1987 GL10 Turbo, 7 codes and intermittent CEL

[newrider3]

I've been having some check engine light issues with my '87 GL10 turbo wagon and thought you guys might be able to help me make sense of it.

 

A few months ago I would occasionally get a flash from the CEL along with a misfire/stutter from the engine if I pushed it too hard (~4000 rpm, 10-11psi according to boost/vac gauge) that I attributed to running it too hard with regular fuel.

 

More recently I had the light come on after accelerating normally onto the highway. It stayed on, so I pulled over after about a mile. Aftermarket oil pressure, water temp, and vacuum gauges were all within spec, engine was running great, no abnormal noises. I decided to keep going, and then the CEL was off by the time I got out of the gas station parking lot.

 

Now it came on and has stayed on for a week or two (in college so it sits mostly), with occasional flashes off when idling, or when turning a sharp corner(?). I pulled the codes and this is the results (and some of my observations):

 

21-ECT (coolant temp sensor) - Tested it with my DVOM and the resistance values are within spec at different temps. The temperature sender has been removed to make room for a mechanical gauge probe, but the electronic sensor is still in place.

 

23-MAF - The only manual I have access to is the Chilton Online database, and it doesn't list a test procedure for the MAF. Tapping on the sensor housing while idling causes no change, and engine rpm drops severely to a stumbling idle if unplugged. Is there an actual test procedure/resistance values for the MAF on these cars?

 

32-O2 sensor fault - I haven't crawled under the car to test this yet.

 

33-VSS (vehicle speed sensor) - The car is lifted with 27" tall tires, so of course the speedometer is way off. Will this even effect the way the car runs, or trigger the CEL by itself?

 

35-Purge Control Solenoid - Again, Chilton's Online is crap, they don't list a test procedure or even show a picture of this. Any suggestions?

 

42-IAC (idle air control valve/switch/circuit) - This seems the most suspect to me. Cold starting the car is fine, but it revs up to 2000-2200 RPM then slowly idles down over the course of 5-10 min. Is this normal for warm up? Then the car is difficult to start after a heat soak, I have to hold the throttle WOT when cranking to get it to start. Occasionally after this hot start it will have a severely low idle.

 

I attempted to test this component, but again, Chilton's sucks, and only shows a picture of the IAC for a 2000+ Impreza. So I basically probed every wire off the throttle body for resistance values, and unhooked connectors while idling (no change in idle speed). Oreilly's doesn't even list the IAC in their catalog, so I couldn't get one to look at/price check. Does anyone have a pic of the IAC and it's harness connector so I can see if I'm testing the right thing?

 

51-Neutral Safety Switch/Circuit - This is obviously faulty, as I can crank the car in gear, with the clutch out, or a combination of both. This isn't a big deal to me (every manual vehicle we own is older and does this), is it a big deal to the computer? Will this cause a CEL by itself?

 

 

I really have no idea if any of these codes are uncleared legacy codes, and the system won't let me clear the codes until it thinks everything is fixed.

Does anyone have any input or suggestions?

Thanks!

[naru]

The ea82 manual here should help.

http://www.finleyweb.net/JonsStuff/SubaruDocumentation.aspx

 

That many codes at once suggests a wiring problem to me.

Maybe a connector or ground.

 

I`d consider measuring the ECT resistance at the ECU plug while moving things around.

Erase the codes and see what comes back too.

 

Hard hot starts may be related to code 51.The ECU doesn`t know it is supposed to be in start mode.

Edited October 30, 2012 by naru

[newrider3]

Thanks for the link, very helpful.

 

Can I erase the codes? Following the procedure in the code reading thread from here, hooking up both connectors does nothing. I was under the impression that you can't clear the codes unless the computer detects no faults?

 

Also, I just finished testing the O2 sensor. Key on, engine off, reads .60v.

Engine running, it is steady around .58-.59v, I can't tell if there is a true wave function with this cheap multimeter though. Pulling a vacuum hose results in no change.

Edited October 30, 2012 by newrider3

[naru]

All the codes are still there?

 

If the codes are still there,the faults are still there.

 

O2 won`t come alive until it is good and warm.

[newrider3]

 

O2 won`t come alive until it is good and warm.

 

 

Yes, I know. Engine was at full operating temp.

Funny thing is, right after I posted last I went and started the car again (hot start) and it fired right up, no CEL. Went for a drive and it's running noticeably better (never really noticed the change before).

 

Got back, shut it off, checked the codes again and they were all present. Hooked up both connectors and started the engine, CEL began flashing , unhooked them and the codes are deleted.

 

Seems like a ground or connection problem somewhere is a likely cause. I'll have to see if the light returns, in the meantime I'll do a little digging in the wiring diagram to figure out potential problem areas.

[scoobiedubie]

You said that you were runing on regular fuel. The gas cap cover states that you should not be running on 90 octane or better, which is not regular fuel. Just pony up for some better gas and problem sovled. The engines respond well to super gas. My experience with regular gas is that they have problems climbing hills without having detonation.

