Injector power supply questions

Headlights to tailights and everything in between.

Moderators: Helpinators, Moderators

92 ra-r
First Gear
Posts: 57
Joined: Sun Dec 16, 2007 8:39 pm
Location: L'ville, GA

Injector power supply questions

Post by 92 ra-r »

Can someone explain injector circuitry and power supply? Looking at the diagrams, it seems that yellow wire supplies power full time, and then the other wire supplies a signal from the ECU when the injector is supposed to fire. I have power on both wires full time, even with key off. I'm worried that the injectors are either firing all the time, or not at all. I have good fuel pressure, but car just cranks and cranks. I've pulled plugs and they are wet, but not dripping, so I think they are working normally, but car will not start! I've done every step of diagnostic in the book and all systems check out fine. I have good compression, yet car will not run!
Is there maybe some sort of vehicle theft control I don't know about? I've found several strange boxes under the steering column, but I think they are cruise control boxes, and perhaps power door locks, as one has wires running to the door locks and a receiver antenna. What would I need to do to find out if this car had an alarm on it, and whether I need to have the remote to defeat what ever sort of start defeat/whatever it does? (I do have spark by the way, so that isn't it, but I don't know what alarms do, so .....)
vrg3
Vikash
Posts: 12517
Joined: Sun Nov 03, 2002 2:13 am
Location: USA, OH, Cleveland (sometimes visiting DC though)
Contact:

Post by vrg3 »

The yellow wire does supply power all the time. That's how it's supposed to work.

The other wire is connected to the ECU, and when the ECU wants to open the injector it grounds that wire.

When you say you have power on both wires full time, are you measuring with the injector connected? Because that's what you'd expect to see, since the injector's coil would conduct the ~+12v to the other side too.

Have you tried a noid light to see if the injectors are being pulsed?
"Just reading vrg3's convoluted, information-packed posts made me feel better all over again." -- subyluvr2212
92 ra-r
First Gear
Posts: 57
Joined: Sun Dec 16, 2007 8:39 pm
Location: L'ville, GA

Post by 92 ra-r »

I've been trying to find a noid light, but I ask for them at parts stores, I just get totally blank stares! It's extremely frustrating! I'll be finding myself a tool truck tomorrow and buying one.
vrg3
Vikash
Posts: 12517
Joined: Sun Nov 03, 2002 2:13 am
Location: USA, OH, Cleveland (sometimes visiting DC though)
Contact:

Post by vrg3 »

Today's young whippersnappers don't know what it's like to have to actually troubleshoot.

The stores probably do carry them even if the clerks don't know it. They'd be with all the other diagnostic and repair tools, and might be called something like "electronic fuel injection test light" or something.

You could just use a W3W bulb like the one in your sidemarkers, though. The only difference between that and a noid light is that the noid light has a connector that fits injector harnesses.
"Just reading vrg3's convoluted, information-packed posts made me feel better all over again." -- subyluvr2212
92 ra-r
First Gear
Posts: 57
Joined: Sun Dec 16, 2007 8:39 pm
Location: L'ville, GA

Post by 92 ra-r »

I just monitored the power out of the ECU with my volt meter and I see a constant voltage while cranking, there is no interruption/pulse. I'm thinking security has this car disabled! Thing is, I don't think I got any kind of remote from the guy when I bought the car. I guess I'll start looking through the wiring looms and seeing what taps into what to try and find a switch or something else that might be the problem!
vrg3
Vikash
Posts: 12517
Joined: Sun Nov 03, 2002 2:13 am
Location: USA, OH, Cleveland (sometimes visiting DC though)
Contact:

Post by vrg3 »

If the ECU is firing the plugs but not the injectors, then it can't be a security system problem as far as I can tell.

MAF sensor, perhaps? Maybe even coolant temperature sensor?

Keep in mind, though, that the injector pulses are very short when cranking, so you need to make sure you're using a voltmeter that can respond very quickly. That's why noid lights are used; light bulbs illuminate even if they're given short bursts of power.
"Just reading vrg3's convoluted, information-packed posts made me feel better all over again." -- subyluvr2212
92 ra-r
First Gear
Posts: 57
Joined: Sun Dec 16, 2007 8:39 pm
Location: L'ville, GA

Post by 92 ra-r »

I hooked my test light up to the signal wire, it stays on until I turn the key on, then it flashes once. It stays on solid when I crank....no pulsing. I'll try to find a proper noid light, or perhaps an LED with pigtails I can plug into the harness. This is so frustrating! I just put in a different ECU because my original one had a charred chip on it, but that did not help anything!
vrg3
Vikash
Posts: 12517
Joined: Sun Nov 03, 2002 2:13 am
Location: USA, OH, Cleveland (sometimes visiting DC though)
Contact:

Post by vrg3 »

The test light approach is good enough; if you're measuring with it you don't need a noid light.

