EA82 Idle Issue (Cyclic Idle)

[jj421]

I am so tired of dealing with my car's idle problem. If I can get it fixed once and for all, I'll be the happiest man alive.

 

Now my car has developed a new idle issue. I still have my other idle issue, where I have to rev it to keep it from stalling. However, this issue has started appearing more and more often over the past few weeks.

 

Basically what happens is the idle goes up, cuts out, goes up, cuts out, goes up, cuts out, etc. Much like "cyclic idles" that you see on tuned cars. It does then when it's supposed to be idling (obviously). If it's idling and the car is moving (like if I'm coasting in neutral to a stop sign), it'll do this cyclic idle. However, once I come to a stop, the idle will return to normal. Very recently, it is now doing it all the time though. Here's a video of what's happening:

 

https://www.youtube.com/watch?v=6xBu3RJIZrk&feature=youtu.be

 

It's like it's trying to rev up to a kinda high idle, but then it get cut off. I don't really know how to explain it. The only CEL code I have is for "air control valve or circuit". I have replaced and cleaned the IACV twice now using junkyard units. Whatever I've done has made absolutely no difference. I've thought about replacing the wiring harness that goes to the IACV, but my testing says I might not need to. I've got power going to the IACV plug and resistances are within tolerances.

 

The only thing I haven't tested is the ECU. Again, my previous idle issue is still there. It kinda feels like if there's water in the intake or something, since I've driven into deep mud before. However, I have sprayed every little and big vacuum hose in the engine bay with carb cleaner and haven't noticed any leaks. Sprayed the intake manifold gaskets, TBI, intake, PCV, and I didn't notice anything. My car has an exhaust leak and a misfire when idling, although it drives fine. Also, I have replaced my CTS twice. Currently have a brand new one in there, and it reads correct resistance at all temperatures.

 

1990 Subaru Loyale, EA82 N/A

 

So, I'd like some help with this. I'm hoping that this new issue could be a possible fix for my other idle issue. My other idle issue seems to be OAT related: when it's warm outside, it'll idle fine. When it's cold outside, it won't want to idle and I have to rev it to keep it from dying. Has nothing to do with engine temperature; just outside air temperature.

 

Any ideas? I can't stand this idle issue. My main issue is just plain dangerous, and this cyclic idle is annoying (and embarrassing, really).

[Subruise]

hows your 02 sensor, Grant? did any wires get messed up when you lifted it?

[jj421]

Nope, not when I lifted it. I've wondered about that as well. I believe at some point, I tried unplugging the O2 sensor with the engine running, both hot and cold. I don't think it made a difference, either way. But I also read that the O2 sensor doesn't do much at idle, so unplugging it shouldn't do anything. I tried revving the engine with it plugged in and unplugged; engine acted about the same on both tests.

 

Looks like an O2 sensor is only like $20 at O'Reilly. Might be worth replacing it anyways, since my current one looks a little worse for wear. Gotta by a socket too, since I don't have a wrench big enough to take it off.

 

I do remember that Marcus had stated that my car smelled lean when I started it cold. I don't have the greatest sense of smell, so I can't really say otherwise. But that could be a thing too.

 

I've also looked into the timing as well. I plugged in the green connectors and put a timing light to the flywheel. Seems to be where it should be. Although my idle problem wasn't as bad with the green connectors plugged in.

Edited March 21, 2014 by jj421

[Subruise]

im thinking unplugging an O2 sensor thats bad wont have any effect. hondas w bad ones sound like that.  also, are you still running  a TB w/o that base spacer? are you leaking there?

[jj421]

That's what I was originally thinking when I did the test. But I looked up online and the general consensus is that a bad O2 sensor won't really affect your idle much if it's bad. Rather, you'd get poor performance and poor gas mileage. My performance isn't that bad, neither is my gas mileage. Still, I don't think it would hurt to change out the sensor, although I would like to know of some tests that I could do beforehand.

