hard start now no start

[hedron]

2011 Subaru Impreza 2.5L premium EJ25 engine

I bought this car oct 2023. About six months (maybe a year) later I noticed the RPMs dip abnormally low, and a clacking noise coincided with it. It didn't happen very often, and I thought it's probably just the water pump bearings have a grain of sand in them or something. I posted the noise here and people told me not to worry, it's normal.

So a week or two ago, the car would start normally in the morning. But after I drove it somewhere, like to a store, and parked, I'd come out and it would hard start. I'd turn the key all the way clockwise, and it would just chu-chu-chu-chu-chu infinitely.

Sometimes I'd get 335 crankshaft CEL, and the cruise control light would be on. This only happens when cranking abnormally long. Never got this while just driving or when it starts up normally.

Mostly it would start eventually. On advice, someone told me to 'prime' the pump by turning it to 'on' briefly, then off, then on again, three times. And this seemed to work for a while. And this was "proof" it was the check valve in the pump. Further evidence pointing to fuel was that it would start with starter fluid.

Twice while driving the RPMs dipped, as mentioned above, and the car sputtered and died. I started the car up again no problem. Typically the RPMs dip while slowing to a stop quickly. Rolling to stop without braking hard does not trigger this.

I tried a fuel pressure test , and the pressure is 52 psi or so.

I swapped out the fuel pump and the problem remains.

When the car is driving, there is no power loss or any other problems.

Now, it won't start at all. I get chu-chu-chu-chu and maybe a sputter or two. Tried 'priming' the pump, tried starter fluid, nothing.

I can't help but see the connection between the hotter weather and this problem coming about. Maybe it's just a coincidence?

Any idea what the problem could be?




tldr version:

>RPMs dip to near zero while stopping and sometimes while idling
>when car warmed up, turn ignition, no start, nothing but chu-chu-chu-chu-chu-chu-chu to infinity

* Registered users of the site do not see these ads.

[hedron]

I pulled the crankshaft sensor (which appears to be indentical to the camshaft sensors) and there was a bit of black powery gunk on it. Did the multimeter test, and it appears to be good. No, real surprise there. I guess the only thing left to do is open up the fuel injectors and timing cover.

[hedron]

The a spark plug, and it was fine. Ohm rating .7 at 20k.

A fuel injector was working fine.

Both components were in good physical shape. I only checked one of each, so I suppose it could be a different one? But I really doubt it.

So, I guess all that's left is the timing belt, and assorted pulleys and pumps. Unless it's the valves, but then if it was something that bad wouldn't I get more of a CEL than a crankshaft sensor code?

[hedron]

Apparently, the crank shaft sensor was dirty. It had some soot all over the sensor end and that was causing the problem.

Although, I can't help but think that the oil leak is related to this. That soot didn't just appear there through witchcraft.

[Charlie-III]

Originally Posted by hedron

Apparently, the crank shaft sensor was dirty. It had some soot all over the sensor end and that was causing the problem.

Although, I can't help but think that the oil leak is related to this. That soot didn't just appear there through witchcraft.

Soot? More likely dust off the timing belt. Dust/dirt won't usually mess with the signal, metal particles or rust dust on the face of the sensor could.

[hedron]

It just looked like soot. I guess it would have been more accurate to say, "it had something that looks like soot." But wiping it off fixed the issue. At least, temporarily since it's probably going to build up again. I suppose I should have kept it and sent it to a lab for analysis. But it was a powdery black substance.

It could have something to do with the oil leak? I guess next time I change my oil I'll have to open the valve cover and poke around. Not sure I'll find anything. Although, I did discover the top center bolt was loose, so that could explain the leak. But I'm not convinced.

>dust off the timing belt

How would that get into the crank case?

[Charlie-III]

I don't know what engine you have, but the EJ's are belt drive timing and the crank sensor is under the belt cover outside of the crankcase.

Are you talking cam sensor?

I believe newer engines are chain driven timing which means the crank sensor is in oil.

[hedron]

I have EJ25 engine. No turbo.

By appearance alone, the sensor looks exactly like a cam sensor, only called the crank sensor by the computer. Even some auto stores don't even carry the crank sensor, because it's the same thing, only located in a different location. I mean, maybe the sensors' sensitivity is different? Like the resistor or whatever is, is different to allow different amount of magnetism to trigger it. IDK. But I did the ohms test with it, and it seemed operational.

Anyway, whatever sensor it is, it's located on the top of the engine, in the center, directly under the alternator. You could take it out without removing anything, but I removed the alternator just to be able to put it out without a pry bar since it was stuck pretty good, and I didn't want to accidentally break it causing this problem to spiral out of control.

Unless I'm royally confused about something, the crankshaft is in the middle of the engine on this car. The harmonic balancer is right there directly under the alt and the crank sensor.