An LED wouldn't do because it turns on and off so fast you wouldn't necessarily be able to see it flash.

That initial pulse is the priming pulse.

It seems pretty clear that the ECU's not firing the injectors while cranking.

What chip was charred in the ECU? Do you have an explanation for why that happened? If it was an external short, the new ECU may have been damaged too.

I feel your frustration.
"Just reading vrg3's convoluted, information-packed posts made me feel better all over again." -- subyluvr2212
92 ra-r
First Gear
Posts: 57
Joined: Sun Dec 16, 2007 8:39 pm
Location: L'ville, GA

Post by 92 ra-r »

I'm giving up until the weekend, and then I will have some help to crank while I test the signal from the cam and crank sensors. I would think if those were bad, I would be getting a CEL, and the only one I have is the purge solenoid, and I can't see how that would keep the car from running!

The old ECU damage was suspected, not confirmed. I found the fusible link had been replaced with speaker wire, so I got a new one and replaced that before even starting this ordeal. I found a slightly odd looking chip inside that is all hooked to F47 plug, and all the issues I had were with wires coming from F47, so it made sense. Now, nothing makes sense!! I have checked the new ECU since trying to start the car, and it looks to be fine, so I imagine whatever destroyed the original fusible link took out the ECU too.

I'll have more to report next weekend, or a small bonfire if I can't find anything!! =)
vrg3
Vikash
Posts: 12517
Joined: Sun Nov 03, 2002 2:13 am
Location: USA, OH, Cleveland (sometimes visiting DC though)
Contact:

Post by vrg3 »

I would think that if the cam and crank signals weren't coming through, you wouldn't get spark.

And I agree that a CPC solenoid code wouldn't prevent the car from running.

A slightly odd looking chip inside? Weeeird. Any chance you could take a picture?

Can you measure the signal from the MAF sensor and coolant temperature sensor?
"Just reading vrg3's convoluted, information-packed posts made me feel better all over again." -- subyluvr2212
92 ra-r
First Gear
Posts: 57
Joined: Sun Dec 16, 2007 8:39 pm
Location: L'ville, GA

Post by 92 ra-r »

Coolant temp sensor I can't get to all that easily, I can check the MAF in a few minutes though.

Wierd thing happened this morning. I decided to try and take a fuel injector out and see if it was spraying or not. I'd run the test light out of the ECU again this morning and still no blinks for ground pulses. I got the rails loose, but couldn't get them out, so I just put it all together, and when I got ready to crank, I was getting ground pulses from injectors #1 and #4, but not 2 nor 3. Wierd!! I wonder if it's just loose connectors, grounds or something. I did check the crank sensor circuit and I guess my VOM is not sensitive enough, because I couldn't get any kind of signal out of it, nor would in register sensor resistance in M-ohm. I tried to crank the car while it was unplugged and I got code 11, so it must be signaling the ECU properly when plugged in.

Now that I have two injectors firing, it really sputters and sounds like it will start if I get atleast one more going, but what are they drawing a signal from? What can I check on those last two injectors to get them to work? What is even stranger is the fact that they are on opposite banks, which almost means a wiring issue! Grrr.....so close!!
vrg3
Vikash
Posts: 12517
Joined: Sun Nov 03, 2002 2:13 am
Location: USA, OH, Cleveland (sometimes visiting DC though)
Contact:

Post by vrg3 »

Hm!

The injector ground is a separate wire from all the other grounds. It grounds at the grounding bolt near the rear of the driver side fuel rail. The black-with-yellow-stripe wire. It goes to the 16-pin square connector behind the battery and then to the two injector ground pins on the ECU.

Maybe you could take a look at those wires, at least the easily visible parts, to see how they look. Maybe you'll be really lucky and it'll just turn out that that 16-pin connector wasn't properly fastened.

