Lots of cranking, weak spark, and no starting. Help!

[subaruimpossible]

Hi all, I've posted this on nasioc but gotten nothing so far, so I figured i'd post here.

I am almost done replacing the engine in my 2000 legacy outback wagon after an engine failure (metal scraping noises!) I have everything back but the thing won't start I have been troubleshooting all week after work and all day last Saturday but still haven't fixed the problem. 

Here is the full story....

After not getting a start I started doing the basic checks. Fuel, check. Injectors firing, check. Good battery, check. Timing(just did it during install) check. Crankshaft sensor test and replaced, check. 

While checking the spark I found that it was non-existant and or weak and inconsistent so I tested and swapped in a spare coil pack I had from the donor engine...no change. I tried checking power to the coil, that was good and did NOT dip during crank. I checked coil pack signal, I got some pulses but I do not have a good enough oscilloscope to really see what I was getting(OC is from the 50s...)

After this I started reading through the FSM religiously I found some info about the crankshaft sensor which I know drives the timing for spark, which might explain my spark issues. So I went through the troubleshooting steps for that sensor. The sensor checked out via a resistance check, but I replaced it anyway simply to rule it out.

Next, the manual said to check the wiring by unplugging the sensor and checking resistance to ground. Pin 1 (left most pin) is supposed to read between 10ohm and 100K ohm. I tested it at 90K ohm...so a little bit high. 

Pin 2, supposed to be < 10ohm to ground. Tested, 1.5k ohm. So bang out of tolerance. For this the FSM says to 

"In this case, repair the following:
I Open circuit in harness between crankshaft
position sensor and ECM connector
I Poor contact in ECM connector
I Poor contact in coupling connector (B21)"

Going off this I figured it was best to isolate parts of the wiring and test them individually. So I tested E10 to E2 AKA the connector to the sensor that then runs to the gray connector that is on the intake manifold wiring said. This circuit checked out A-OK with both wires showing 0-ohm resistance from both ends. Next, I check the other side so B21 to B135 AKA bulkhead side wiring to ECM wiring(middle connector). This also checked out a-ok. 

After this, I deduced that it must be the connectors, so I connected everything back together, and unplugged both ends, ECM and sensor side (B135 and E10). Then I tested resistance from end to end and across the wires to make sure there was not a short. Again, everything checked out a-ok.

At this point I was beyond stumped, I seemingly had checked everything and it still wasn't working. As soon as I would plug the wire back into the ECM my test results would be bad. I began trying different ground test points, start, battery, intake manifold, ECU housing, intake manifold. All tested the same or within the margin of error. I even checked the same measurements on a 04 legacy outback I have and the readings I got off that working car for the sensor wire to ground measurements were well within spec.

By now I am pretty convinced that it is a ground issue so I look up all the ground points on the ECU as defined by the FSM. The only one that did not test 0 ohm to ground was ground on pin 6 of B135 ECM connector(middle orange for me). I got a reading of about 320ohm to ground which, is not high, but I think higher than it should be. I chased this wire using the FSM to pin 16 on B22(big brown connector engine bay) I tested from ECM to B22 and got about the same reading again. Which when you think about it is actually interesting. Why between two points that are "supposed" to be directly connected is there the same resistance as there is to ground...that's weird! 

I started digging through the FSM again checking the wiring diagrams and finding all the places this wire connect. According to the FSM I have this wire connects to the, fuel gauge sub-module, neutral position switch, vehicle speed sensor, check connector, and data link connector. 

That is where I am now, I'm trying to identify all the places it runs and start testing parts of the circuit. But, I figured it was finally time to ask a question on a forum or two and see what you guys think. Am I missing something super obvious? Any tips for tracking this issue down?

The weird thing about this is that the car ran before I pulled the engine and dropped in a new one. I started the process about 3 months ago, and have been slowly working on it till now. During the swap I made sure to keep my intake manifold, and crankshaft pulley. It was an auto to manual swap so there was actually quite a few parts to switch over...so maybe I missed something? 


ANY help would be truly appreciated, below is a link to the relevant sections FSM if anyone wants to take a look and double check my findings. I am reallllly stumped at this point.

