 

I had a 91 Chevrolet Camaro Z28. Almost every time I drove it on the highway the check engine light would come on. It would only come on with a sustained speed of at least 50 mph that lasted at least 5 minutes. Sometimes it would take a half hour or more of sustained highway speeds to get the check engine light to come on. Furthermore, the computer (this was pre OBDII) did not have a code stored after shutting the car off. This goes back to the day of blinking out the codes in the computer but you had to start the process with the car off. The check engine would always be off after restarting the car and no codes were ever shown through the "blinking" method. The engine always ran perfectly. There was no reason to suspect anything was wrong with the car at all other than the light on the dash.

 

I talked to my mechanic who asked me to get the light to come on; drive to his station and not shut the car off. With the car still running he was able to retrieve the soft code in the computer. The code was related to the EGR and a quick check found a very small tear in the vacuum line that went to the EGR. In two minutes the hose was replaced and that was the last I ever saw of the check engine light.

 

In short, if you have an EGR check it and the vacuum line connected to it.

 

Steve

[subsince77]

Thanks everyone, sorry for the silence. We have suddenly entered into an unexpected, extremely busy couple of weeks. Extended family health issues, a death in a family very close to us, and some planned special events at work. We were supposed to be leaving for Albuquerque this morning. I had thought it worth leaving the vacuum gauge connected because there is no doubt that somewhere on the road the light will come on, and probably a fairly short distance out of town.

 

Now that trip has been moved to next week. I probably will not have time to take the car for any longer drives between now and then with everything we have going on right now. It may be a few days before I can continue all of this, but I am not dropping it.

 

Now for specifics. On a CCR rebuild, my understanding is that they transfer the intake manifold, injector rails, all the externals like started, alternator etc. I would guess pulleys, I had them put on new belts. But as for sensors, I don't know. Emily, if you're following this, maybe you could chime in there.

 

As to whether I've noticed anything unusual while watching the V gauge, I was never sure what I was supposed to see. It idles at about 15 or 16 when warm. On a straight run at 65 or so it will hold perfectly steady at around 10 to 12 (depending on up valley or down - there is no real level here). On a long steep hill when I am pushing it it will drop way down to under 5, then recover instantly with the slightest decrease in pedal pressure. On a long downhill or fast decel, it goes up to about 20. It does all of that seemingly without any variation, so although it is moving around a lot it seems very predictable. It is nearly impossible to find a stretch of road around here where you can just hold the gas perfectly steady for very long. Too many hills and curves.

 

The EGR problem description above certainly fits the symptoms. Just from engine compartment memory, I'm sure this car has an EGR sitting right in the location described. I can check the hose pretty easily. In fact, would hooking the vacuum gauge up to it and seeing if the vacuum is steady help? Otherwise, I can just plug it and run it. I just filled it with regular, so it would be a while before I could change the gas. One note there, remember my elevation. Even Subaru does not say i need premium at 8000'. I have used premium for 6 moths at a time, just to see if there was any difference in this problem, milage, power, anything - there was none. Just thought I'd head that one off. I understand why McDave was suggesting it for this experiment, I am just addressing an inevitable suggestion that regular gas is causing the misfire in general. i don't think so.

 

Sorry this is os long. I'll get on this asap. if you don't hear from me for a few days, I'll be back.

[OB99W]

Sorry to read that you're having issues beyond the car itself.

 

[...]Now for specifics. On a CCR rebuild, my understanding is that they transfer the intake manifold, injector rails, all the externals like started, alternator etc. I would guess pulleys, I had them put on new belts. But as for sensors, I don't know. Emily, if you're following this, maybe you could chime in there.[...]

Yes, it would be good to know if the cam and crank sensors got transfered.

 

Although the vacuum gauge readings that you're reporting would be low in my area, at your elevation they're probably about right.

 

 

The EGR problem description above certainly fits the symptoms. Just from engine compartment memory, I'm sure this car has an EGR sitting right in the location described. I can check the hose pretty easily. In fact, would hooking the vacuum gauge up to it and seeing if the vacuum is steady help? Otherwise, I can just plug it and run it.

Vacuum to the EGR valve is ported/controlled, so it shouldn't be steady. When you get the chance, pull the hose at the valve, plug it, and see what happens. Don't run the car that way any longer than is necessary to determine if the misfiring situation changes.

 

 

I just filled it with regular, so it would be a while before I could change the gas. [...]I understand why McDave was suggesting it for this experiment[...]

Just in case anyone reading this doesn't understand why McDave suggested premium gas while running this "experiment" -- exhaust injected into the intake when the EGR valve is open cools things a bit, and regular grade would be sufficient. When the EGR is disabled, a higher octane might be required to prevent knocking that could cause the ECU to retard timing.

[subsince77]

Don't run the car that way any longer than is necessary to determine if the misfiring situation changes.

So, how long would be safe? I understand you to say, unhook, plug the hose, drive as normal until the light comes on, then read and report. How much driving around without the EGR is OK?

 

Sorry for the "experiment" terminology. I meant Diagnostic Procedure.

[OB99W]

So, how long would be safe? I understand you to say, unhook, plug the hose, drive as normal until the light comes on, then read and report. How much driving around without the EGR is OK?

If you find the engine performance to be significantly affected, quit at that point, whether or not anything has been determined by then; it could mean that there's knock (which you might not even hear). On the other hand, as long as the car seems to be running okay, use it with EGR disabled until you either get a CEL and read codes, or you've run it longer than it ever took for the misfire codes to be set.

 

 

Sorry for the "experiment" terminology. I meant Diagnostic Procedure.

It is an experiment, but "diagnostic procedure" is certainly a more professional way of describing it.

[hankosolder2]

One thing with the EGR experiment; the experiment itself will set a CEL! The computer monitors EGR flow (by looking for a change in manifold vaccum when the EGR valve is open) and will set an EGR flow code.

 

Nathan

[McDave]

One thing with the EGR experiment; the experiment itself will set a CEL! The computer monitors EGR flow (by looking for a change in manifold vaccum when the EGR valve is open) and will set an EGR flow code.

 

Nathan

Doh! Why didn't I think of that? How will you be able to tell when it sets the misfire codes if the cel is already on?

[OB99W]

One thing with the EGR experiment; the experiment itself will set a CEL! The computer monitors EGR flow (by looking for a change in manifold vaccum when the EGR valve is open) and will set an EGR flow code.

That's true, which is why the last sentence of my post #89 of this thread says "If you get a CEL while the EGR is disabled, please read the code(s) and report what you've gotten, since it might be other than the misfire ones under these conditions."

 

 

Doh! Why didn't I think of that? How will you be able to tell when it sets the misfire codes if the cel is already on?

By reading all the codes after running for a while -- more than one can be stored.

 

 

EGR codes could be ignored during the "diagnostic procedure", but I still would like to know of any code(s) being set.

[subsince77]

Just a quick update. I have had no time to work on the car since last report. All is well, but we have been way too busy. We had an out of town trip to help ailing relatives planned for this week so I was going to work with both the vacuum gauge and the EGR to try to get some more information to you guys. But, because we have fallen so far behind, my wife took the car on the road trip, and I am staying home to catch up on things.

 

The cel had never come back on until just yesterday, right after a fill up - it does not always do that, and I have already replaced the gas cap. I didn't have time to check it, so I told her, "I got the cel on for you, now just ignore it."

 

I hope to get back on this right after labor day. Sorry to ruin our momentum.