The Mysterious #3 & 4 mis-fire solved - Maybe

[subsince77]

This will take a few paragraphs, so get comfortable.

 

OK, I have posted a few times about the problems I have had with a persistent CEL that is always a mis-fire on both cylinders 3 and 4 at the same time. This has been going on for years, but in the last 18 months or so, it got a lot worse. It had gotten to the point that any time I drove the car at highway speed for more than 45 minutes to an hour, It would always throw this code. I never felt anything, or heard anything, the car ran perfectly except for throwing this code. I have read many posts by others with the same complaint.

 

I have replaced the plugs and wires a few times, always with the correct OEM stuff. I finally replaced the coil pack last fall - still got the stinkin' code. I have done the Seafoam white smoke clean, replaced fuel filters, air filters - anything suggested - still got the stinkin' code. I finally decided it must either be an injector problem, or a valve hanging, but why would either of those always happen on two opposite cylinders at the exact same moment every time?

 

Then came the HG problem and this code was the least of my worries. I couldn't run on the highway anyway. I secretly hoped that with the new engine the code would go away, but no, just shy of one hour up the hill from Denver, with a brand new engine (of course the plugs, wires, injectors etc. were all my old ones) on comes the light - #3 and #4 mis-fire:mad:

 

I pulled over and called CCR. After a bit of deliberation, and discussion about the various causes of this problem, I elected to head on home instead of going back down into Denver to have them look at things. I was sure this was some gremlin in my equipment, not a problem with their rebuild.

 

The next day when I got home from work, there was a message on my machine from Rick Johnson, the owner of CCR. He said something like, " I know you will never believe this, but if you will replace the neutral safety switch on that car, it will stop throwing that code." He said, "I know it doesn't sound right, but I'm telling you, on that car, with the manual transmission, that switch will cause that code.":eek:

 

The switch is only $30, so I figured it was worth a try. It screws into the left side of the transaxle, about 6 inches behind the back up light switch. CCR offered to get me one, which was going above and beyond as far as I am concerned. So they did, and it arrived yesterday. I put it in last night, and today I drove for 2 1/2 hours on the highway (1 1/4 each way) with no CEL. I can't say this is absolutely conclusive, but the car hasn't done that in a long, long, time. In two weeks we'll be doing a 3 hour one way drive, so If I make it through that, I will consider it a sure fix.

 

Again, this is for the 2.5L MT. Mine is a 97 OBW. Not sure if it applies to others. Sorry this was so long, but I found this kind of fascinating. Maybe it's not new info. But I hadn't read it before.

[Skip]

I thank you and Rick very kindly for this

post.

 

It should be some how saved in the archives

 

Thanks again. You,

Rick, Emily and

CCR rules!!

[davebugs]

I may be chasing the exact same problem on the exact same car. The car doesn't get used much. New NGK plugs, new Subaru wires, coil pack was next.

 

Thanks for the heads up.

 

Dave

[Cougar]

I wonder if the old switch contacts were causing noise and it somehow got into the circuit for the knock sensor.

[screwbaru2]

I thank you and Rick very kindly for this

post.

 

It should be some how saved in the archives

 

Thanks again. You,

Rick, Emily and

CCR rules!!

 

Good thing to add the the USMB repair manual. Thanks for the post I have a 97 OB MT also that was doing the samething.

silly me I was thinging the ECU. I'll but this on my todo list when I get a 2.2.

[EVOthis]

yeah thanks for the info....sure thing ill remember.....

[OB99W]

[...]The next day when I got home from work, there was a message on my machine from Rick Johnson, the owner of CCR. He said something like, " I know you will never believe this, but if you will replace the neutral safety switch on that car, it will stop throwing that code." He said, "I know it doesn't sound right, but I'm telling you, on that car, with the manual transmission, that switch will cause that code.":eek: [...]

