There's only one wire going to my starter solenoid and there's not one coming off so are you saying that in the crank signal delivery from the ignition switch there is a wire that is joined with the crank signal wire that goes to the PCM?

Thank you ill try that

As soon as I get back to my PC I will download that PDF and start tracing wires thank you guys very much this forum is awesome

There is a start signal input wire to the ECU, but it just splices off of the start wire from the switch to the solenoid. It's buried in the harness in the dash and not easily accessible.

The ECU does NOT NEED that input in order to start the engine.

 

Like GD said it sounds like Something is losing voltage when the key is in the start position. Would have to check voltage at the start switch, then on each side of the main relay under the dash.

A noid light will only tell you if you have a trigger if there is voltage AND ground present at the injectors when the ECU closes the ground. Test for 12 volts at the injectors when cranking. If you have no voltage look at injector power supply, ignition switch, main relay, etc.

If you have voltage but no ground look at ECU power supply and if that checks out then consider swapping ECU.

Edited January 5, 20188 yr by Fairtax4me

Utilizing the Noid light does tell you if there is voltage to the injector because without the voltage pulse being grounded by the PCM or ECM the Noid light would not Flash or am I missing something in my small brain of electric knowledge :-)

The injector must get a pulse at every revolution of compression otherwise the engine will not run......

First, I think you are wrong about this.  If the engine is just cranking and the injectors squirt gas on every compression stroke, you could flood the engine.  It is hard to say because new cars with everything working start so quickly that they only get the one pulse and their running.  Of course once running, they get a shot of gas at every compression cycle.

 

But on re-reading your original post, I seemed to have missed something, or didn't see the significance of it the first time.  You stat that once you stop cranking and hit the starter again, it starts right up.  That leads me to suspect that you are losing fuel pressure during long shut downs.  There could be a number of reasons for this, but there are a couple of things you could try.

 

If you own a fuel pressure gauge, hook it up, start the engine and then shut down.  Make a note of the fuel pressure, then check it after a long shut down and see if it held pressure.  If you don't own the gauge, then after a long shutdown when you normally have this problem, turn the key to the run position for two seconds, then off, and then start and see if it starts immediately.  That would confirm that loss of fuel pressure is the issue.

Incorrect I must have Miss stated. The Noid light would flash when the key was let go and the engine did it spinal rotation it did not happen all the time and the engine would not start up immediately it would take some more cranking when it decided to but the only time the motor would start up was after it had flashed once letting the key up so I'm going to follow that ignition wiring

The noid light flashed when you let up because it thought the engine had started.  It doesn't really know that the engine has started, only that the key is in the run position and it got a signal from the crankshaft position sensor or cam position sensor.  But by all means, continue to check your wiring, I am only offering an alternative to consider.

Edited January 5, 20188 yr by keith3267

Yes, I totally agree but it should pulse while the engine is cranking otherwise the ground signal coming from the computer is not reaching its destination and that is where my problem lies is it one of the four wires from the ignition switch or is it the splice from the ignition crank signal to the ECM or other?

So maybe it's time for a new ignition switch again LOL

When I get that PDF download I'm going to trace every single wire from the ignition switch power distribution and crank signal distribution and then start probing and testing to find out where I'm losing continuity between the ignition and the PCM because it is not the crank signal cuz that's just 12 volts to the solenoid

Unless it technically is in the splice as somebody described earlier

Unless it technically is in the splice as somebody described earlier

 

Correct - the small lead that runs to the PCM comes directly off the 10 AWG solenoid wire. But it's an internal splice inside the harness and as Fairtax mentioned it is completely inaccessible, buried behind the dash. 

 

I doubt you have a fuel pressure issue - the PCM has no knowledge of fuel pressure and would continue happily pulsing the injectors without any fuel at all. What you seem to have is either a slow cranking speed coupled with no start signal to the ECU, or a loss of 12v to the ECU. Did you say that you had solid 12v at the injectors?

 

As for the ECU start signal.... they will eventually struggle to life without it if you turn the engine over fast enough, but the ECU gives additional fuel and changes timing to allow the engine to easily start at the ~300 RPM cranking speed. Without that signal, it may be difficult to start and the slower it cranks the worse that would become. Also typically harder to start when cold as the ECU's only inputs are crank and cam signals and without the start signal it just starts seeing engine rotation - it may not idle up and provide cold start enrichment. 