[CrownLoyale]

The purge solenoid should have about 40 ohms, and shouldn't be able to blow through the ports when not activated. I had 5 codes stored in my 90 Loyale with one of them being a purge solenoid code. Mine was open and it let vacuum through it. I had a similar one lying around and wired it up to the harness. Cleared the computer and haven't had the light return, been about a month. Although I didn't have driveability issues, the light just bugged me.

[scoobiedubie]

Pull one of the battery terminals if you want to clear the codes.

[skishop69]

Check all your grounds. Specifically the ones on the engine and for the ecm.

[AKghandi]

Extra boost on reg fuel? No bueno.

 

A perfect running ea82t at most, can handle 12psi and you said your running 11psi? You sir are playing with a grenade with the pin pulled.

 

The maf gets heated by the ECM and cooled by incoming air. So more air means the ECM is working extra hard to get the correct reading.

 

I had my car on extra boost for a while.i ended up replacing all the sensors you mentioned, but it didn't get fixed until I replaced the intake pipe and maf. So I turned the boost down to be safe. No problems since.

 

The studder you get when your pushing it is over boost because of a lean condition.

 

 

I also get the occasional CEL which I think is related to poor a/f ratio. Because on stock boost I get overboost sometimes.

[newrider3]

Thanks for the suggestions. I have seen no documentation on the car anywhere calling for anything other than regular unleaded, so that's what I've been running.

Also, the engine is 100% stock, no wastegate mods or anything to raise the boost.

[Caboobaroo]

Ceck for vacuum leaks too. The old, hardened rubber lines for the boost control solenoids can crack which will let you run more boost pressure depending on which one it is. Also, the rubber turbo boot can crack, letting in unmetered air while accelerating but can seal back up decent enough for it to run good at idle. Easiest thing to do would be to grab a can of brake clean and while it's idling, spray it around all the rubber vacuum lines, the rubber couplers between the MAF and the throttle body, base of throttle body, and the intake gaskets. When you find a leak, the idle will raise and then drop back down.

[AKghandi]

with this engine you would be better off to run premium.

i always run premium, granted its a little more expensive but its better for the car. especially when the motor should have sploded 10 years ago.

 

if your boost gauge says 10-11 psi, you are over boosting.

stock boost on the EA82T is 7 psi. i recommend finding the cause of the extra boost first. and after thats fixed im sure most of the codes will disappear.

 

i hope you get it figured out

 

also can we get some pics of the "offending vehicle"?

[Quidam]

With a stock EA 82T the effective compression ratio is somewhere around 11.5 1 at full stock boost. If you don't have required octane for that the knock sensor pulls timing back.

 

If it's gone as far as it can with that it will ping. And there is pinging not readily noticed unless you have the hearing for it.

 

Any ping is destructive. The reason some people just aren't equipped to deal with a turbo car, and really shouldn't own them.

 

Doug

[newrider3]

The CEL is back, narrowed down to two codes. First was an O2 sensor fault, attributed to a 'stuck' sensor (no signal change even with drastic AFR changes). I replaced this and saw improvement in the way the car runs, it's probably been running for quite some time without ever getting into closed loop due to the fouled sensor.

 

Next issue is an ECT sensor code. I've been experiencing occasional rich stumble/other hesitation when accelerating and decelerating, as well as an excessive high idle at cold start that never fully drops to idle. I'm having trouble with finding information about the coolant temp sensor and circuitry itself, it seems for each source I try I find conflicting information.

The Chilton's diagram shows a single wire from the sensor to the ECU, white/black, while the manual from the link posted earlier in the thread shows it having two wires, WB and BR.

From looking at photos of replacement temp sensors and temp senders online, I'm getting even more conflicting info.

 

Here's a couple pics showing what I'm working with:

 

The top of the t-stat housing, what I've been assuming is the temp sensor. The two wire plug matches up, but the blue/black wires don't, and I'm not sure what the sensor is physically supposed to look like. Probing the male side (sensor) gives me resistance readings at different temps matching the ECT specs. Probing the harness pigtail with key on, engine off shows nothing, and I read alternator voltage (~12-14v) with the engine running. Both sides also show continuity to ground (??) so I believe there's a short here somewhere.

 

 

Pic two is where (what I thought was?) the temp sender was, now replaced by a mechanical gauge sender. You can also see the white/black stripe wire that connected to it (can also see a white/green stripe wire that has been hanging since I got the car, anyone know its purpose?). If I go by the Chilton's manual, it would lead me to believe that this should be the temp sensor rather than the sender, but I don't believe this is correct, can anyone shed some light on this?

 

[AKghandi]

your a little off. i believe the thing on the thermostat is the iacv.

 

when your standing at the front of the car, the engine temp sensor is a little south east of the turbo, its on the coolant crossover pipe.

 

 

 

BE CAREFUL WITH THE CONNECTOR, they get very very brittle sitting next to the turbo.

Edited December 3, 2012 by AKghandi
pics

[newrider3]

Thanks for the pics, that's exactly what I was looking for. Got to the ECT sensor plug (managed to break the egr vac solenoid whilst down there) and probed it for voltage, shows a steady 3.14v both key on engine off, and key on, engine on. Is this the correct reference voltage?

[AKghandi]

i'm not well read on the tech specs, but the sensor was like $19 iirc.

you can pull it and test it with hot water, to see if its resistance changes as the water gets hot.