[Charlie-III]

OK, so like I thought.
Yes, crank is in the middle.
Belt is just below your crank sensor.
Yes, it's belt dust on it.
No, it's NOT inside the crankcase, it's inside the timing belt covers.

[Elbert Bass]

Originally Posted by Charlie-III

I don't know what engine you have, but the EJ's are belt drive timing and the crank sensor is under the belt cover outside of the crankcase.

Are you talking cam sensor?

I believe newer engines are chain driven timing which means the crank sensor is in oil.

The FA/FB crank sensor is on the left rear of the bell housing near the knock sensor. It reads a reluctor plate that is fitted between the flywheel/flex plate and the crank shaft - not in oil. The camshaft sensors are in oil - at each corner of the timing chain cover reading reluctors on the front of the cam sprockets.

[Elbert Bass]

Originally Posted by hedron

I have EJ25 engine. No turbo.

By appearance alone, the sensor looks exactly like a cam sensor, only called the crank sensor by the computer. Even some auto stores don't even carry the crank sensor, because it's the same thing, only located in a different location. I mean, maybe the sensors' sensitivity is different? Like the resistor or whatever is, is different to allow different amount of magnetism to trigger it. IDK. But I did the ohms test with it, and it seemed operational.

Anyway, whatever sensor it is, it's located on the top of the engine, in the center, directly under the alternator. You could take it out without removing anything, but I removed the alternator just to be able to put it out without a pry bar since it was stuck pretty good, and I didn't want to accidentally break it causing this problem to spiral out of control.

Unless I'm royally confused about something, the crankshaft is in the middle of the engine on this car. The harmonic balancer is right there directly under the alt and the crank sensor.

NO - the crankshaft sensor and camshaft sensor are different part numbers. You can have a sensor that "measures" good but has a fault, and yes electrical components can be temperature sensitive. That "soot" is not the root of your issue unless it has metalic particles in it that want to stick to the sensor face.
Based on information you have provided - you have a hard start issue, a crank sensor code, and a no start issue, the first step should be replacing the crank sensor - with the correct part.

Ideally you would have a scan tool that reads live data and you could verify the fault - you would not have a valid engine RPM reading during cranking.

[hedron]

I mean, OK, but I wiped the soot-like stuff off and it starts fine. I think the sensor is just fine. It's just finding out where that stuff is coming from.

Rubber is an insulator. Assuming it was shavings from the timing belt, that would provide enough insulation to dull the sensors ability to monitor the crank speed. I think that's a rational explanation, given the fact this issue slowly crept up on me. The timing belt shavings just slowly built up over time, until the crank sensor couldn't determine any magnetic force from the harmonic balancer.

It's likely the timing belt is worn, and I think the last time it was changed was about 180k or so, and it's now at 245k. So, it's in need of changing.

[Elbert Bass]

Originally Posted by hedron

I mean, OK, but I wiped the soot-like stuff off and it starts fine. I think the sensor is just fine. It's just finding out where that stuff is coming from.

Rubber is an insulator. Assuming it was shavings from the timing belt, that would provide enough insulation to dull the sensors ability to monitor the crank speed. I think that's a rational explanation, given the fact this issue slowly crept up on me. The timing belt shavings just slowly built up over time, until the crank sensor couldn't determine any magnetic force from the harmonic balancer.

It's likely the timing belt is worn, and I think the last time it was changed was about 180k or so, and it's now at 245k. So, it's in need of changing.

Rubber is an ELECTRICAL insulator - that sensor works off of MAGNETISM.
What is probably happening is there is a short or open in the coil in the sensor. You removing and installing the sensor stresses the housing causing the connection to work. Heating & cooling cause the sensor housing to expand/contract breaking the connection again.
Unless you have a scan tool or oscilloscope to properly diagnose the issue you will just keep swapping parts - put a new crank sensor on it or perform a proper diagnosis.

[Charlie-III]

Yes, rubber would not affect a "hall effect" type sensor. Rust, metal shavings, "metallic dust" from rubbing metal parts on the face of the sensor will affect the reading.

I hate intermittents, whether heat or connection type. Have the thing just die.

[Elbert Bass]

Originally Posted by Charlie-III

Yes, rubber would not affect a "hall effect" type sensor. Rust, metal shavings, "metallic dust" from rubbing metal parts on the face of the sensor will affect the reading.

I hate intermittents, whether heat or connection type. Have the thing just die.

These are not even Hall Effect sensors (3 wire)
Just plain old 2 wire inductive sensors...

[Charlie-III]

Originally Posted by Elbert Bass

These are not even Hall Effect sensors (3 wire)
Just plain old 2 wire inductive sensors...

Thanks for clarification, thoughts still apply in terms of what does/does not affect operation.

Did I say I hate intermittents?