this one beats me

[loyalist]

Hi, all:

My name is Lewis and I'm new to this forum. I have a 90 Loyale that I've purchased new. It has 180,000 miles on it, but only about 100,000 on the engine. I've lost the original engine due to a broken waterpump bearing on a cold January night up in the mountains close to the Cal/NV border when I tried to get to a place to call for help. Purchased one of those "Low mileage used japanese engines" and installed it myself, relocating the intake manifold and up, because the replacement engine was a carburetted one, compared to the original SPFI. Anyway I've just done some work on this engine lately, namely: new belts, tensioners, idler, oil pump, hydraulic lash adjusters on the driver's side, waterpump, clutch, rear main seal, throwout bearing, alternator, a Bosch starter, clutch and accelerator cable.

It ran like new again until about two week ago, when it started acting up like this:

* the idle started fluctuating. Dropping down to around 500 and the rising to over 1000 rpm and back to normal.

* the engine stalls after taking off, driving about 100-200 yards. Starts up okay, but it never did that before. And it did not do it all the time, only a couple of times a week.

* after a 200 miles drive when I slowed down at a Hwy exit, the engine died. Started up easily. Gassed up the car and it was hard to restart.

* another 150 miles okay on the highway, 75-80 mph, but arriving to destination the engine dies when lifting foot off the accelerator and applying brakes. Restart after every stop at a red light.

* next day starts up right in the morning, dies at stops, running excellent on highway. Got back home safely.

* started up next morning, but won't pick up rev over about 1000 rpm with the pedal floored.

* next time it started up and ran "normal" - I mean it is driveable, but dies at stops. It actually can be kept running by paying attention to not to allow it to down to idle. But after that short critical seconds it is willing to idle normally again.

 

I'm not a trained mechanic, but I'm attached to my vehicles, don't replace them every 5 years. Beside this Subaru we own - and drive daily - a '72 Land Rover and a '71 Citroen Mehari. I keep them running. I'm - as they say - mechanically inclined, most of the time even enjoy working on our cars, but this one beats me. Anyone can help me to solve this problem?

 

As a possible clue another detail: I replaced a cracked vacuum line that goes to the brake booster. It is not original part, not even metric, but the closest SAE size (I believe 3/8") and I couldn't really push it on the little upward curved metal pipe that 's sticking out from the booster drum. It was also a soft-wall hose. Then I replaced this with a heavy-wall one and put clamps on both end. When I disconnect it at the manifold end, leaving the booster end attached and blow/suck into it, placing my fingers at the booster drum where the hose is attached, I can feel air coming out when I blow into the hose (clamp is tight).

So I thought there is an air leak there. Then I disconnected the hose at the booster drum, leaving the manifold end attached and plugged the loose end of the hose. It did not make difference.

 

Any ideas? Please Help!

 

Lewis

[northguy]

I'm certainly not a mechanic, nor am I too scoobied up on Loyales, but your problems smack of a vacuum leak or a plogged fuel filter (or both). Just my guess, and I am sure others will chime in with more informed prognoses.

[soobme]

It sounds like a vacume leak to me 2, but fule filters are cheep, eazy to re-place and can solve LOTS of strange problems.

I would start the car up, let it warm up, and then start spraying brake cleaner arond the manifold/vacume ports &lines. If the idle drops, U know U have found your vac.leak;)

[loyalist]

Thanks guys!

I hope it will be that simple. The fuel filter is about 8 months and maybe 6000 miles old. Since the vehicle comfortably cruises at 75 mph is it still possible that the fuel filter is - even partially - clogged?

Definitely will try the brake cleaner - or starter fluid - sprayed around the top of the engine where it can be sucked in by the engine - if there is a leak there. I tried to shake every visible hose and cable connection on the top of the motor to see if the running would change.

So you think it is not some electronic engine control, or sensor problem? Because those I don't know nothing about.

 

Lewis

[JonOfScio]

when you changed your engine from SPFI to carb, did you remember to change the pump? or get a FPR?