 

GD

Starter switch is pin 81 on your ecu.

If you don't get 12 volts (pinout says 8-14v) supply 12 volts to that pin using a power probe or another battery.

I have my starter switch coming from the little spade connector on the starter, straight to the ecu.

Not to be confused with ignition switch pin 79?

Edited January 6, 20188 yr by sparkyboy

Awesome, will do that when my effing son comes back from the store with my car after he left his friggin subaru 10 miles away at another store.....

 

And hes been gone over 2 hours to go buy one item.... Ugh, old rump roast kids...

You know what's funny, is that i may have some insight to what is going on with my ej22. The starter switch may not be getting a proper signal. I stuffed my ecu into the glovebox so i can check. At temps below 10*f she takes up to 10 tries on the starter, quick bursts of less than 2 seconds. When she starts she idles fine, but will stall immediately if you touch the gas. After about 30 seconds of idling i can pump the gas (being completely serious here) to get her to a steady 1500 rpm for another 30 seconds and she is perfect. Just like you my car starts very easily when even a little warm.

My efi acts like a carb in extremely cold weather, i think we have the same problem, but i have my fuel pump wired to run all the time with the key in run, while you only have the little squirt of fuel for priming.

Just for grins try short bursts on the key. Start in off and turn straight to start no more than 2 seconds at a time, try to get the pump to prime several times.

Edited January 6, 20188 yr by sparkyboy

Regardless of temperature once this engine fires over it has absolutely zero issues whatsoever I mean it's like its brand-spanking-new

I think this may work, try it for fun.

Turn the key to on, let the pump prime, turn off and repeat. Dont hit the starter until you have primed a few times, i suggest 5 or more. You may have to turn the key to off for several seconds to get the pump to prime again, I'll bet you'll at least get a cough.

Mine fires up with just a tap on start after she has ran a bit, even in sub zero.

At least you can try something until your kid gets back hahaha!

Edited January 6, 20188 yr by sparkyboy

That was one of the first things I did the day I got the car and started having this issue then I said to hell with it and replace the fuel pump I have 36 PSI from 3 seconds after I turn the key on for eternity

Well damn, maybe it's a rare ecu failure. I'll dig out my spare if you need one.

I'm going to go fetch the car in the morning I'll let you know what happens

Do you ever step on the pedal just a little when it won't start?  If you do that with the noid light hooked up, you may start seeing pulses.

I have in the past but not while the Noid light was hooked up. It had no effect on starting whether the throttle position sensor new the throttle is open or the mass airflow sensor knew that there was more air passing none of those changed anything

I should rephrase/clarify what I said earlier.

A noid light tells you basically two things:

Either you have BOTH voltage AND ground (light blinks)

Or you Do NOT have voltage, or ground, or possibly both (light does not blink)

But it cannot tell you WHICH ONE you don't have. This is where the multimeter comes into much better use, and why I personally don't even own a set of noid lights.

The injectors have constant power when the key is On. The ECU grounds the circuit to pulse the injectors. In the case of this particular Subaru, It grounds each one individually in sequence (Sequential port fuel injection) with the firing order. And yes, the injectors SHOULD be pulsed while the engine is cranking. Otherwise it won't start (which is the exact problem we have in this thread).

 

Back to injector power/ground. You're missing one or the other, or maybe both.

ECU and injectors are both powered through the Main relay, along with the ignition coil and a handful of other things. So to break it down, check for power reaching the injectors during cranking. You only need to check one, they all share the same power. You can also check for power reaching the coil on the center wire (yellow IIRC) during cranking.

You should have power at both with key On, AND during cranking at key Start.

 

No Power at all, look to the main relay. Power with key On, but not when cranking, look to the ignition switch.

I am taking my multimeter and a spare battery and other tools this morning to retreive the car along with the all the wiring info I can get.  I will defeat this gremlin,.  Thank you guy's I'll let ya know what happens, or you'll see the smoke from the car fire in the sky over gig harbor caused by a small 5 gallon fuel leak....