 

Yes, I'm still running without the toilet bowl gasket. Not leaking there currently. I'm losing a bit of coolant, but not much. Maybe 0.5 cups of water/coolant every two weeks. Well, maybe not even that much. Not enough to where it'd cause such an issue. Not to mention that when I bought the car, it idled fine. Over time, it developed an issue. It has had this issue before I messed with that gasket, swapped the engine, and even before I swapped the transmission.

 

I'm pretty sure it's gotta be something electrical. It either idles fine or can't idle at all, as if a switch was being flipped between the two. If it were a mechanical problem or a leak, it wouldn't be so, "one way or the other."

 

I've also cleaned the EGR valve quite a bit, but that made no difference whatsoever.

Edited March 21, 2014 by jj421

[jono]

OK, so it is mpfi or spfi, and 1990 so it is not the idle air control valve mounted on the thermostat housing like 1985,86 model mpfi's.

I run both early and late without idle control valves able to contribute air anywhere - just a bit boggy on start up - just like starting an old car with a choke knob, or without, may take a little juggling. Try blocking it out for test purposes. Have you checked the PCV valve ? If the rocker breathers are blocking up the crankcase may be getting a bit too much air in it with an escape out the PCV valve causing a hunting effect.

Try a spark plug I think, blocks the PCV hole for testing ???

[jj421]

It is SPFI. I changed out the PCV valve with a brand new one from the dealership about a year and a half ago. the PCV valve fixed my white smoke on right turns, and haven't seen that problem since.

 

If I unplug or remove the IACV hose with the engine idling poorly, nothing changes. However, if I do that when the engine is idling fine, the car will bog down and it wants to die. I can explain more when I get home and get on my computer.

 

Before I drove to work today, I unplugged the O2 sensor. I noticed no difference whatsoever. Drove perfectly normal. No difference at high RPMs, full throttle, up hills, and no obvious change in the fuel consumption. My drive to work is about 30 minutes.

 

This leads me to believe that my O2 sensor is bad. So I'm gonna replace that anyways, though I don't think that will fix my idle.

[mudduck]

My 91 would do this. I cleaned out the pcv lines and the IACV and it seemed to solve the problem. And how are you running without the toilet bowl gasket? I just rebuilt a throtltle body, and dont see any way it wouldn't leak without that gasket.

[MilesFox]

MY 88 sedan used to do this. I went through a series of used junkyard engine temp sensors. I would suggest testing or replacing the temp sensor if you haven't considered it yet.

[jj421]

My 91 would do this. I cleaned out the pcv lines and the IACV and it seemed to solve the problem. And how are you running without the toilet bowl gasket? I just rebuilt a throtltle body, and dont see any way it wouldn't leak without that gasket.

 

Hmm, I'll check the IACV lines and stuff, but I'm pretty sure they're clear and free. Could be wrong about the PCV lines, since I haven't touched those much. And don't ask about the toilet bowl gasket; I honestly don't know. I went to install a new gasket, but I couldn't find an old one. There were witness marks from one, but I couldn't feel it. Keep in mind I didn't fully take apart the throttle body; I just undid it enough to change out that gasket. I installed a new gasket, started adding coolant, and the coolant level wasn't staying steady. Next thing I know, the engine is hydrolocked. I removed the gasket and viola, it was fine. Unless it was stuck on the top part of the TB and I couldn't feel it, but I ran my fingernails on it and it felt like flat metal.

 

I'm just putting the toilet bowl gasket thing on the list of, "It's a Subaru thing." It works, and I ain't complaining. I've been meaning to take apart a junkyard throttle body to see if I missed something, but I keep forgetting.

 

MY 88 sedan used to do this. I went through a series of used junkyard engine temp sensors. I would suggest testing or replacing the temp sensor if you haven't considered it yet.

 

When you say engine temp sensor, you mean the coolant temp sensor, right? Unless there's an engine temp sensor that I've never heard of before, haha. I've replaced my CTS twice, and am currently running a brand new one. Resistance readings are within tolerances as per the FSM at all temperatures, ergo it's working fine.