[loyalist]

I was not clear I believe. I transferred all the manifold/fuel injection stuff over from the trashed engine to the "new" used - originally carburetted - unit. I understand the cams of the fuel injected engine are slightly different, but that time (about 5-6 years ago) I was not aware of this. New fuel pump installed three years ago.

 

Lewis.

[GeneralDisorder]

One thing you should check right off is the ECU codes - are you pulling any codes from the computer? I have had amongst other things bad 02 sensors cause very similar things to what you are describing. Also check what other people are saying - vac leaks is high on the list - and fuel filters. I would probably replace those on general principle....

 

GD

[Skip]

I like GD's sug on the 02 sensor. If it is functioning somewhat it will not set an error code.

Check it's operation by viewing the 02 monitor LED on the end of the ECU. (the same LED used for checking for stored error codes) If you have not found the ECU yet, it is to the left and under the steering column, you must drop the trim panel to see it. A gold/silver box, the end facing you will have a dark circle, this is the LED. With the engine warm and idleing you should see a green LED flash at random intervals at a freq of about once every 5 to 10 seconds. If you do not see it flash, then it has probably run it's course and wants replaced. A universal single wire Bosch is around 20 dollars.

 

BUT I think the idle / no idle problem is the IAC or Idle Air Conrol valve. This electrical devise located on the very front of the throttle body, is the idle speed control. It gets gummed up over the miles and wants it's air control valve cleaned. It can be removed and cleaned with carb cleaner.

General assumtion - this car has SPFI

HTH

[loyalist]

Thanks GD and Skip!

Yes, the vehicle is a SPFI. But now we are getting where I have no knowledge of - the electronics. I still have a healthy amount of common sense, so if someone would explain how those things work there is a fairly good chance I will understand.

I understand this car has at least a half a dozen different sensors monitoring the operation of the engine and other circumstances. I have a list to tell what those code numbers mean, but have no code reader. I tried at Pep Boys but they don't have one that covers my car. Is there a sure place to find one or I have to call around all the different suppliers? How much is it roughly?

In case that green LED on the ECU is blank, do I have to take out and dismantle the unit in order to replace the LED, or is it like replacing a lightbulb?

Skip said that the O2 check/indicator LED is supposed to flash green if the sensor is still OK. No light means burnt out LED. What should I see if the LED is good, but the sensor is bad?In case the O2 sensor is bad, where is it? How can I access it? Do I have to rise the car up, or from a pit?

I know where the idle control thing is. Is there anything that I need to know to avoid possibly ruining something in the process of taking it off for cleaning? A special gasket, seal, or similar to pay attention to?

 

Lewis

[GeneralDisorder]

Ok - well one thing you'll be happy about - no need to buy a code reader as these cars don't use them. You read the codes by connecting two connectors together under the dash (or under the hood in some cases I think), and reading the codes from the long and short flashes of the light on the ECU. Do a search on EA82 ECU codes on the board and I'm sure you'll find which connectors to use - or someone else here (like Skip) can help with that as I don't know EA82's.

 

The 02 Sensor is located in the exhaust manifold - it's right where the two pipes from the heads meet - at the "Y" in the pipe. Check it's operation via the ECU light first tho before pulling it. They can (and almost always are) a REAL pain to get out. Not because they are hard to get to, but because they are usually frozen solid with rust, and the alternating hot / cold of the exhaust seems to cause real problems that way. If you do replace it you will most likely need a special 02 sensor wrench or socket. And remember to use a liberal amount of anti-seize on the threads so it can come out if it ever needs replacing again.

 

As for the LED in the ECU being out - I've never seen one go out. LED's are rated for insanely high hours of use. If it is out - best to just ask on the board if anyone has another ECU.... unless you are good with electronics, then you might be able to solder one in... I wouldn't personally bother - there's plenty of ECU's out there to be had.

 

I think Skip meant that the 02 sensor had "run it's course" - not the ECU LED. If you see no light, then the 02 sensor is most likely bad. Replace it, and then check for a light.