Do you think there's any chance you could get your test light's probe to the ECU's pins? You could check to see if the injector ground wires are well grounded, and then see if the injector drivers are firing.

http://www.graphics.cornell.edu/~v/ecupins/
"Just reading vrg3's convoluted, information-packed posts made me feel better all over again." -- subyluvr2212
92 ra-r
First Gear
Posts: 57
Joined: Sun Dec 16, 2007 8:39 pm
Location: L'ville, GA

Post by 92 ra-r »

I've been using the ECU pins all day, as I can't crank the car from the engine bay. I do notice this, and I think it's a ground issue. When I have the test light grounded on the center console grounds, and I have the light on the ECU pins, the car will cough and sputter like it's trying really hard to start, but not getting fuel. I do smell fuel during these times.

When I disconnect the test light and just crank the car, it does nothing, so it has to be a ground. I have run countless continuity checks from those pins to the body and they are grounded. I even relocated that specific ground and ran a 10ga wire directly to the body rail below the washer bottle, just to make sure it is done!

I tried to "roll start" the car, and force it to run, but that didn't work either! So, tomorrow is another day, but at this rate, I'm wondering why I didn't buy a truck, and get my rally car tuned and go racing!!
vrg3
Vikash
Posts: 12517
Joined: Sun Nov 03, 2002 2:13 am
Location: USA, OH, Cleveland (sometimes visiting DC though)
Contact:

Post by vrg3 »

Well, maybe you could just tap the wires to all six ground pins of the ECU (F47.14, F47.15, F47.24, F47.25, B48.21, and B48.22) and run a new ground for them?

The body is not the ground reference you want; to make sure something is grounded properly you need to bolt it to the engine.
"Just reading vrg3's convoluted, information-packed posts made me feel better all over again." -- subyluvr2212
92 ra-r
First Gear
Posts: 57
Joined: Sun Dec 16, 2007 8:39 pm
Location: L'ville, GA

Post by 92 ra-r »

I ran a quick jumper wire off both injector grounds, then tapped that wire to the body, the negative battery terminal and two places on the intake manifold. The car sputtered more, and sounded oh so close to starting, but did not actually fire up.... I ran the test light off the ECU like I have been and still get no pulse from injectors 2 and 3. I'll probably search for all grounds and run new ones, and ground each injector cap. I'll also take all the plugs out and crank the engine for a few minutes to blow stuff out a little bit. It smells very strongly of fuel, so I'm thinking the injectors that are working might be putting too much in, and I'd like to dry it all out before continuing.
Any other advice or suggestions? I've never seen anything like this---EVER!!!
vrg3
Vikash
Posts: 12517
Joined: Sun Nov 03, 2002 2:13 am
Location: USA, OH, Cleveland (sometimes visiting DC though)
Contact:

Post by vrg3 »

That's so weird. Pulses on injectors 1 and 4 but not 2 and 3?

That sounds like a problem with the cam and crank position sensors, doesn't it? Or maybe even an actual cam timing problem?
"Just reading vrg3's convoluted, information-packed posts made me feel better all over again." -- subyluvr2212
92 ra-r
First Gear
Posts: 57
Joined: Sun Dec 16, 2007 8:39 pm
Location: L'ville, GA

Post by 92 ra-r »

Yeah, I'm really scratching my head here!
I get 135 and 140 for compression on the left bank, and that one has the tricky intake cam that jumps when you are changing the timing belt. I was worried that I set it off time, but I have verified that everything is still lined up 3 times now! It was all original on this engine when I received it, original timing belt, etc so I have no reason to doubt it was messed with before I got it.
I get no CELs except for the purge solenoid, and from the 11 code from when I left the crank sensor off intentionally. I can't clear anything until the car is running. I left the battery off all night, and disconnected the ECU wiring harness and once I hooked everything back up, and plugged in the black check plugs, the same codes came up as last week. I hope I can clear them once the car is running!
vrg3
Vikash
Posts: 12517
Joined: Sun Nov 03, 2002 2:13 am
Location: USA, OH, Cleveland (sometimes visiting DC though)
Contact:

Post by vrg3 »

You've done other Subaru timing belts, right?

