Dead Legacy RS - please help!

[Cret]

Horror of horrors, yesterday afternoon the unthinkable happened and after 2 years of not missing a beat my beloved motor has decided to go on strike.

 

Was in the middle of fitting my new backbox after driving it down to my mum's garage to do the job, when the unmentionable happened.

 

Half way through cutting & welding the backbox (off the car) I started up the car and it sounded rough, ran for about 10 seconds or so, stalled, and now refuses to start.

I'm completely certain the exhaust has nothing to do with the problem as it was the same on the car when it conked as it was before.

 

It turns over normally but will not fire in the slightest. There's plenty of fuel in it and no fuses appear to be blown so I'm really not sure what can have packed in to make it do this.

 

If anyone has any cunning ideas of likely causes I'd be incredibly grateful as having just spent loads of money on various things for the damn car, and getting some bodywork done etc I'm cleaned out, meaning a hideously expensive trip to a garage is not really an option right now.

 

I'm really p1ssed of about it as I was meant to be going karting today but instead I'll be scouring the web for possible causes and leaning over my engine bay getting backache. What an utter bummer.

[Opie]

Cranking but not starting usually means you are missing one of two things, fuel or spark.

 

Are you getting spark to the spark plugs? If not the coil pack might have gone bad, just as a bad plug or bad plug wire could also cause this. Has the fuel filter been changed recently as it could be clogged, or the fuel pump could have gone bad.

 

Hope this helps!

[theotherskip]

that sounds like you did something to the fuel pump electrical lines while you were working under there. what you probably got is the car starting on the last of the pressure in the fuel system. you probably either nicked or completely cut a wire to the fuel pump. start by checking the fuse to the fuel pump, then have somebody turn the key from off to on while you are listening under the gas tank for the fuel pump. if you can't hear a whining noise, open the access panels to get to the top of the tank and see if you are getting voltage to the pump... your answer is probably not under the hood, more likely in the area you were working...

[Cret]

Thanks for the suggestions guys.

 

I had wondered about coil packs as it was recently misfiring when cold but an ECU reset seems to have cured that.

 

Not really sure how I can check for sparks on account of where the plugs are, hidden away like they are.

 

I'm quite happy to do my own mechanics as I'm pretty handy with the spanners but I've not needed to do any work on the scoob before so the engine bay looks a bit intimidating.

 

My mate suggested the exact same thing that maybe I'd sabotaged a fuel pump wire and that it was running on the last bit of juice in the system, but I checked under the car and could even find any wiring or even see the pump itself. Do I get to it from inside the boot?

We did have a listen and I suspect the pump may be ok as I can hear a slight whirring noise from near the back that goes for maybe 5 seconds when the ignition is turned on, then stops. I would imagine that's the fuel pump priming the system maybe, which makes me think that it's probably ok.

 

Is there any way that I can check in the engine bay if fuel is actually making it to the engine?

Likewise any tips on how I can check for sparks without the aggro of removing the plugs, and where exactly the dizzy is hidden in the engine bay.

 

Anything like that would be most useful, or any kind of links to how to's for common jobs like that on a scoob.

 

All the fuses in the driver footwell box were fine, and all the ones by the battery were ok as well but I've no idea which would be the fuse for the fuel pump anyway as everything's in japanese and I've no kind of haynes book or anything like that.

 

I'm really grateful for any ideas as I'm a bit stuck without the car. More than that I'm just really gutted that something like this has happened just after I've spent a load of money on the car to get it really nice.

 

Cheers!

[frag]

I have no pratical knowledge of eclectrical welding (only have an oxy-acet system.) and volonteer this only because you seem to be out of fresh ideas.

Is it possible that using an electrical soldering system (if it's what you used) on the car without grounding it could play havoc on the lectronics of the car?

Just a wild idea... :-\

[Cret]

Thanks for the suggestion but I can rule that out straight away as the welding was done with exhaust off the car not on it (didn't fancy doing welding near the fuel tank!!).

[theotherskip]

you should be able to get to the top of the gas(petrol) tank through access panels in the trunk(boot). they should be under the carpet behind the rear passenger seats and before the spare tire(tyre) well. as for which fuse, be sure to check all of the fuses under the hood as well as the ones in the passenger compartment. remove them and check them with a continuity meter, as it is not always apparent if they have blown. alternately, you can remove one of the fuel lines from the fuel filter, hold it in a glass jar, then turn on the ignition and see if you get fuel out of the line, but i still think that your problem lies with the fuel delivery, not ignition system, especally since you were working on the rear of the car...

 

good luck!

[Cret]

Thanks very much the advice. I really hope you're right as it will hopefully then mean it's going to be a simple & not too costly fix if any parts do need replacing. If it's a matter of a wiring fault then I can easily do that myself, so I think I'll be heading down there tomorrow with my multimeter.....

