Trying to diagnose 98 LGT vibration issue.

[Zoo Mob]

Just got my 1998 Legacy 2.5 GT wagon on the road after a auto to manual swap, at the time, heads came off and were resurfaced, all gaskets replaced, new plugs and wires etc. etc. The car runs and shifts pretty well, but I know its just not where it should be. It idles very smooth, but once the car gets past 2000k, it begins to have some vibration issues. The vibration is not that bad, but I've got a friend who has the same car and it much smoother running throughout the powerband. The check engine light is not on. The vibration happens even if the car is not in motion, so any driveline issues can be eliminated.

Possible thoughts thus far are:

 

it may need a valve adjustment; there have been some who suggested at 140k it could be necessary, though there is no apparent valve ticking etc.

 

crank pulley is wearing out and could cause the engine to be out of balance

 

motor mounts?

 

Not really sure what else to consider or how to proceed from here, and thoughts or suggestions would be welcome.

[nipper]

IMMEDIATLY check the main pulley/vibration damper. It sounds like the engine was apart before this. This is a not-so-uncommon error in assembly where main pulley is not torqued high enough. Double cjeck that first and we will go from there.

 

Inspect the pully for seperation also.

[Fairtax4me]

Double check your cam timing. I'd be willing to bet one is off by a tooth.

[SubieDaddy80]

Just throwing this out there, but possibly a flywheel issue?

[Zoo Mob]

I'm pretty sure the crank pulley was properly torqued down and it is not off by a tooth, but I will verify both. The flywheel was resurfaced and it has a brand new clutch. Seems pretty unlikely, but could it have been resurfaced improperly?

[nipper]

Does it do the vibration in neutral?

[Zoo Mob]

It is a has a manual transmission and will vibrate if the rpms are brought up in neutral while standing still.

Edited July 2, 2010 by Zoo Mob

[logic23]

Could be an injector that has a bad spray pattern but if it ran smoothly pre swap I would go after things that you touched> motor mounts>pulley>timing belt issues. It could be a plug wire that is shorting out. Ive seen these damaged when the engine is taken out or installed. They get pinched and the shielding is compromised. You can check this in the dark> look for the spark jumping out of the wire to the block. Be careful not to get your hands close to it....its an awakening experience to be shocked by a plug wire.

You need to find out if the shake is there at all rpms.

Put you hand on the engine and feel it running at idle.It shoud be smooth no jiggle. Grab the throttle and slowly raise the idle keeping your other hand on the engine. You should be able to feel the point where it "cuts out". Let us know what you find.

[nipper]

A vacume gauge would help too.

[Zoo Mob]

Well, spent some time, tried a few things and got nowhere tonight. The crank pulley was removed and checked for separation, there did not seem to be any, but we swapped on another with no apparent effect. Next, visually inspected the plug wires which looked fine, then we turned off the lights and looked for arcing and there was none. Lastly, we swapped the coli pack for another and it made no difference.

 

We did a little experimenting with trying to see where the shake begins and ends; it seems to come on around 1750rpms, is the worst around 2k, and might get a little better towards 3k, it was not revved past 3500 and was still noticeable at that level.

 

This leaves valve shimming as a possibility, but as mentioned, there is no noticeable valve noise present.

 

Injectors also seem a possibility, but as I recall, the shake was not present or at least I did not notice it prior to the auto-manual swap. If we do swap the injectors out, the plan was to swap them two at a time and leave them connected to the rail. Does this seem the right way to do it to you guys?

 

Next time we tackle this, the timing covers will come off for inspection, but my friend who did the work is pretty certain the belt is on correctly.

 

Is there anything else we are missing here that you guys can think of?

 

Also, how would we incorporate a vacuum gauge into the diagnosis as was suggested?

 

Many thanks!

[Fairtax4me]

A fluctuating vacuum gauge or low reading would indicate an improper valve seal, incorrect valve timing causing compression to bleed out of a valve that isn't closing as soon as it should, or a vacuum leak.

Just hook it up to intake manifold vacuum. There is usually a perfect port for it sticking straight up (towards the hood) about an inch in front of the throttle body. Disconnect the evap hose and hook up your gauge there.

[Zoo Mob]

Thanks, I'll give that a shot.

