rough running after engine reseal

[Marck]

Hello,

I am stumped here :-\ . It runs so rough that when I begin to release the clutch pedal and apply gas the car starts to shake violently like an earthquake. Although it idles rough, it revs very nice and fast. It is also very hard to start (more so when cold) unless the accel pedal is pushed to the floor. There is a squeaking sound but I am unsure if it is related to this problem. I think the squeaking is coming from one of the timing belt tensioners (good thing I ditched the covers). Also another thing that may be related or not is I am getting a code 22 (knock sensor) from the ecu.

 

Here is whats been done:

ported cylinder heads

new oem turbo head gaskets

delta cams (mild grind)

lightened XT6 flywheel and new clutch (haven't adjusted the clutch cable yet)

TWE Header

Stock downpipe and cat

O2 sensor has been replaced right before the car sat two years ago

turbo spider intake manifold swap (all electrical plugs have been cleaned out and checked for corrosion)

fuel injectors have been cleaned by RC Engineering

new fuel filter

new spark plugs and wires

new oem dirtibutor cap and rotor

old gas has been pumped out of the gas tank and a couple gallons of fresh 91 octane gas has been added

look at my signature for anthing I forgot to list

 

My thoughts:

Will the delta cams make the car run this rough? I tried to adjust the TPS while the engine was on to see if it had any effect and it doesn't. I also tried turning the distributor 180 degrees. I have been troubleshooting the spark and fuel for a long time and replaced a lot of stuff. Could it be I installed the timing belts incorrectly or could it be some stupid sensor on the spider intake manifold is not working?

 

What are your thoughts? What should I check next? This is the first time I attempted a project this big so it maybe or not something trivial that I forgot to do. Also note that the car ran fine (with cracked heads and crossover pipe) back in March.

[Adam N.D.J.]

Hmm, an earthquake huh. Start with the simple stuff, if you had the engine out, make sure that the motor mounts are good and tight, been a couple times that I've put the nuts on, but got sidetracked and forgot to tighten them. Then work your way up from there. Make sure the timing is on, both cam and ignition. Make sure all the plugs are fireing right, no crosssparking, check fuel pressure and injector operation. Sounds like you did quite a bit to the engine, so just double check that everything on it is right, the answer most of the time is the most simplest though. Lotsa luck.

[idosubaru]

the knock sensor code is definitley telling you something. if that sensor is cracked up it could be retarding the timing.... and definitely will cause the engine to run terribly. i'd find out why that code is flashing....check the sensor first make sure it didn't crack of fall apart. if it's the original it's probably old and brittle. check the wiring just prior to the knock sensor harness and the harness/connectors as well.

 

how did you address the HLA's? did they sit for awhile? any tapping on first start up? subarus can run terrible immediately after a rebuild/reseal due to the HLA's not quite functioning right until they are fully pumped up and lubricated. and that's assuming they are working appropriately.

[Marck]

The motor mounts are torqued to 25 foot pounds. Whats crosssparking and how do you check for that? How do you check fuel pressure? I placed a steciscope (long screw driver) and heard each fuel injector clicking.

 

I am curious about the knock sensor too. How do I check the knock sensor electronically? I have an el cheapo digital multimeter that I can use.

 

The HLA are tapping, but I am not worried about it. I think it will go away after awhile once the car is back on the road. I did take the valve covers off to re-torque the cylinder heads and everything looked in order with the HLA and cams.

[nvexplorer]

You might also want to take a look at the air intake setup. If there's a crack in a hose or fitting after the MAF or something's not connected it'll idle rough but drive fine at higher RPM's. Hasn't happened to my Soob but did happen to the 89 camry with the V6 I used to drive. Ran excellent once I replaced the air intake hose.

[MilesFox]

maybe a crossed plug wire or the timing belt is off

[idosubaru]

HLA's can affect how the engine runs. any backfiring or other noises?

 

lots of questions can narrow things down - why did you do this reseal? headgasket issues or anything else? just oil leaks?

 

if you didn't have the knock sensor code before the reseal and have the code now, i'd check into that before checking zillions of other options. start with what you know before making random stabs.

 

first i'd pull the connector to the knock sensor. check the sensor out, make sure the plastic isn't cracked or falling apart on the top of the sensor. then inspect the plug that plugs into the knock sensor and see if it looks clean and in good order. then check the wiring just prior to the plug and make sure there is no exposed or kinked wire. with the engine running you can even grab the wire between two fingers and move it around a little at the plug and just prior to the plug, this will make the engine change sounds and idle if there's a short in the wire.

[idosubaru]

definitely double check your timing belt marks and make sure they are dead on and also the distributor timing has to be perfect as well. the mechanical distributor timing first - then the elctrical timing of the distributor. distributor sounds set correctly if the car will actually run and drive. have you set it with at timing light yet?