The neutral switch (note that I didn't include the word "safety") is monitored by the ECU, since engine run parameters are changed dependent on whether in-gear or not. If the switch becomes intermittent, the ECU can get "confused". Although not directly related to the trouble being discussed, here's something interesting:

http://endwrench.com/images/pdfs/DiagTroubSum04.pdf

[hankosolder2]

Do you guys realize the possible side benefit of the existance of this neutral switch?

 

Remote start on a MT vehicle!

 

You'd probably want to add a relay to isolate the ECU's neutral sensing circuit from the remote starter's circuit, but it should be doable.

 

 

Nathan

[Skip]

Good observation Nathan, you are correct.

 

I have see this switch

used for this purpose for years.

[subsince77]

The neutral switch (note that I didn't include the word "safety") is monitored by the ECU,

I noticed that my manual refers to it either as the Neutral Switch, or Neutral Start Switch. I like Neutral Switch - it's shorter. The reason I wrote Neutral Safety Switch, is because that is what I thought Rick said on the answering machine. Our machine is kind of garbled, so whatever.

 

I want to reiterate a couple of things. First, I am not 100% positive that this fixed the problem - maybe 90%. My confidence has much more to do with the source of the information, than with yesterday's drive. I have thought I had this thing fixed before, only to have it reappear. After a few more longer drives, I will be totally convinced.

 

Also, my car always threw both #3, and #4 together, never just one or the other, and never any other cylinder. What I gathered from Rick was that this switch specifically throws both of those codes when it does this. This isn't about other misfire codes to my knowledge. If that's not right, maybe Emily or Terry, or someone from CCR will chime in.

 

And finally, I never had any problem with the switch as far as doing it's normal job. This car has not started exactly once in the years I've owned it, and that was because I left an interior light on all night. And, it would not start without pushing the clutch in. So I had no hint that something might be wrong with this component.

[johnceggleston]

i want to know how rick johnson of ccr knew this? who did he learn it from? and what else can we learn from him / them?

[ShawnW]

This isnt the switch that has anything to do with the clutch pedal being in or not for starting. Merely a switch that is moved on/off with gear selection.

 

CCR knows this from a Subaru specialist in Denver called Strictly Automotive. Their recommendation along with the dealer stocking the part due to repeated need prompted the replacement suggestion.

[subsince77]

This isnt the switch that has anything to do with the clutch pedal being in or not for starting. Merely a switch that is moved on/off with gear selection.

 

That explains why it wasn't located under the clutch as I first thought. So, for those of us that are ignorant, maybe just me, what does this switch do? What does it tell the ECU?

 

The dealer in Grand Junction had two of these in stock too. I had ordered one before I realized CCR had already sent me one. It must be a commonly needed part.

[ferret]

This is a GOOD ONE to remember. I have seen the neutral switch cause the P1507 ( IAC in Safe Mode ) because if you are coasting in gear with a M/T, and the ECU 'thinks' the tranny is in neutral ( defective switch ), it closes the IAC to lower the idle to the preset ( usually 750 RPM ) and obviously can't. So it thinks the IAC system is not working.

 

But the real culprit is the neutral switch. It's also used to dis-engage the Cruise Control. That's another symptom I've seen. But this is new ........ so noted in my OBD manual notes. Good Call .... THANKS for sharing and Posting.

[cabron]

I have 01 legacy with about 100,000 miles; auto. I have also had the dreaded code. I too changed everything, plugs, wires, etc, etc. What fixed it for me was front and rear o2 sensor.

 

Very frustrating issue.

[subsince77]

Well, I had to make the same trip again today to a neighboring town. Generally, its about 75 minutes each way all highway. But today, due to snowpacked and very icy conditions, it took over 90 minutes one way, and about 80 on the return trip. At the mid point of my trip, while driving around in town, my check engine light blinked a couple of times and came on. Just like the old days.

 

At that point, I figured that this whole neutral switch thing didn't work after all. But when I got home and pulled the code, I discovered that this time there was only one misfire, #4, not the 3 and 4 that I always had previously. So now I'm stumped. I'm not sure if the new switch fixed one problem but I still have another, or if this is just more of the same.