 

Whereabout in the engine bay do the fuel lines come in/the fuel filter be found? I'm sure I'll spot it but it's just that bit easier if I know where to head straight away.

 

Just remembered something else my mate mentioned as well. When I was cranking the engine over he stuck his face up the backbox and said he couldn't smell any fuel through the exhaust, so that's another indicator that the fault is indeed fuelling & not ignition.

 

Here's hoping.....

[theotherskip]

i don't know how much is mirrored for a rhd, but probably on the above the left fender you will find a round can with 2 rubbers hoses connected to it, one going in, the other out...

[adge_082]

to check for spark on my old subie (not at the plugs though)i just stuck a screw driver (wear heavy gloves) in the lead and held it a couple of millimetres above something metal that doesnt mind a bit of electricity. If you see pretty lights, you have power to your plugs.

 

Then check plugs i guess. If your sure you have fuel.

[alias20035]

I think all possible items have been covered, but I will summarize for informational purposes.

 

I used to say that three things are required for an engine to run, I rethought the process and came up with a fourth item:

 

1. Air

2. Fuel

3. Spark

4. Ability to eject exhaust

 

You first item to check in all cases should be the timing belt alignment and condition. Remove the left and right timing belt covers, rotate the crank to TDC (notch in crank accesory pulley aligned to the zero mark on the center timing cover), and note the location of the timing marks on the camshaft pulleys. The camshaft pulleys should read 12 o clock on the timing mark, or 6 o clock, 180 degrees out from the timing mark, in which case you need to rotate the crankshaft one complete rotation back the the zero mark and the cam shafts should now be at 12 o clock.

 

I always check the timing belt because it is easy to do (6 or 7 timing cover bolts only), and it eliminates the possibility of doing further damage to the valve train if the belt has in fact slipped or broken.

 

Air:

 

If the Idle Air Bypass (IAC) failed or is not connected properly (both electrical and its air plumbing), the engine will not run. You can try to start the engine with the throttle partially open, if it starts then the IAC is a possible cause.

 

A failed Mass Air Flow (MAF) or Manifold Air Pressure (MAP) sensor can sometimes prevent an engine from running, but the ECU is supposed to enter "Limp Home" mode and fall back on predefined Air/Fuel tables and use the oxygen sensor to compensate. If the ECU does not enter "Limp" mode as it should, you can force it to by simply unplugging the MAF/MAP sensor.

 

Fuel:

 

You have heard the fuel pump prime the lines when the key is turned to "On". This is the first item to note. The pump should prime for between 1-5 seconds and then turn off, and if the key is left in the on position without starting the engine, there should be a slow and continous intermittant pulse from the pump as it is maintaining the fuel pressure (fuel is returning to the tank all the time).

 

Next thing to do is to connect a fuel pressure tester with a T connector in place of the fuel filter. The fuel pressure should be between 30 and 60 PSI (I think 47psi is the official spec?). No pressure = clogged line or incorrectly connected hoses.

 

I have never seen a fuel filter clog completely nor prevent an engine from running. Icing up yes, clogging completely no.

 

Note that the fuel pickup is on the right side of the tank and a venturi pump using the fuel return pressure is used to pump gas from the left side to the right over the axle hump. If something is wrong with the gas tank plumbing this "pump over" may not happen, and there may be no gas on the right side to be picked up, although in this case the fuel pump will usually try to prime the fuel lines for a very long time without stopping. With this condition your fuel gauge will still show some gas in the tank as there are two fuel level floats and the two readings are averaged for the gauge.

 

Does your "Check Engine" light come on when you turn the key to "on"? If not the ECU may not be working. Engines with failed ECU's will crank forever but never start as the fuel injectors are not working. One way to test the ECU is to do the code download process (causing the check engine light to blink the condition codes), if the ECU enters this mode and indicates ok or a problem, the ECU is at least working.

 

Spark:

 

Coilpacks do fail, but usually it is one cylinder bank not both and it is usually a mid to high RPM misfire. If you have a timing light you can connect its "high tension trigger" lead to each spark plug wire to test for spark at each plug (or at least the fact that the coilpack is trying to spark each one). The timing light will flash if the spark plugs are firing. Using a timing light is the safest method, but the screwdriver/thick glove technique will work, although there is enough current to stop your heart and also destroy electronic components, so please be careful.

 

If no spark, check to see if the coil pack is receiving power when the key is in the "on" position. If so, then swap the coil pack temporarily and if the no spark problem still remains, the problem lies with the ECU, the crank position sensor, the cam angle sensor or any associated wiring. This is where it can be a pain in the rump roast if the ECU doesn't tell you what specifically is wrong.