 

and about the delta cams, i just installed a set and they idle just fine so far. no adjustments needed. myphalx and i installed a set on his XT6 a year or so ago and had issues with the car idling. not sure if it was cam related or not though.

[Marck]

I just rechecked the timing belts and they are dead on. BTW, I did not use the three marks on the XT6 flywheel to set the timing belts because I know those are off. I engraved into the XT6 flywheel a new set of marks so that it would match the stock flywheel. I also tried running the engine while using a screwdriver to jiggle the knock sensor wire. I did not have any trouble codes prior to the engine reseal. I resealed the engine because I had a crack in the exhaust outlet of the turbo side cylinder head. I also used this opportunity to replace every single seal and do many mods. There are no backfiring noises, but there is a lot of lifter tapping or knocking.

 

Now about distributor timing, I did use a timing light to set the timing to 20 degrees BTDC. That is stock for this car. But what I found is that the engine runs much smoother (like the way is is supposed to run) when I advance it to about 40+ degrees BTDC.

 

Test Drive:

It was hard to drive, but I managed to drive the car down the street and back. It blew a lot of white smoke with maybe a little blue and gray smoke out of the tail pipe. There were also smoke coming form the turbo. I think the smoke is mostly condensation in the exhaust pipe burning off along with engine degresser and WD-40. The Turbo spooled very good (starting at 2000 RPM). It also felt like the engine was holding back a lot of the power.

[Marck]

I also thought that I would mention that there is oil pressure at idle and that I am using all emissions equipment that came with the spider intake manifold (ie. EGR, PCV, AAV). BTW, the AAV does not seem to work at all. The engine harness had the same plugs as the car so there was no rewiring involved.

 

Another thing I noticed that has stumped me:

I poured 4 quarts in this engine to begin with which was just below the full mark on the dipstick. Now I am reading way over the full mark and feels kind of thin. Where did the extra oil come from? Did gas get into the oil? The oil looks like clean oil and not like a milkshake, so I know that there is no water or coolant in there.

[ShawnW]

Tooth off on the distributor Id say. Then re-adjust timing.

[idosubaru]

i'm with shawn, i have seen a distributor be off and the car be driveable. reset your distributor.

[Marck]

I just pulled the knock sensor connector off and took a good look at it. There is no corrosion in the connector or on the knock sensor and no broken plastic. I used my FSM to test the connector and the sensor. It says to measure the resistance between the inside connection of the knock sensor and the body of the knock sensor and make sure that resistance value is in the given range of the FSM. Then it says to do the same with the connector. It passed both cases, so my knock sensor must be good. Then why the bogus ecu reading? I tried reseting the code using the directions in the FSM, but it doesn't work.

 

About the distributor, do I just take of the two bolts and pull the distributor off the cam hole and turn the distributor one tooth clockwise or counterclockwise? Because I already tried that. I go ahead and try that again though. Is 11:03pm here and I am not giveing up yet .

[idosubaru]

i've found the distributor really annoying to line up after having the engine apart. if you set it using the FSM - it says to bring the #1 cylidner to TDC..that's easy enough. then set the distributor pointing to the #1 cylinder on the distributor cap. well that's the tricky part....there are multiple positions where the rotor can point at cyilnder #1 depending where in the slots the distributor is bolted down. stupid instructions in my oppinion. in other words, if you're following the FSM it's very easy to be off because the directions are ambiguous at that point. i have a bunch of XT6's lying around and find the easiest method of lining up the distributor to be making it exactly like the others! if someone could post a picture of the distributor rotor position when the #1 cylinder is at TDC you'd be golden. then you'd know which way you need to go with it.

[Adam N.D.J.]

Is your cap and rotor new? And if they aren't are they actually in good shape, and is the bolt in the rotor tight? I know that a loose rotor bolt can cause a lot of problems. Also did the you set the timing with the D-check connectors connected?? This can also cause some problems if it isn't.

 

Hmm, definatly sounds like you have a pretty good problem going on, and it sure is hard to work on a car when your not with it. But we'll all try to help you out.

 

As for the knock sensor, it may not be the problem. If you are having a problem thats making the engine shake it's self to death then it can actually make the knock sensor over active, then the computer will retard the timing as far as it can and still make it run, then it will set a code when that doesn't stop it. I've never heard of a resisitance check on a knock sensor, I don't see how it would work at all. Most knock sensors that I know of (including subie's) are piezo electric generators, so here is how you should test it, hook up your multimeter to both leads on the knock sensor, put it on AC voltage, the lowest voltage setting if it's adjustable, then tap lightly on the block near the sensor with a hammer handle, every time you do this, you should see a voltage spike, if you do not, then there is a problem with the knock sensor. Hope this helps you out, let us know if you come up with anything else.