[Zoo Mob]

Well, finally got around to taking the timing cover off and triple-checking that everything was lined up correctly.....it was. Haven't checked the vacuum stuff yet, any other thoughts?

[nipper]

WHere do you feelthe vibration. What happens when you pop the car in neutral at speed (taking your foot off the gas pedal of course). Is it a constant vibration or does it come and go.

[snowscooby]

If you suspect a vacuum leak, as I did, you can get a can of ether and spray it around all your hose connections and gaskets. I had a rough idle and my engine would shake back and forth. Did the ether spray test and found that I had to gaskets leaking. When you spray the ether, if there is a leak, the engine idle should spike a little because the mixture becomes ritcher.

[Setright]

Just a heads-up on the crank bolt:

 

Dip the entire bolt (thread and shoulder) into fresh motor oil and torque it down to 140Nm in one full swing.

 

The oil on the thread ensures that torque comes from the tension in the length of the bolt and not friction in the bolt thread. The bolt must clamp the crank pulley hard onto the crankshaft end.

[nipper]

Just a heads-up on the crank bolt:

 

Dip the entire bolt (thread and shoulder) into fresh motor oil and torque it down to 140Nm in one full swing.

 

The oil on the thread ensures that torque comes from the tension in the length of the bolt and not friction in the bolt thread. The bolt must clamp the crank pulley hard onto the crankshaft end.

 

http://forums.viperclub.org/rt-10-gts-discussions/615323-anti-seize-torque-setting-follow-up.html

 

You are overtorquing the bolt.

[Setright]

Nipper, I don't disagree with the link, but the Service manual stipulates a lubricated bolt.

 

So many crankbolts have worked lose because they have been torqued too low.

[nipper]

Nipper, I don't disagree with the link, but the Service manual stipulates a lubricated bolt.

 

So many crankbolts have worked lose because they have been torqued too low.

 

 

Oh I agree.

[Zoo Mob]

Motor is out as we and the professional shop we took it too suspect the flywheel is the cause. A few questions; the flywheel was a from a '98 OB, and this car is a '98 GT any chance we have the wrong flywheel? Could the machine shop have resurfaced it poorly and caused it to be out of balance? How can this be tested, and how would you recommend I proceed?

[nipper]

Subaru flywheels really do not like being resurfaced unless it only needs a really fine cleaning up.

 

May want to replace it.

[Zoo Mob]

Took the flywheel and clutch to another shop to check balance, and they were seemingly ok. The flywheel by itself was only 5g out, and was fixed. With the clutch attached, it was only 2g out, so I'm thinking this is not the issue.

 

Correct me if I'm wrong, but these engines are internally balanced correct?

 

There was a thought that there could be a fault with the pressure plate, but that was not something they could get on their balancer. My friend suggested that bolting that back on in a different orientation might help.

 

Leakdown test has not been completed yet, but we're really freaking stumped here. HELP!

[nipper]

All engines are internally balanced.

 

Subarus more so as far as wear is concened.

 

Is this a new flywheel or a resurfaced flywheel?

 

Have you tried removing the accesorry drive belt then watching the harmonic balancer?

 

Possibly a bad valve adjustment?

 

Have we done the vacume gauge thing yet?

[Zoo Mob]

This is a resurfaced flywheel. I read what people have said about not resurfacing Subaru flywheels, but since it has been checked, and is balanced, what might be the problem there?

 

No vacuum test yet. No leakdown test yet. We've removed the harmonic balancer and checked for separation, but have not inspected it while running if that is what you are suggesting.

[nipper]

A flywheel has several jobs. One is to keep the engine spinning at idle. Another is to help absorb vibration at speed, and the last one is to attach the rest of the driveline to the car.

 

If the flywheel is to thin, the harmonics (vibrations at X rpm) may be changed.

 

4 cylinder have specific RPm points where they "bang" or "Roar" . You can hear this when you stomp on a 4cyl's gas pedal. this is why there is all this wierd snorkus tubing, to try to reduce this noise. if you look 6 cylinders or more do not have this. That Roar can also affecxt engine vibrations if anything is fiddled with.

 

I can not say 100% if this is what it is as I have not driven the car, but after everything else I would suspect this.