 

Ability to eject exhaust:

 

If the exhaust is plugged (usually a plugged catalytic converter) the engine may run poorly with little power, or may not run at all. Often the engine will loose all power and once the engine is stopped it can not be started again.

 

If you your engine is running badly due to a clogged catalytic converter, the catalytic converter will usually glow red (most noticable in the dark).

 

You can disconnect the Y pipes from the underside of the head and try to start the engine, if it starts shut it off immediately, running an engine without an exhaust system attached will destroy the exhaust valves (but not in the brief 1-3 second long period to test for a clogged exhaust).

 

If the engine can not eject exhaust it can also not pull in the air fuel mixture.

[Tiny Clark]

This has been said by myself and others on this board, and I will repeat it once more.

 

The easiets way to check for spark is to use an extra spark plug, new or old, doesn't even have to be for your subie.

 

Find a good piece of metal (ground) on the motor and clamp the plug to it with a pair of vise grips.

 

Pull one (or all) of your plug wires off and connect them to the test plug. If you have spark, then it's a fuel/air problem.

 

Tiny

[Cret]

Thanks guys (especially Alias for that big long post!)

 

I'll be heading down there after work this evening to see what I can test and rule out now that I've a better idea how to approach it.

 

I'll let you know what I find out - fingers crossed it's just the pump or something.

 

I can't see that I'd have managed any kind of catastrophic engine failure just idling it like that out of the blue...........

[Supaglu]

Hi,

Sounds to me that you have just flooded the engine with fuel, under the dash - drivers side you should see two black single pin connectors - connect these together and this will stop the fuel system from over injecting when starting, put the gas pedal flat to the floor and crank the engine over until it fires, let the engine warm up and then disconnect the two black connectors.

 

These connectors are used when the vehicles are new in the showrooms so that they can be started and stopped continuously when moving them around.

You can also check the operation of your fuel pump by connecting the two green single pin connectors together and switching on the ignition - the fuel pump should then pulse on and off.

 

Hope this helps !

Dave H

[Cret]

Thanks for all the help so far. Sadly, things aren't looking good.....

 

Well folks, it's time for depression factor 11 Mr Sulu.

 

The good news is that my fuel pump is just tickety boo. This at least gives me the tiny bit of satisfaction that I was telling my mate it sounded like it was fine all along. I like being right.

 

The bad news is that my timing belt has slipped. I don't know how much, but it's not hopeful as nice things like it only being a tooth or two out are things that happen to other people.

 

I guess now my only option is to get a new belt (anyone know what they cost & a good place to order one? I know they aren't cheap), time it all up properly with the new belt on & see if it's been lucky enough for the pistons & valves to have avoided that most unholiest of unions that I'm dreading/expecting, or if the engine is now a cambelt murder victim. :(

 

If anyone has any cunning suggestions on other ways of finding out if it's wrecked before I fork out for a new & potentially useless cambelt I'd be obliged as I'm hardly flush at the moment having just spent loads of money on getting the car all shipshape and treating it to lots of goodies.

 

Merry ****ing Christmas to me - ho ho ho.

The best bit of all is that I just finished paying my mum back last month the money I borrowed off her to buy the treacherous chunk of crap.

 

Well, I'm off to get drunk now, and sulk. A lot.

Anyone know what a second hand scoob engine goes for roughly? Time to get saving. :|

[tcspeer]

How do you know timing belt has slipped? just seems strange the way it happen.

[theotherskip]

i agree with tcspeer. what are you checking the timing marks against? it seems very unlikely that you were able to drive the car in, then were working on the rear end of the car and it won't start now. slipping a timing belt at the same time would be very unlikely...

[1 Lucky Texan]

I think if you have a bad crank and/or cam sensor you won't get spark. Maybe one ha sbeen bad for some time and now the other is bad?

 

justa wild guess.

[alias20035]

Originally posted by 1 Lucky Texan

I think if you have a bad crank and/or cam sensor you won't get spark. Maybe one ha sbeen bad for some time and now the other is bad?

 

justa wild guess.

 

The ECU is supposed to be able to fire the spark plugs with either the cam or crank sensor not working, but not both. This is according to the Subaru technical manual. I can confirm that a failed cam sensor will not prevent starting, but I never been able to start a Subaru engine with the crank sensor disconnected, so perhaps the technical manual is incorrect, or perhaps the engine will simply continue to run with a failed crank sensor, but not restart.

 

However, if the ECU detects incorrect timing between crank and cam, it will not fire or inject fuel as a safety precaution. I think the threshold for misalignment is about 2 or 3 cam belt teeth, but on the DOHC models it could be even lower.