[idosubaru]

forgot to comment on the knock sensor. not sure, but something must be wrong with it if it's showing a code on your ECU. you said you tested the wiring harness side connections...if not i'd say check the plug and make sure one of the pins isn't bent or something wrong with it and make sure the engine harness plugs are seated tight.

 

years ago i had an errant knock sensor problem on an XT6, a new sensor did nothing for the problem. i pulled the sensor and wrapped the threads about 20 times with plumbing tape and the car ran great. have no idea why and haven't seen that problem duplicated since so i doubt it'll work for you, but just throwing it out there.

 

the ECU uses the knock sensor to adjust timing so i still think this needs to be addressed if you're getting the knock sensor code. if the timing is being adjusted by the ECU because of this problem then setting your timing might not work so well.

[Adam N.D.J.]

Thats what D-check mode is for, when you plug in the connectors under the dash (or is it in the engine compartment by the fire wall for the Gen III) either way, you all know the ones I'm talking about. When you plug them in, it shuts up the computer, so that it does not control the timing or fuel mixture anymore, basically the computer goes into UA, or uncalculated algorithm, or as most now it, limp in mode. This is when you set the timing, while the computer doesn't control it, then you disconnect the plugs, and if your still watching, the computer will then move the timing to where it wants it.

 

Hey Marck, that is actually a pretty good idea. You should plug in the connectors, set the timing to base, then while still watching with the timing light, disconnect them, and watch what happens to the timing. If it goes to way retarded, then you know that there is definatly something wrong with the knock sensor.

[Marck]

I tried unplugging the green test connectors as I was checking the ignition timing with a timing light and it turns out that the green connectors being plugged in or not has no effect on the timing.

 

I tried turning the distributor to many different places and it just runs worse than it did before. Then I made the mistake of leaving the breaker bar on the crankshaft bolt while trying to start the engine. I am lucky the distributor was still off so the car wouldn't start. That would have been ugly. Of course the enigne still turned over though, so the breaker bar hit the oil pressure sender unit and broke it and loosened the crankshaft bolt.

 

Since this happened, I decided to take off the v-belts and timing belts off and investigate the squeaking noise. I turned both tensioners, timing idler, oil pump sprocket, alternator, ac compressor, power steering, water pump, and the v-belt idler. Everthing turned smoothly and nothing squeaked. So where else can this squeaking noise come from? Could it be one of the cams or the throwout bearing? And yes, I did replace the throwout bearing.

[idosubaru]

hard to say, sometimes the bearings may not be noisey unless they are under load (have belt tension on them). seems like you should notice it though.

 

if it always squeaks you can start the car without the accessory belt and run it for 20 seconds to see if it squeaks still. this would narrow it down considerably to timing belt pulleys or drive belt pulleys.

 

have you done a compression check?

[Adam N.D.J.]

The green connectors are your memory connectors, they are for checking the codes in the computer. The D-check connectors, or test connectors as they are sometimes called, are the other set. I do believe they are black, but can't specifically remember, I know they are on the Brat. Also you have to make sure that your TPS is adjusted correctly. If the idle switch in the TPS isn't closed, then the computer won't go into test mode, and you can't check the timing.

I agree with the run the engine without the accessory belts, that will narrow things down a bit. Your cams and cranks should not squeak, if they do, then you have a pretty bad problem (i.e. no lubrications). I would go after the T-belt tensioners though, they do have a pretty rough life, and those bearings only last about as long as the belts do. A typical belt job for me includes replacing the idlers and tensioners.

Don't worry too much about the breaker bar. The worst thing about that is that you broke the sending unit. In fact, this is how many of us loosen up the bolt on our rigs (least I do when I don't have my air tools).

This is a pretty tough problem, lots of things that got changed at one time, a whole lot of variables. But we should really start with the timing. With these optically controlled trigger fuel injection systems, a lot of problems can be had by the dizzy being off by even the slightest. We will get it fixed for ya though.

[Marck]

Are you sure? I am pretty sure the green ones are the test connectors and the black ones are the memory connectors. Can anyone else verify this?

 

I did run a compression check before I started this thread and the compression ranges from 125 to 140. I think thats fine, because that is how it was before I did the reseal.

 

After I pulled off both timing belts and reinstalled them, I went ahead and ran the car without the accessory belts. I couldn't believe it. Not only was the squeaking sound gone, but the car ran so quiet and so smooth. It was almost perfect. Then after I reinstalled the accessory belts, it when back to the way it was before I removed the timing belts. So then I started to move the wiring around distributor, battery, fusible links, AC compressor, and alternator while the enigne was running. When I was doing this, my hand or arm rubbed up againts the alternator body. I noticed that the alternator was kind of hot. I figured that the car ran so good when the accessory belts were disconnected, I should see what happens when I disconnect the alternator electically. So I shut the car off, disconnected the battery, and unpluged the alternator. Then I ran the enigne off the battery. It ran so much smoother than it is was before. Not as good as when the accessory belts were off, but the RPMs didn't fluctuate like it did before, the car was not vibrating as bad, and I think the squeaking was gone too. I repeated the process of connecting the alternator and disconnecting it over and over again to make sure that the problem lies in the alternator.