 

Currently uploading a short video to YouTube. Maybe it'll give you guys another view on the issue. One thing to note that I didn't really notice: the car won't start without playing with the gas pedal. If you don't touch the gas, it won't start. It'll just crank and crank, as you'll see on the video. Also, when it's idling poorly, disconnecting the electrical connection to the IACV does nothing. However, disconnecting the vacuum line will cause the engine to die.

 

I'm hoping to get a new O2 sensor from O'Reilly tomorrow, if I wake up early enough to go before work. It'd probably help to pass emissions in August, but my exhaust leak is probably the biggest issue for that.

 

I really wanna believe it's something to do with the IACV. If I start the engine and it idles fine, I have no CEL. Then, once it decides to idle like garbage, the CEL comes on with the "air control valve or circuit" code. But so far, all the tests I've done to my IACV have proven it to be working, as far as I know. Resistances are within tolerances as per the FSM. With key on engine off, you can unplug it and plug it back in and hear it actuating. Got voltage going to the IACV plug and I believe I tested the big round black plugs near the battery, and that checked out fine as well.

 

I truly appreciate you guys helping me with this. I never got any responses on my previous posts that I've done on this problem. I would like to put this issue to bed once and for all. I'm trying to give you all as much info as I can, and hopefully you guys will stick with me to the solution.

 

Okay, here's the video:

 

https://www.youtube.com/watch?v=AS4sJPCvA3Q

Edited March 22, 2014 by jj421

[jj421]

Well, replaced my O2 sensor and haven't noticed any change whatsoever. Still can't idle, still has the cyclic idle, still misfiring, same performance. Tested the O2 plug with my multimeter and it's fluctuating like it should.

[Silas Cruse]

On my 1991 Loyale if I unplug the green "test" connection then this idle cyclic happens continuously.

I plug the green back together and the idle returns to the "normal" 1050 - 1000 rpm

So hoping that someone has the solutions for your problem.

 

- watching this threat Intently ! -

[jj421]

I tested the coolant temp sensor when I got to work today. With the engine running, I unplugged the CTS and the idle increased in RPMs for like 10 seconds, then it went back down again. Plugging it back in changed nothing.

 

This was with the engine warm. I'll try it when I leave work today.

 

Also, with the engine running, I plugged in the test connectors, but it did nothing. I'll try again when the engine is cold.

[jono]

Well, you may have worked out why it did not change on disconnecting the IACV but did when you pulled the hose off ??

 

If you pulled hose off and not blocked the hose, you allowed unmetered, unfiltered air into the intake boot, leaning it out instantly

 

Not your answer though.

 

As this thread grows, so too are repeat questions. Can we do signatures in here? So you can list what you have done in point form in a sig to follow you about ?

 

If this was my problem at this stage I would be plugging up PCV valve hole solid and let rocker hoses breathe to atmos for testing, block off IACV totally.

 

I'd also be looking at all hoses running off inlet and throttle, one by one, disconnecting from inlet to rule out weird things - including brake vacuum booster - but don't drive it disconnected or unclamped !! For safety sake !! Not ever had a bad booster in a subaru, but imagine if fluid leaks back of master into the booster, forms a puddle and the booster diafragem has a hole, air blubs it way through ......to inlet manifold , weird idle - straw clutching now ....

[jj421]

Here, I'll make a short list of the things that I've done, off the top of my head. Keep in mind, none of these things made a difference to the idle problem, how the car runs, or anything.

 

• Replaced CTS with a junkyard one
• Replaced CTS with a new one
• Replaced IACV with a junkyard one, cleaned it before installing
• Replaced IACV with a junkyard one again, cleaned it as well
• Tested CTS, IACV, MAF, TPS, and O2 sensor. All checked out fine according to the FSM
• New O2 sensor
• Cleaned MAF
• Replaced intake from airbox to throttle body with a spare one
• Cleaned throttle body
• Cleaned EGR valve
• Replaced PCV valve with new one from dealership

Just played with the car a bit. Did some things that may or may not indicate stuff. With the engine running, I started disconnecting vacuum lines. IACV line causes the engine to wanna die. The hose going to the passenger valve cover causes the car to wanna die with detached. However, when I take off the vacuum line going to the driver side valve cover and PCV valve, nothing changes.