Any chance you could swap in another cam sensor and another crank sensor? Maybe from your original engine?
"Just reading vrg3's convoluted, information-packed posts made me feel better all over again." -- subyluvr2212
92 ra-r
First Gear
Posts: 57
Joined: Sun Dec 16, 2007 8:39 pm
Location: L'ville, GA

Post by 92 ra-r »

Yeah, I've done a US WRX before, and have all the print outs for the DOHC engines. I am confident that those parts are done correctly. I will pull the airbox and snorkus out and pull the right side plugs, check compression there, look for loose wires and such this weekend, and hopefully I'll find something that is blatant! If not, I'll be taking the IM off and rewiring everything, then putting it back on. I really didn't want to have to do mechanical work to it, but I might have to.
92 ra-r
First Gear
Posts: 57
Joined: Sun Dec 16, 2007 8:39 pm
Location: L'ville, GA

Post by 92 ra-r »

I spent most of the weekend out of town, but did get into the garage for a few minutes this evening. With a noid light, I confirmed what I've been testing with the test light. NO pulsing on the two injectors. I checked continuity between the ECU and the injector lead, and there is indeed continuity there. That was my last thought! I guess the intake manifold is coming off next weekend!
Without having to rob a new harness, where can I get new pins for this thing. I'd almost rather make a new harness than take this one apart! I guess I could pull one from a junkyard and break up the connectors and use those wires with leads on them. I'd so much rather do it from scratch though, so I know the wires are good.
92 ra-r
First Gear
Posts: 57
Joined: Sun Dec 16, 2007 8:39 pm
Location: L'ville, GA

Post by 92 ra-r »

It's been atleast two weeks since I even touched the car, been doing research, school, gf, etc... Today I figured I would try a few things. First I swapped the injector plugs on each side, but I still only got signal to 1 and 4. I swapped them back before proceeding. I thought maybe the cam or crank sensors would be feeding signals to the ECU that would convert into injector pulses. I left the EJ20G sensors in place originally, so today I swapped those out for the EJ22T sensors and noticed a few things. I swapped both at the same time. I immediately heard the fuel relay clicking again as I cranked, not just as I turned the key on. It had not been happening before. I checked the ECU out side for the injectors--no pulse to any injectors! Crap!! I swapped the cam sensor back to the original G sensor, and the fuel relay stopped clicking, still no pulse. I decided to use the 22T cam sensor as that atleast turns on the fuel relay. I swapped the crank sensor a few times, but never got injector pulse back!

I think this weekend the IM will come off and I will pull the 22T engine wiring harness in favour of the 20G harness and repin the grounds. I can see that I have a short or something going on, but I keep chasing my tail. I just want to drive already! What gives!
Any other suggestions to try before I get deep into this?
92 ra-r
First Gear
Posts: 57
Joined: Sun Dec 16, 2007 8:39 pm
Location: L'ville, GA

Post by 92 ra-r »

I've spent a fair amount of time on this the last few days and I still haven't gotten any further than I was! The frustrating details: I swapped crank sensors and cam sensors, both of which I checked with a piece of metal to make sure I was getting a signal from them. Then I installed them in the car, and verified signal while cranking. I then verified signal at the ECU from the crank sensor. Everything was getting to the ECU fine ( only CEL is all clear), but I was still only getting a signal to injectors 1 and 4. I checked all the ground plugs and all the power plugs and everything looks fine. I swapped in my other ECU which I thought was burned up, and I get the same issues as the "new" ECU that I got from here. I took them both apart, and there is nothing obviously wrong with either of them, but neither of them will start the car!
I don't know where else to go on this one! Both the cam and crank sensors are okay. Signal from both of them are getting to the ECU. The ECU has proper grounding from what I can tell. I still get only 2 injectors pulsing. Anything? Anyone?? This is getting so old!
Saskatoon Subaru
Second Gear
Posts: 395
Joined: Wed Aug 09, 2006 6:36 am
Location: Saskatoon,SK

Post by Saskatoon Subaru »

i'd recheck the timing belt
2010 Impreza WRX - 265hp

1990 Legacy 267,000km(Rebuild complete 265,423km----March 9/2008)

1991 Legacy Turbo 180,000km
92 ra-r
First Gear
Posts: 57
Joined: Sun Dec 16, 2007 8:39 pm
Location: L'ville, GA

Post by 92 ra-r »

I've checked the timing belt twice, removing all covers, resetting the tensioner properly, spinning it over numerous times while the covers were off and checking all the marks. I've got good compression in all 4 cylinders....but still not running!
vrg3
Vikash
Posts: 12517
Joined: Sun Nov 03, 2002 2:13 am
Location: USA, OH, Cleveland (sometimes visiting DC though)
Contact:

Post by vrg3 »

Hm. So very weird. Is it possible that one or more of the driver's side cam sprocket's pips are damaged? Cuz it does kind of seem like the cam sensor's not getting a good signal.
"Just reading vrg3's convoluted, information-packed posts made me feel better all over again." -- subyluvr2212
Post Reply