 

I feel like it is an 80% chance it is the alternator causing all the problems. It may also be the reason the check engine light is on too. I have one of those crappy NAPA remanufactured lifetime warrantee alternators that like to break down all the time. So I went down to NAPA to get a free replacement but they didn't have one in stock. They ordered one for me and it should be in on Tuesday. So I let you guys know if that solves my problem(s).

 

Until then, I will try to get the TPS set correctly using the DMM and my FSM. The turbo spider intake manifold came with a 4 pin TPS and my old MPFI turbo intake manifold had the three pin TPS. Does this mean I should use the directions given in the SPFI part of my FSM instead of the MPFI Turbo part? Anyone know if the TPS for the XT Turbo w/ spider man is the same as the one in the SPFI cars? Anybody have a FSM for the 87 plus XT cars that can verify this? Also, should the TPS be adjusted with the battery and/or alternator connected or disconnected and ignition key on or off?

 

Last thing, anyone know if this oil pressure sender will work with the stock gauge:

http://store.summitracing.com/default.asp?target=partdetail.asp&autofilter=1&Ntt=oil+pressure+sender&N=0&part=ATM%2D2242&autoview=sku&Ntk=KeywordSearch

[Marck]

Hello,

I got the alternator today, but it made no difference except my check engine light is now gone. The alternator must have been the reason for the false knock sensor code. Since this didn't improve the problem, I tried the following:

 

-Moved the passenger side T-belt one tooth off

-Moved the driver side T-belt one tooth off

-Moved the passenger side T-belt back to where it was before

-Disconnected the alternator to see how it runs without it plugged in

 

The squeak is still present and acually occurs with the accessory belts on or off. The check engine light is off. The distributor is timed to 20 degrees BTDC. And the thing still vibrates. I haven't adjusted the TPS yet, but it really makes no difference what position it is in when moving it around with the engine running.

 

Let me try to describe what it is doing better. I went under the car while the engine was running and found that the squeaking is louder underneath. I also noticed that the squeak doesn't so much sound like a squeak you would here from a tensioner bearing or something, but more like a chirping like a VW bug. The the car vibrates because the engine is chugging (when idling only). It is not sound bad like an earthquake, but its chugging bad enough to feel it while sitting inside the car. The exhaust coming out of the tail pipe is not a continous flow of exhaust, but a bunch of discrete puffs.

 

Does anybody know what could be the problem? What should I check next.

[Marck]

I noticed a dew other things that are add on this car. When the power steering wheel is turned all the way to one side, the RPMs will drop almost 1000rpm. When the steering is turned just a little, the rpms will drop a few hundred. The RPMs will also drop some when the heater/AC/blower fan is working.

 

This makes sense since the engine runs a whole lot better without the accessory belts. So does this mean that the engine cannot idle underload? Now how do I fix this please? I really need some feedback on my past three posts.

 

I took it for another test drive (1mile) and I found that the temperature will cliimb a little after running it down the street and letting it idle after that.

[hooziewhatsit]

out of morbid curiosity... are you sure each cylinder is firing? After I tested the injectors on my 85 twagon, I managed to make two of the plugs look connected, but not be. This made the engine shake like crazy at idle. However, it still revved up fine until I tried to drive, then it was a turd.

 

easiest way is to carefully pull each plug wire off the disty while it's running. If the idle changes, that cylinder is firing. If the idle doesn't change at all, there's something wrong with that cylinder causing it to not fire.

 

anywho, just a random thought I had

[Marck]

That was a very good suggestion hooziewhatsit. I just now tried that and the car runs worse and I get shocked when I pull out each plug wire except for the number four cylinder. When I start to connect the number four plug wire back on the disrtibutor, I do see the sparks flying from distributor to the plug wire just like each of the other ones. So it is getting spark, but not enough spark to make a complete burn.

 

I am thinking about repeating the test again and then pull the number four spark plug to see whats going on there. Maybe the gap is too big or something.

 

Update: Okay I did the test again and now the number four spark plug does make a difference when pull it off. It gave me a good shock too like the other ones.

 

If the spark plug gaps were too big, would that make it run this way. I am thinking about pulling all four spark plugs and regapping them. I originally set the spark plugs gaps to 0.050 because I am running that MSD Blaster coil.