 

I took the hose off, it's clear. The was some oil residue on the PCV valve and valve cover, where the hose attaches. I removed the PCV valve, and doing the shake test, I could hear it moving around in there. Wasn't exactly rattling, but it made noise when I shook it. Installed everything back again, started the engine, and no change. May sound like a stupid question, but how far is the PCV valve supposed to go in? I can still see some thread on the valve, but it's pretty tight. I'm pretty sure that's how it was before, but I cannot remember. I don't wanna over-tighten it, since it's pretty darn tight right now. However, I did notice a rapid clicking from the throttle body. I don't know if it was there before, but I hear it now. Can't tell exactly where it's coming from, but it's a rapid clicking from the throttle body.

 

With the engine running, if I disconnect the vacuum line from the PCV valve, the engine will slow down and it won't run very well. I can put my finger over the PCV valve and it has suction. Reattach the vacuum line and the engine performance improves.

 

I then disassembled the top of the throttle body, but didn't see anything out-of-the-ordinary. Put it back together, started the engine, and no change. I decided to pull the two small vacuum lines at the front of the throttle body that go to the canister, but pulling those lines didn't change the engine performance at all.

 

I also disconnected the brake booster vacuum line with the engine running. Obvious vacuum leak. You could hear it and the engine just about died.

 

So, there's that. Also, not sure if it could be relevant, but I'll throw every bit of info out there: my engine does tick when I start it up cold, but goes away after a few minutes. Not obviously related to the idle issue, but maybe that can give an idea to someone.

Edited March 23, 2014 by jj421

[cumminsea82]

mines doing the same exact thing. subscribed.

[jj421]

mines doing the same exact thing. subscribed.

 

I always find this interesting; it seems like a lot of people have the same issue, or at least a similar issue. But nobody posts the specifics of their problem or any solutions. Oh well; hopefully we can get your car fixed too.

[cumminsea82]

I've done nearly the exact steps as you. but one thing I noticed awhile ago was that if I jiggled the wire harness all around and found the "sweet spot", my idle would go back to normal. eventually this lost its effectiveness, like the wires had a weak connection. so I replaced the wire harness from the junkyard but nothing changed.

[cumminsea82]

my idle goes up then cuts down only when I apply a little bit of throttle. (1000-1500) but if I don't keep a bit of throttle while coming to a stop the rpm's drop so fast it kills it. when the engine is cold it won't idle at all though. when warm it will idle sometimes, sometimes it won't and will die on me.

[mudduck]

Didn't you mention swaping the engine, and had the problem with the old engine?  You might want to check harness.

[jj421]

Well that definitely sounds like a connection issue or something. I'll try that on my car. I'll try jiggling the harness around, especially around the two black plugs near the battery.

 

Yes, I had the issue with my old engine as well. I swapped only the long block, so the intake manifold, throttle body, and everything else is still original to my car. I'm gonna check and play with the harness today.

 

Oh, one thing I forgot to mention: when I swapped the long block, it had a brand new water pump on it. That new water pump only lasted 400 miles before it started leaking, but during those 400 miles, I had no problems with my idle. It was idling fine and I never once had to rev it to keep it alive. But once it started leaking, I installed my old, original water pump. The idle issue returned.

 

That is the only thing that I've done that's affected the idle: swapping the water pump. Could just be a coincidence. but: new water pump=good idle. Old water pump=bad idle.

[MilesFox]

My reference to the CTS sensor is the one with the green pigtail with 2 wires. the single spade terminal is or the dash gauge.

 

With the mention of juggling wires, make sure the intake harness is grounded to one of the intake bolts on the far end of the harness.

 

Check or a corroded pin where the intake harness meets the body.

 

There isn't really much more to diagnose there than the idle is being manipulated by the ACV, by either some sort of input issue, , or some fault in the plumbing.