 

GD

[Skip]

Lewis, please excuse me for going on like that.

First I would like you to test the vacuum brake booster.

With the engine off. Pump the brake pedal several times and hold it down. Start the engine -does the pedal fall towards the floor slightly? If it does it's probably okay.

Now

GD is correct, I was refering to checking the 02 monitor for proper 02 sensor operation.

Do as I said and watch for the flash, if no flash then talk about changing the 02 sensor. GD is right on this one to. If you are replacing it you may just cut the wire and use a deep well socket.

If you have never changed it and have owned it that long well sorry to say, it probably wants changed.

He is also quite correct in that we do not need a scan tool of any sort.

Here is a link for a chart and instructions for checking codes, but if you have not seen a "Check Engine" light, there will not be any codes stored.

Your connectors are under the hood, in the area of the drivers side hood hinge.

http://www.troublecodes.net/Subaru/

The IAC can be removed, cleaned internally and replaced, use care and you will not need a new gasket.

You will do fine, you have a lot of "auto-sense". If you have made it this far with no FSM (Factory Service Manual) I have faith it the fact you can fix this.

Let me know if I left anything out

[loyalist]

Thanks guys again for replying so quickly!

Today I started the engine up. First it did not want to rev over a 1000, but then it changed its mind and actually became very responsive to throttle. Almost peppy.

Then I sprayed starter fluid all over the hose and flange/seal connections and it did not make difference. I expected the idle to rise, but it did not happen.

During initial startup it backfired once or twice into the air cleaner and as I'm writing this I remember that it happened earlier during these experiences. Once, when my right foot was on the brake pedal (left one of course on the clutch) I felt a strong kick on the brake pedal, when the engine backfired. Is that normal? Isn't that is supposed to be a one-way thing (vacuum only from booster to intake manifold)? My Land Rover's booster drum has a replaceable check valve, where the vacuum hose is attached to it. Also let me mention again that when I blow into the vacuum hose's intake manifold end, placing my fingers down at the hose/booster connection (clamp on this end of the hose is tight) I can feel - and hear - the air coming out. Is this normal?

Today I also discovered the ECU. There is a LED on it, which is black until the engine warms up. Then it gives RED - not GREEN pulses. I've also found connectors hanging there: a brown one, with two female spade connectors in it (perpendicular to each other), a green one with four similar connectors in it, (parallel in pairs), then a blue one with three connectors (two parallel, one across), and two white/clear connectors - one of them has four round female thingis in it and the other one has five connectors with a sixth space left blank. I used once a pair of matching green connectors - plugged together for the test - to adjust ignition timing. But it was in the engine compartment. A friend suggested it, I did not just figured out it myself. This friend of mine unfortunately left town since then.

The LED - after the engine warmed up - started pulsing. Those pulses were of different lenghts.

Now I will look up the pages you suggested and take it from there - if I can. But I really appreciate the way you guys hold my hand and try to walk me trough all this!

 

Thanks again! Lewis.

[MilesFox]

those green timing clips are what you are looking for. the test clips are under the hood for

SPFI, under the dash for turbo

[loyalist]

Thanks, MilesFox!

But, please explain in this case what am I supposed to do with those green plugs that I plugged together for the ignition timing.

And what to look for or check after I connected them together again? We are talking about that matching - maybe two-wire - green plugs hanging behind the brake fluid reservoir, right?

 

Lewis

[Skip]

Lewis, sorry on the color of the LED, the newer cars changed to red. It sounds like the 02 sensor is funtioning normaly.

Guess you missed me saying

Your connectors are under the hood, in the area of the drivers side hood hinge.

Didn't care to do my brake booster test?

That was suggested from you saying "the booster end attached and blow/suck into it...I can feel air coming out when I blow into the hose "

To answer one of your questions, yes there should be a check valve in the large line to the power brake booster.

You may have a crack in check valve from your blow/suck air coming out description.

Did you spray this area with starting fluid to see if there was a leak?

