Forever cranking, never starting. No spark & ECM not communicating

[ChrisInVT]

I'm at wit's end with my car. I've already posted on LGT & NASIOC with this issue, hoping I could get some insight here. I'm on day 2 without a running car, and even had to take the day off work to try and get his fixed, so I can actually go to work.

 

I brought my car into my shop, changed the oil, went to drive it out - and the damn thing wouldn't start. I have no spark and my scan tool (I've tried three) will not "see" the ECM. As soon as you turn the ignition to ON the cooling fans kick on, the screen on the code reader dims and the LCD goes completely blank. The car will crank endlessly, but will never even attempt to start.

 

I've unplugged and re-plugged every connector I can get my hands on with no change. Also noticed the coolant temp gauge has pinned itself so far past the H it's stuck. 

 

The Check Engine light is coming on when the ignition is turned to ON, and I've swapped the ECM from this car into another Legacy and it started right up so that's not bad. (even tried two additional computers in the car with no change) I even swapped the ignitor for the hell of it. Nothing explains why the cooling fans are coming on and the coder reader is powering off when the ignition key is in the ON position. I've gone over every fuse in the car also and even disconnected the factory alarm from the ignition harness. The green diagnostic plugs are not connected to each other.

 

 I disconnected the three plugs from the engine to the harness, they were clean but I cleaned them real good with electrical spray, waited for it to dry, and plugged them all back in - same issue.

 

I've been out in my shop all morning trying to get this damn thing running with no success.

 

I tore apart the harness from the firewall to the engine connectors and found no problems. I've unplugged every single connector, box and everything I could find under the dash minus the ignition plug. I've unplugged every single sensor and everything I could get my hands on in the engine bay. I've removed every single fuse except the 10A METER SRS LAMP and 15A E/G IGN SRS AIRBAG with the same issue. 

 

As soon as the key goes from ACC to ON the screen on the code reader goes blank and the backlight dims. It won't do this with the ECM unplugged, or the 15A E/G IGN SRS AIRBAG fuse removed. It seems that the fuel pump IS running when in the ON position. Not sure if that's new, or I just didn't hear it before over the sound of the cooling fans (which are now unplugged).

 

So now it seems like there's no spark, and the ECM isn't communicating with the car. If I unplug the bottom connector on the engine harness and plug it back in, I'll see a quick spark in my spark tester - so that should be OK.

 

Now the question is, with the engine unplugged from the harness, every fuse removed from the car minus the two essential ones to make the ECM and gauges work, everything unplugged under the dash - WHY is the code reader dimming and the LCD screen going blank as soon as the car is turned from ACC to ON where it would normally be able to read the ECM?! 

 

Mind you, same issues with two other ECMs plugged into the vehicle.

[Dinky26]

Almost sounds like it could be a battery deal.

 

How old is your battery?, sometimes a battery will die even when new off the shelf.

 

Batteries do wierd things, just saying, and also your connections to the battery are they good?

 

Okay..never mind..it would not crank if the battery were dead.

 

 

Sure does sound like something is not getting the power it should.

 

 

What year and model do you have?

I see what you have now. 98 Outback, cool. I had a 96 loved it to death, then I donated it

 

Have you checked the ignition module?, not sure how one checks it though.

Edited February 25, 2013 by Dinky26

[mikaleda]

It's a 98 outback, I found your post in nasioc

[mikaleda]

Okay I read a little about your problem on nasioc and you said this happened after you changed the oil, you drive it in there fine but after the oil change it wouldn't start. Imho I would check your dash gauge temp sensor and make sure that's not grounding out on something, I don't necessarily think that would make your car not work but it does kinda sound like something somewhere is causing a short. Next I would check all connections going through the firewall and make sure nothing is shorting out there. Other than that I would verify the ECM is getting the correct amount of power

[ChrisInVT]

By ignition module do you mean the ignitor? If so, I've completely replaced it. I can't physically get to the temp senders without disassembling everything on top of the engine and removing the intake manifold. I don't think it's that anyway, because the OBD reader goes dead with the engine electrically disconnected from the main harness. I have the connections at the firewall all torn apart but I haven't found anything yet. My search continues, soon I'm going to be cutting my losses and searching for an OBW with a blown engine.