FSM: Look at engine sections for crankshaft sensor testing and Wiring diagram for plug locations and wiring diagrams.

https://drive.google.com/folderview?...UE&usp=sharing


P.S. Hopefully it didnt take you as long to read this as it did for me to write it!

[1 Lucky Texan]

hmmm....just a longshot but - make sure the battery is disconnected and try the resistance test again?

[subaruimpossible]

hmmm....just a longshot but - make sure the battery is disconnected and try the resistance test again?

Tried that...same thing /: But, ill give it another shot since I havn't tested it in that way in a couple days.

[Caboobaroo]

Was one an manual and the other an automatic? They do have different cam and crank sprockets between the two and those will cause your issue. You either have a 6/7 setup or a 36-1-1-1 setup. If you have one set on an engine and trying to use the computer from the other, it won't be able to read the pulse correctly.

[987687]

I was gonna post just that ^^^^^

 

You can swap the computer from the other car... HOWEVER, make sure they use the same style O2 sensor. I did this from an 00 outback into an 02 forester (both manual, but weird optioning makes the forester have an auto-style engine setup). The outback used the older style O2, forester used the newer style O2. Changing the computer actually screwed the sensor up, net result was buying a new sensor and swapping the cam and crank gears.

 

I'd just put the proper cam/crank gears on it.

[subaruimpossible]

Was one an manual and the other an automatic? They do have different cam and crank sprockets between the two and those will cause your issue. You either have a 6/7 setup or a 36-1-1-1 setup. If you have one set on an engine and trying to use the computer from the other, it won't be able to read the pulse correctly.

Yes, donor engine was an automatic and my car is a manual. Donor was also an RS and i'm a legacy. But, I actually used my crankshaft pulley and  sprocket so that shouldn't be an issue. 

 

I was gonna post just that ^^^^^

 

You can swap the computer from the other car... HOWEVER, make sure they use the same style O2 sensor. I did this from an 00 outback into an 02 forester (both manual, but weird optioning makes the forester have an auto-style engine setup). The outback used the older style O2, forester used the newer style O2. Changing the computer actually screwed the sensor up, net result was buying a new sensor and swapping the cam and crank gears.

 

I'd just put the proper cam/crank gears on it.

 

Didn't know about the sensor issue, although I pulled the ECU from another 00 legacy manual so almost the exact same car. So I wouldn't think it should be an issue, but even an o2 sensor shouldnt keep a car from starting right? Plus, the cnrankshaft sensor  wire still tested outside of subaru spec.

 

I was thinking, since the wire is supposed to be connected to ground, why don't I manually connect it to ground? The two pins from the sensor run directly into the ECU, but If I am testing their resistance to ground and one is supposed to be < 10 ohms then surely its directly connected to ground? I am just weery of breaking anything else....

[porcupine73]

 But, I actually used my crankshaft pulley and  sprocket so that shouldn't be an issue. 

 

You may also need the LH camshaft sprocket (see other posts above).

[subaruimpossible]

You may also need the LH camshaft sprocket (see other posts above).

Sorry, should have said it differently. I used my original crankshaft pulley and drivers side cam sprocket.

[Fairtax4me]

You used the orignal crankshaft timing sprocket?

And the original drivers side camshaft timing sprocket?

[subaruimpossible]

You used the orignal crankshaft timing sprocket?

And the original drivers side camshaft timing sprocket?

 

Wait.....timing SPROCKET? I did NOT swap that from my old engine. I swapped the "pulley" the thing all the accessories belts attach to. Now looking at my old engine that is what the sensor gets signal from, that's gotta be what is causing it. Going to work on swapping it out now......will report back with what I find. 

 

This also lines up because the car caught enough to throw a code p0336 aka Crankshaft Position sensor range/performance which sounds exactly like "hey, this is not the sprocket i'm expecting"

[Fairtax4me]

Yep, need the crank timing sprocket. Gotta have the correct reluctor tooth count for the crank sensor.

[subaruimpossible]

Still working on swapping out the sprocket, working on the timing is much more difficult with engine installed! But, im calling it a night working in 80% humidity and 85 degree weather is brutal! 

[987687]

If you have already filled the cooling system, remove the radiator fans. You'll have plenty of access.