The back fire is an indication of running lean, the brake booster line check valve if broken

could be causing a large vacuum leak and be the source of your problem.

The GREEN connectors (a matched pair of single wire connectors above the master cyl.)

are to be plugged together for setting the ignition timing as you say. The link I gave you

will discuss the pair of white/maybe black connectors

and their use for reading trouble codes. These will also be in the same area as the green connectors.

 

Have fun and good luck, I tried.

Sorry I could not have been of more help.

Skipout

[loyalist]

I've re-read the whole thread and couldn't find where you mentioned the location of those plugs.

I forgot to do the brake booster test you suggested - it was not out of disrespect. But I sprayed the starter fluid in that area too and the rpm did not pick up. The cracked original vacuum hose I replaced did not have a check valve between the manifold nipple and the booster.

The link you suggested is for model year '96 vehicles and the locations and colors of connectors on the drawing are different from mine. All the connectors I've found sofar under the dash are female. Or is my car a lesbian? I will check the ignition check connector area for another pair today - along with the booster test as you suggested.

Never had the "Check Engine" light on with the engine running since I have the car.

Someone told me that there are "Read Memory" and "Test Mode" connectors that need to be plugged together depending on what you want. Retreiving the stored codes, or do a test on the running engine.

Looks like the problem is either the Idle Air Control valve - or the brake booster/vacuum line/check valve. All on the top of the engine fortunatelly, because I have no access to lift or pit to work on the underside.

Will take it from here, but I applied the starter fluid very generously and the idle did not change a bit at all.

Thanks for your patience and help Skip!

 

Lewis

[calebz]

Loyalist. Take another look at that link. It covers 83 and up trouble codes.

 

Your connectors are under the hood, in the area of the drivers side hood hinge.

 

Did you spray the brake/carb cleaner around the mating surfaces of the intake manifold?

[loyalist]

I've found the codes for my vehicle, but the connectors' location and color in the article are different from mine. The link says: "Retrieving Trouble Codes 96 & Later ... " Mine is a '90.

As you - and others - suggested I have a pair of connectable green plugs at the firewall (engine side) that I used for checking/adjusting the ign. timing. Are these the ones that I have to use to read stored codes too? I don't want to make such foolish mistake like shorting out something that will make finding the cause of the problem harder.

Yes I sprayed starter fluid around the intake manifold/head flanges too. Almost used up a whole can to make sure I don't miss any connection, or location for possible air leaks. No change in rpm.

Aren't there supposed to be separate pairs of connectors for retrieving stored codes and for diagnostic testing?

 

Lewis

[Skip]

I guess my email did not make the trip?

You are correct Lewis the headline is a bit confusing

The link head line reads

"Subaru Trouble Code Info

Retrieving Trouble Codes

96 & later are OBD2 systems"

This is to tell you IF you have a 96 or later, the car is equiped with the OBD 2 system and the procedure outlined in the link will NOT work.

ALL systems up to 96 are covered, please scroll down the page you will see what we are trying to tell you.

Here is a link to another site

Early Subaru Trouble Codes

And yes, in the area where you found the green connectors (these are called "test mode connectors") should be another pair of single wire connectors. They will be of another color, probably white. (these are called "read memory connectors")

I did a search on the board for you.

Please look at this post and read the first post by Asavage he explains it better than I can.

http://usmb.ultimatesubaru.org/forum/showthread.php?s=&threadid=4449&highlight=trouble+code

 

I do not think this is going to help your current problem, I'm sorry.

The part about blowing and sucking on the booster hose has me confused as you are about the problem.

Could you please tell us a little bit more about this b and s

you did.

[loyalist]

Thanks Skip for claifying the connectors for me which one is the diagnostic pair and which one is the stored code retrieveng pair. And also their location. That link has a drawing showing all of them under the dash.

I received your e-mail, but I wanted to write you after I checked the booster as you suggested.

This is the area that is the most suspicious !