[mikaleda]

I know this may sound dumb but maybe as a last resort try running some extra chassis grounding maybe it just isn't getting a good enough ground.

[nipper]

Since they were under the car, check to make sure the Engine Temp Sensor is not unplugged.

 

Also a failed crank position sensor. Sensors do fail at the oddest time. A bad crank sensor will keep you from starting.

 

You will never get a code on crank or "on" position, as everything is not functioning at that point.

 

And a helpful hint, do NOT go nuts just swapping things out. You can make problems that were not there, especially if you do not return the car back to its original state.

[ChrisInVT]

I changed the oil myself. I don't think it's the crank sensor, I've never seen a single mention of the crank sensor failing and causing the diagnostics port to die and the cooling fans to turn on when you put the ignition in the ON position. It's just so strange...

 

And you do get your OBD-II codes in the ON position, that's the only time you can get them. Even OBD-I cars the only time you can pull codes is when the ignition is in the ON position (but OBD-I doesn't support that while the car is running like OBD-II does).

Here's some videos...

 

This is with everything plugged back in:

https://www.cubby.com/p/_bcb53300b15649a986b3b4643100bef3/IMG_9425.mov

 

Everything plugged back in and a quick cranking:

https://www.cubby.com/p/_649c919a926c49ee97ff647efa409371/IMG_9427.mov

 

This is an on/off cycle with the temp sensor unplugged, watch the thermostat:

https://www.cubby.com/p/_79d3e3511dec499f96dcd37d8034d16f/IMG_9429.mov

[nipper]

What the ....

 

 

There have been quite a few bad crank sensors that have caused no start symptons.

I suggest you get a FSM or something to diagnose instead of just guessing things.

 

 

http://www.doityourself.com/stry/symptoms-of-a-bad-crankshaft-sensor#b

 

http://www.ultimatesubaru.org/forum/topic/53131-he-says-its-a-crankshaft-sensor/

http://www.ehow.com/about_5860665_bad-crank-sensor-symptoms.html

 

 

 

Also can be a bad cam sensor, both of which have shown up here on this board often.

 

Good luck, as this has seemed to turn into a mess with all the part swapping. General rule, when a car ran then suddenly stopped after being shut off, look at timing belt, cam, crank sensors.

 

Crank sensor is ignition timing, cam sensors.

 

For the record I am an automotive engineer by trade so really you may have made things worse then better by swapping everything out. 

 

And of course no one plugged in the green connectors under the dash, the ones only used for diagnostic purposes at the dealership and reflashing the ECU. When those are connected the car will run but the fans will always be on along with some other oddness.

 

Also logically you wont get trouble codes on crank untill the car starts and runs, it is part of the code written in an ECU, otherwise it would be throwing all sorts of bad codes.

 

Your crank sounds odd. It is not unheard of for a timing belt or chain to jump when a car is shut off or when the throttle is lifted. the timing drive is suddenly unloaded and then there is slack and you can get a tooth jump. If the crank and cam sensors are too far out of synch you wont get spark or fuel as the ECY will not allow raw gas to be pumped into the catalytic converter.

 

Never forget that you still need suck squish boom blow to make a car run. The ECU and OBD is JUST a tool, it is not where the solution to every problem is.

[ChrisInVT]

I don't see how I could've made anything worse by verifying the ignitor and ECM work in another Legacy. If the cam or crank sensor went bad that still doesn't explain why the ECM isn't functioning correctly when the ignition is turned on, and the cooling fans are kicking on high. Timing belt has about 5k on it, I would hope it didn't jump timing, but if it did that still doesn't explain the ECM not communicating. The car has always sounded like that when it cranks, the 2.5 in my 2.5GT sounded the same, as does the 2.2 that's in there now. Must be the phone and the fact the car was in the shop, kinda makes everything sound strange.