 

My next move is to install the NGK wires that I got. The #4 wire on my OEM set is way too short, and stretches tight to make it from the plug to the coil pack. The new NGK's are 4 different lengths like my original Subaru set that came on the car when it was new. I am pretty sure they will be long enough to fit properly.

 

The only other thing I noticed was that the light came on while I was changing lanes, and right when I ran over a manhole cover. Not a real deep hole, but a bit of a bump. Don't know if it had anything to do with it or not.

[OB99W]

[...]The only other thing I noticed was that the light came on while I was changing lanes, and right when I ran over a manhole cover. Not a real deep hole, but a bit of a bump. Don't know if it had anything to do with it or not.

It's possible that there is a relationship between the bump and the misfire. If the ignition wire is barely fitting, a jolt might cause it to loose connection long enough for the ECU to notice. Hopefully, replacing the plug wires will be the cure. Also check the primary connection to the coil pack. In the same vein, make sure that the connection to #4 fuel injector is tight (and perhaps #3 as well).

[subsince77]

I put in the NGK plug wires today. They definitely fit much better. Each one is exactly the right length for this engine. We'll see if it makes a difference. I will check the connections on the injectors, maybe pull them off and push them on a few times, or shoot a little contact cleaner in there. Our high today was 13, so I didn't want to be pulling on any brittle connectors today.

[Gloyale]

I just got brought a 2000 OB throwing codes for 3+4 mysteriously. It's an Automatic:confused:

 

I will be curious to see the end of this thread

[subsince77]

I will be curious to see the end of this thread

 

Me too. I didn't get 5 blocks with the new wires before I had a check engine light. First #3, then later #4. I put them in fast (didn't remove the washer fluid tank or battery) and it was pretty cold, so I'll check tomorrow to see that everything is seated well.

 

If I have to go back to OEM wires that are too short, I'm going to be frustrated. Maybe I'll try magnacore next and see if they actually make a wire that fits the car. I keep reading that they make a compatible wire for this car. CCR uses these NGKs all the time, so maybe I just didn't get them pressed in well.

[FSRBIKER]

Anyone have a diagram or picture of exactly where the switch is located they can post.

[subsince77]

Anyone have a diagram or picture of exactly where the switch is located they can post.

Hopefully someone will have a picture or diagram. But on my transmission, on the left side, maybe halfway back, there are two obvious switches with wires coming out of them. They are about 6 inches apart. The rear one is the neutral switch. The lead connects on top, and can be easily seen looking down through the engine compartment, right toward the back of the compartment. I had a friend there, so I handed him the lead, and he hooked it up after I screwed the new switch in. But you could do it by yourself pretty easy.

[OB99W]

Since this problem is affecting both cylinders #3 and #4 (paired in the wasted-spark arrangement), it would seem that the problem is ignition-related. While secondary ignition problems are a common cause of misfire (so good plug wires are a starting point), they're not the only possibility -- the ignition path begins at the ECU and goes through the igniter and coil primary before getting to the coil secondary, wires and plugs. A loose connection or faulty part on the primary side could be the culprit.

 

I'm not sure useful info would be captured, but if there's access to a good scan tool, the freeze frame data that set the misfire code might reveal something.

[Cougar]

To see if the coil is causing the problem you could try swapping the coils and see if the problem moves with the coil change. If that doesn't change the problem you might try installing a new knock sensor to see if that is causing this trouble.

[subsince77]

To see if the coil is causing the problem you could try swapping the coils and see if the problem moves with the coil change. If that doesn't change the problem you might try installing a new knock sensor to see if that is causing this trouble.

I already replaced both the coil and the knock sensor. I suppose I could have gotten a bad part, but there was never any change. I have never done anything with the igniter, ECU, or fuel injectors other than cleaning them. This last episode yesterday, happened immediately after the wire change. Prior to that, I can go many weeks with no codes until I take a road trip. So this last thing seems tied to the new wires. Too bad, I really like them. They fit, seem like quality wires, and come in a great, sort of neon blue color.