The power booster must be working, because every time during this "ordeal" when the engine died I lost not only the power steering, but the vacuum assistance on the brake too. Actually the engine died every time when I used the brake to slow down the vehicle to full stop. But if the engine is just idling and I step on the brake, it doesn't make any difference. If I pay attention and don't let the rpm drop down to idle while stopping/slowing down the vehicle the engine keeps runnung.

But back to what I did in that area:

About 6 months ago I replaced the vacuum hose that goes from the intake manifold to the power booster. The old one was cracked. I did not even put clamps on it, just pushed it on at both ends. The manifold end went on completely, the other one only about a half an inch or so. It was a piece of 3/8 rubber hose, with realtively soft wall. But it worked even though I did not like the softness of it. A good month ago I replaced this one with a thicker wall hose and put calmps on both ends.

After this stalling problem occured - being suspicious of evrything I've done on the car lately - I first disconnected the hose at the manifold nipple and taking the loose end in my mouth I blew air into it and applied vacuum alternately. Under vacuum I expected a certain amount of air coming out, then choking off, but I can suck air out of there indefinitely. Then I blew air into the hose and heard it coming out at the booster drum. Placing my fingers down there I could feel the air coming out. Isn't it supposed to be sealed? Like blowing into a bottle? There must be some flexibility felt because of the diaphram in the booster, but basically it is a closed space isn't it? So this is suspicious enough.

Next thing to do was to disconnect the booster end of the hose, reconnect the intake manifold end of it (clamped tight) and then plug the loose end. If the booster has a leak that coused the problem (leaning out the mixture) this would have corrected it, right? Would loose vacuum assistence on the brake, but the engine should go back to normal. But it did not cure it.

I plugged the hose with the tip of the rubber coated handle of a pair of pliers, so maybe I will have to place a rubber plug on the nipple on the intake manifold to make a positive seal there.

What I'm going to do now is to go down to the vehicle (nice sunshine and hopefully not as cold as in the morning) and do the booster test as you suggested, after warming up the engine connect the code retrieving connectors and check the LED on the ECU. I'll send you an e-mail what I've found.

 

Lewis

[loyalist]

Hello, Skip:

I've sent you an e-mail yesterday reporting my latest tests and findings. I wonder if you received it (or just got fed up with my problems...)?

Tested the booster. With my right foot on the brake pedal I started up the engine and it sucked the pedal in about a half an inch or so. Found and connected the test code retrieving plugs (a white and a black or brown in this case) and the LED pulsed an 11, 13 and 14 stored trouble codes. Yesterday I purchased the Subaru Chilton's manual and found these areas in it. 11 and 13 means "Crank Angle Sensor" problems, while 14 is an "Abnormal Injector Output" signal.

According to Chilton's the Crank Angle Sensor is part of the Distributor assembly. The Injector's abnormal output must be the consequence of the air leak that I have somewhere. Or it is more correct to say, that the injector output may be right, but the air is too much that goes with it?

Test drived the vehicle around the bloc and had good power and acceleration, but still has the tendency to die upon slowing down the vehicle with the brake to a halt or near halt.

I can keep it alive if I don't let the rpm drop down to idle. Meaning left foot on the clutch to put the tranny into neutral and the right foot on both other two pedals. After the tranny is in neutral, just one foot per pedal (brake and gas).

Also found an illustration in Chilton's about the brake power booster - not a very good, detailed one - but it does not show a check valve. In the text it mentions it. I can't find one on my vehicle! Called the local Subaru dealer and they looked it up and it is supposed to have one in line between the intake manifold nipple and the booster drum. Ordered one (about $26.00), but a matching "Hose Kit" to install it would cost $99.00!). Will use generic.

On a start-up from cold it runs at around 1600 rpm that then gradually decreases and finally goes down to around 750.

Today I testdrived it again to check if it dies if I don't use the brake to slow it down. First three attempt NO, but the last one when I parked it, it died again. Who understands this?

What I still can - and will - do (beside installing the one way valve in the brake vacuum line after the part arrives) is to remove and clean the Idle Air Control assembly. By the way: isn't the Power Steering Reservoir in the way, to access the two lower screws? But as the idle changes from 1600 down to 750 during warm-up suggests that the thing is working (maybe a little dirty after 180,000 miles).