 

I can swap both out tomorrow with two known good ones from an EJ25 I have kicking around, but I don't think that's going to solve my ECM communications problem. If it does, I'll buy you a beer next time I'm down on the island! 

 

I'm getting closer and closer to pulling this engine and putting it in the 2.5GT that I just swapped a 2.2 into, then replacing the non-functional auto with a 5MT and calling it a day.

Edited February 26, 2013 by ChrisInVT

[89Ru]

get a notebook and a voltmeter.  start measuring voltage from the battery and on downstream to the electronics.  I doubt the battery is bad but i'll bet something is dragging the voltage down as noted by the dimming screens and wonky sensors.  computers need stable power.  maybe the voltage regulator in the alt is bad.  you could check voltage in the fuse panels.  sorry if you checked all this already...didn't read your other posts.  what is the voltage on the 12V power input at the cable to the ECU?  

[1 Lucky Texan]

if you disconnected the starter, would the scanner still dim when you went past ON to start?

 

just seems like it could be a low voltage issue, caused by either heavy current draw thru old-ish cables and/or a weak battery.

 

At he first sign of a dimmed scanner, I would be measuring battery voltage, voltage while cranking, etc.

[ChrisInVT]

I've done the troubleshooting with both a battery charger connected and just the battery alone. The power to the diagnostics port doesn't drop when the key is turned to ON (when the fans are disconnected). Voltage on the battery is slightly above 12v, when the charger is connected it's slightly above 14v. Getting the same readings at the diagnostics port.

I didn't check the voltage at the computer yet, I'll do that today.

[ivans imports]

crank senser fault or main engine ground on intake have had 3or 4 bad crank senser latly no spark or meatal bitts have found the magnet on end of crank senser or broken timing belt. Do you here the ig realy turn on with key ? and feul pump realy should turn on and cycle have seen the odd ig realy go bad but whould have no check engine light. A no spark is ushaly the crank senser as it tells the ecu when to spark. The fans staying on could be bad coolant temp senser or ac default if ecu gos into default will turn fans on

[torxxx]

what year make model?

 

I would bypass using a key and physically turn the ignition switch with a screw driver.  I had one the other day at work that for some odd reason, it would NOT start using the key but when I'd unscrew the ignition switch from the column, it would start.  

 

Depending on the year, cam sensors control spark.  I had 00 OB kick my rump roast for a few days at work, come to find out the customer had switched out the cam pulley with a 02+ pulley (different pickup marks for the cam sensor, which controlled the spark)  I eventually got it to fire with the wrong cam pulley/sensor but it was firing on compression upstroke so it would act like waaaay advanced timing and kickback on the starter. 

 

For now to keep your battery from dropping voltage, remove the fan fuses/relays 

[ChrisInVT]

98 Outback, and the fans are already unplugged because i was sick of hearing them.
 

I didn't get to test much today, I had to fix my girlfriend's car and then do some work to my truck, since I'll be driving those while this car is dead. My friend was nice enough to bring over his $2,500 Snap-On OBD-II computer reader (he has lots of drool-worthy tools) and it kept coming up "Failure to communicate with ECM". As soon as the ignition was turned from ACC to ON, it would power off his reader. He did a manual bypass (booted his reader off the internal battery) and it wouldn't see the ECM at all (in the ACC position, the car would power the reader, but when turned to ON it would cut power to the reader, even though there was a steady 12v power).

Doesn't matter if the engine is even electrically connected to the car, the ECM isn't talking to anything (we even tried the other ECMs).

He's coming over Thursday night with some more tools and we're going to tear the electrical system apart. He knows people at the Subaru dealership and some other people that work on Subarus so he's going to run the issue past him. I got "In my 40+ years of automotive repair I've NEVER seen anything like this before"

[uniberp]

"I would bypass using a key and physically turn the ignition switch with a screw driver.  I had one the other day at work that for some odd reason, it would NOT start using the key but when I'd unscrew the ignition switch from the column, it would start. "

 

^This. I have had keys lock in the ignition. Also I've seen the neutral safety switch intermittantly fail. Seems worth a look.