 

One thing to consider is there could possibly be a fault in the ECU board

 

As i stated before ,my 88 gl sedan would do this, sort of whenever it felt like it, but didn't have any performance or stalling issues.

[jj421]

I jiggled the wires before work and just now. Engine cold and engine hot. Made no difference. I couldn't find a sweet spot or anything. It does look like the wire harness is grounded to this bolt, which is located right next to the thermostat:

 

 

This does appear to be a good ground. I unbolted it, and it's clean and free of corrosion. It also has continuity to the negative battery terminal. With the engine running, I put the positive lead on the positive battery terminal and the negative lead on this bolt/wire. Read 14.08 volts, so it's good.

 

I disconnected the two black plugs and all the pins are in excellent shape. One thing that I found interesting was on the smaller black plug, there were six wires on the ECU side and only four wires on the engine side. I assume those two terminals are made for cars with cruise control or something else? But yeah, the connection of those plugs are really tight, and the connections are clean.

 

Is there a way I could test the ECU at all? Are all EA82 SPFI ECUs interchangeable? Can I go to a junkyard, pick one up, install it in my car, and see if it changes anything?

 

This probably isn't related, but in the past couple days, I have found these two wires in my engine bay. They're not plugged into anything, and just hanging out by the air box. The wires go into the harness that hosts the corner marker lights, passenger headlight, and cooling fan. They could easily plug into each other, but that wouldn't make sense, haha. What do these go to? I deleted my car's A/C and power steering, but I don't recall messing with these wires.

 

[MR_Loyale]

You tested  all the sensors and they checked out. But that doesn't mean the ECU sees the sensors are being good. There could be an intermittent issue in the harness wiring anywhere from the sensor to the ECU. Yeah I know you jiggled it, but did you jiggle every inch of it? Since it sometimes works, the ECU must be working. I am guessing the IACV circuit as reported.

 

The thing about jiggling a wire to find something is that it is more likely that the wire is broken and jiggling to make it connect is less likely than if it were connecting and working properly and the jiggling caused it to not make contact and the malfunction appears. I hope that makes sense.

 

In your case I suggest you get the wiring diagram for the IACV wires to the ECU and then disconnect your battery and do a step by step continuity check of every single wire involved by testing from the plug that connects directly to the ECU and the one that connects directly to the IACV. It is a PITA but really the only systemic ways to isolate the problem.

[jj421]

I didn't jiggle every inch of the wiring, but I did mess around with most of the wiring from the black plugs to all the intake components. From there, the wiring goes into the fender.

 

Yeah, that all makes sense. Later on this week when I get some time in the daylight, I'll do some continuity and resistance tests. First that comes in mind is the IACV plug to black plug near battery, then black plug to the ECU plug. I'm not sure where there might be other connections, but I'll check out a wiring diagram before doing tests.

 

I believe it's gotta be the IACV circuit, since that's what the CEL code is for.

 

Only thing that doesn't make sense about it being a wiring problem: whether it works or not isn't random. My car will idle fine if it's warm outside. My car idles fine when I'm offroading and rock crawling. High RPMs, low speeds. Otherwise, it'll idle like garbage. If it's a damaged wire harness or something, I don't think it'd be temperature related. If it's cold outside, my car will never idle. If it's warm outside, sometimes it'll idle fine and sometimes it'll idle poorly.

 

I can also sometimes feel my IACV working, or at least what I believe to be the IACV. The other day, it was idling fine for a bit. I came to a stop and it was idling at like 1200 RPM. Then you could feel/notice the IACV adjusting as the idle went down to like 900 RPM.

 

In case anybody is wondering, here's a quick test on my coolant temp sensor that I just did. I used to have a CEL code for the CTS, but I no longer have that code since installing the new CTS. My coolant is sitting at ~74 degrees Fahrenheit, and the CTS had a resistance value of 1650 ohms. This is within tolerance as per the FSM.

 

I am going to do some IACV testing, as per the FSM.