What still bothering a little is that it was the last time at least 4 years ago when someone worked on the brake master cylinder (got a new one). How could that check valve to be missing? And howcome its absence did not cause problem only now?

 

Lewis

[MilesFox]

try testing the MAF. i have seena car with a bad filament in the MAF, it had flat spots in acceleration, but didnt stall.

 

you caould try testing the coolant temp sensor for the ecu. its the one with teh plastic clip, not the single blade terminal(temp gauge), on the thermostat housing of the manifold

[loyalist]

Thanks MilesFox for your thoughts!

What is MAF? The vehicle accelerates pretty well and smooth and has no flat spots.

Since the idle from star-up 1600 goes down gradually as the engine (coolant too) warms up to a fairly stabile 750 doesn't this indicate that the ECU senses the coolant temperature (and its changes) well? Or is there something else that still can be wrong?

The "only" problem now I have with it is its tendency to stall when I let the rpm drop to idle while applying the brakes. As I mentioned in my first bringing this whole thing up - a week or two before it got that bad - was that I noticed the idle fluctuating. I don't know if it had anything to do with the problem later followed, but it never did that before throughout the 13 years of use. It dropped down from 750 to around 500 then went up over a 1000 then back to around 750 again. Maybe it had nothing to do with the rest, but sometimes there are "early warnings" before the real trouble comes, but one has to underdstand the signs. So, considering the depth of my understanding how a modern, ECU controlled, fuel injected engine works, for me this might have been an early "cry" for help from the car.

 

Lewis

[MilesFox]

the air valve that skip mentions on the throttle body introdices air to keep the idle up. its like a false vaccuum leak how it works.

 

i had found it easiest to remove the throttle body from the intoake, from a complete motor.

 

remove the 4 12mm bolts that hold the throttle body to the intake, didconnect 2 vacuum lines, 2 electricle connectors, and the throttle body.

 

but you will want to get the mounting orint ot renstall it should you remove it. i say this if you want to clean that valve, see if it helps, withput removing the power steering pump and such.

 

the MAF (mass airflow sensor) is by the filter housing. you say the car runs without flat spots, so that shpould not be the problem. if you unplug the MAF with te car running, and the idle changes pr car stalls, then you will then know its working properly.

 

chack all the little tunes and hoses around the spfi intake tube, make surte its all hooked up.

 

does your heater vents selector work properly? because if it only blows out the dash, thats because one certain vacuum line is disconnected. let us know if the heater selector works......

 

it may seem irrelevant, but at least we can elimintate possible causes for effective troubleshooting

[loyalist]

Thanks, MilesFox!

To remove the whole Throttle Budy, with all the vacuum, air intake, fuel line and accelerator cable seems to me more complicated that remove the PAS pump.

I checked all the vacuum lines, hoses, electrical connectors, they are solidly in place.

But now that you mentioned the heater control box behind the dashboard I remember that about 3 years ago we noticed that the air can not be directed against the windshield. That time I still had a Subaru trained mechanic (who was not only knowledgeable, but also thorough and pricewise reasonable) taking care of my problems when they occured - since then the place unfortunately went out of business and himself - while still in the phonebook, never answers my calls. So this guy said that the heater control box is bad and a new one would cost around $400. Buying a new one was out of question, so what I did was - after checking that all the vacuum hoses going to/from that box are connected, but they did not move the lever to direct the air upwards - I moved the lever halfway up and tied it there with a piece of wire to keep it in that position. Now the air is always coming out both at the dash vents and the defroster vents.

There possibly could be a leak there too, inspite that all the hoses are in place. But this poses the same question like the brake power booster line/check valve. Why the ECU - and the engine - did not notice it for years? The car ran (when it ran) flawlessly until now. There were times when it did not run, but it it either ran well, or did not ran at all (due to broken timing belts or similar).

 

Lewis