[ivans imports]

did the ig realy click ? did the feul pump cycle ? did the tach bounce ? at all is there 12 volts at the coil ? a non com with ecu could be ecu memery backup fuse failed or mian ground to body

[Gloyale]

Fans run on key on when the temp sensor or some ther main engine connection are bad.

 

OR when the sensor has failed to a dead short ( thinks the engine is way over temp)

 

I'd try a new temp switch.

 

I'd check you fusible links too. You could have half the ECU not receiving power or getting it through a crispy link that is high resistance.

[ivans imports]

Thats why check for power at coil rules out and power supply problems the coil gets 12 volt when key is on Then we listen for ig realy to turn on that indacates that ig switch is turning on realy Then we listen for feul pump realy to cycle thiss indacates that the ecu is geting crank signal { after the first feul pump cycle when cranking ] Then we put a noid light on injector plug and watch for injector pulse. Spark check well never see spark problems ever unless bad crank senser. Out of 200 no starts one no spark in 12 years was a smashed crank senser got a steel ball berring from coged idler at some point.

[Cougar]

From what you state about the code reader display going dim when you turn the key on it means there is a problem with the connection to power on that circuit. If you don't already have a voltmeter to test with and help you with this trouble then you must get one. I first suggest you check for power getting to the fuses in the panel under the hood while the key is ON (there is a load on them). Looking at the top of the fuses you will notice that there are small slits on each side of the fuse you can place your probe tip on to check for power. Place the common meter lead on a good chassis ground point and then check each side of all the fuses using the other meter probe looking for good voltage level. I think fuse SBF-5 is of particular interest for your problem. I believe it provides power to the Main Relay which provides power to the data port along with a lot of other engine related systems. The relay has two sets of contacts so perhaps the relay contacts may be bad. Another possible trouble area is connection to the Main fuse panel itself located under the hood. So even if the fuse has proper power on it, the output wire connection for that circuit may not.

[torxxx]

+1 for check SBF #5.  no power to that = no ecm power

[Fairtax4me]

This reeks of poor connection to supply power. Check on cougars suggestions. Also download a copy of the FSM here: http://www.main.experiencetherave.com/subaru_manual_scans/ to help track down which parts of the circuit you need to focus on. 98 and 99 wiring are the same with a few small exceptions, but the 99 manual there should have what you need in the wiring diagrams section.

[ChrisInVT]

Main relay is OK, fuel pump is running. Unplugging the temp sensor makes no difference. ECM getting .15v less than I see at the battery. Ignition coil has power also, tach does not move when cranking. Injectors have power, but not getting a pulse from the ECM.

I have lots of crap disconnected under the dash (everything that's connected to the lower plastic piece) but the following fuses have no power: 15A Radio, 10A Illumi, 20A Wiper Washer, 20A Cigar Mirror, 10A Tail Clearance.

[ChrisInVT]

This might provide some insight... The battery is at 13.3v with the charger on it. The OBD2 port has a good 13.16v if I test from pin 16 to a chassis ground.

If I use the chassis ground on the OBD2 port, I get 13.03. When I put the ignition in the ACC position, it drops to 12.38. When I put the ignition to ON it drops to 4.41v! With the ingition on, if I move from the chassis ground on the OBD2 port to an actual chassis ground I get 12.53v. This is with the scanner disconnected. If I try to ohm out the chassis ground wire to an actual chassis ground, I get NOTHING. I'm going to trace the chassis ground on the OBD2 connector - and GOOD NEWS - if I measure the amp draw between the chassis ground on the OBD2 connector and an actual chassis ground, I get a .30A reading AND THE ECM POWERS ON AND I CAN PULL CODES!!!

I'm getting closer....