 

Timing belts do not slip on their own, so you have to look for the cause if it turns out that you were reading the timing marks correctly and it has in fact slipped. Not sure why the belt would pick this particular time to slip though, as your work was unrelated. There are usually two marks on the crank timing belt pulley, if that is where you are verifying the timing, the mark to use is a small arrow on one of the crank angle sensor teeth, not the one on the front of the pulley which is 90 degrees out from the tooth mark. Also check alignment between the exhaust and intake cam pulley, usually these are properly aligned until the cam belt tensioner fails.

 

Inspect the cam belt tensioner, idler pulleys, cam and crank pulleys, water pump, etc. The cam belt tensioner can be tested with a hydraulic press and a pressure sensor, I don't have the readings for the compression pressure required, and the return speed of the piston though, the newer ones are 66PSI, but I think the old style tensioners are much higher. The older Subaru turbo's are known to eat a few tensioners, so I would go on the asumption that it has failed. Idler pulleys should have no play and spin smooth and quiet, as should the water pump.

 

I would replace the cam and crank oil seals when doing a timing belt as well as the water pump if it has 80,000km or more on it.

 

Even if you find no obvious signs of belt damage, change it with a new one, it is not worth the risk. Your engine is an interference type and will be killed by a broken timing belt.

 

In order to check for internal engine damage you need to either take the heads off for inspection, or install a new timing belt, tensioner and whatever else is needed and do the standard cylinder compression test.

 

It might be possible to do a leak down test with the timing belt off, but you will have to align and lock each pistion and its associated intake/exhaust valve in the closed position. This would check the cylinders ability to hold pressure (all cylinders leak, but should leak slowly). A leak would likely indicate a damaged valve in your particular case. This method is a pain in the @ss to do and too dangerous to attempt with the DOHC heads, but it is the only method that can avoid an outlay of cash.

 

With the timing belt off you have to be extremely careful to keep both camshafts on each head properly aligned. Subaru sells a special tool to lock the camshafts, and there are many aftermarket tools as well. If the two cams are allow to spin free, the intake and exhaust valves can strike each other and cause damage. On Subaru DOHC's the intake and exhaust valves interfere with each other, and this is referred to as an "interference" engine. The SOHC models are "non-interference" or "free-running".

 

In Canada, Subaru timing belts vary from $80-$140 depending on model, and the tensioner is between $80 and $220, again depending on model. The water pump is $70-90, idler pulleys are $40-60 and the cam and crank seals are $4-10 each. But Subaru parts costs seem to be highly variable based on things like the weather or some other unexplainable reason.

[Cret]

Thanks for the responses guys but the belt definitely has slipped.

 

When I took the covers off it last night the belt was in one piece but in a right state. I'm not sure if the tensioner had failed maybe but the belt was slack in some points and tight in others.

 

Turning the crank by hand it was very obvious to see it slipping quite badly so I' don't hold out much hope for the engine at the moment.

 

First thing I'll do is fit a new belt, line it all up and see if it starts. There's a small chance it might not have damaged itself but I think it most likely has.

Presumably this means at the very least bent valves, and possibly holed pistons, cracked valve guides etc etc.

 

If it doesn't run with the new belt I imagine I'll take the whole engine out rather than just removing the heads, and I'll then have to decide whether to rebuild it or find a replacement engine.

 

A dark hour for me indeed though.:boohoo:

[Meeky Moose]

sorry to hear about your troubles right at x-mas time.. but i feel your pain. my RX bloew a headgasket/cracked a head saturday :boohoo:

[reeze]

So sorry to hear about your engine trouble.

 

For the edification of a new owner of a '95 Legacy L (2.2 engine), could someone confirm whether these are interfering or non? I've owned a 91 and 92 and always thought they were non-interfering.

 

There's no year or engine model mentioned in this thread (or maybe I missed it?) but from what I've read it seems like this particular model DOES have an interfering engine.

 

This has big inplications for me since my car had almost exactly 60K miles when I got it, and I'm quite sure the belt hasn't been changed.

[Cret]

Can't answer you I'm afraid. I've had a couple of the Aus Liberty RS owners saying they've had belts slip and been ok so does that mean the engine is probably non interfering or that they were just extremely lucky?

 

I take it that interfering means that the valves & pistons WILL hit if the belt goes, and non interfering meaning that they clear each other?

 

If that's the case I'd naturally very much like to know which one it is! It's an EJ20G engine by the way.

 

Have a new cambelt on the way now so I'll know one way or the other before too long......

[Supaglu]

Hi,

If it's a quad cam then chances are very high that valves and pistons have collided 99.9%, if its a twin cam then it's a "safe" engine - no valve damage.

 

Regards

Dave H

[Cret]

Yep it's a 4 cam. Bah! As I expected.