Injectors Won't Fire After Swap

[n2x4]

My 91 Legacy Turbo swap into a 95 Impreza L won't start. I'm not getting fuel. I spent all day trying to diagnose it and I'm scratching my head. Pretty frustrated after staring at diagrams for 5 hours and getting nowhere. I don't know what to do. Here's what I do know:

Turning the key on or start gives 0 voltage at the fuel pump. Turning the key off gives me 12V for about 3 seconds. It's as if the fuel pump primer circuit is on, but backwards (coming on when the car turns off, instead of when the car turns on). If I jumper the power to ground on the fuel pump relay, the car chugs fuel all day, but still won't start.

It won't start because the injectors aren't firing. There is no power going to the injectors for some reason. I do know that the pulse signal from the ECU does work, because with a test light from the positive on the battery to the wire that goes back to the ecu at the injector plug, I can see a pulse in the test light as the ECU sends the pulse when I turn the car over.

Why this is happening? I don't know. I do know that I have great spark, and starting fluid down the TB did let it run for a few seconds.

Something's not right, but I really don't know where. It's funny because I didn't really mess with power wires. I basically cut the plugs off the old ECU, and spliced in the Turbo ecu. I didn't cut any other wires than what ran to the ECU. Fuel pump power, and injector power doesn't run directly to the ECU. What's up with that?

I'm going to study the diagrams more, I know I've done everything else right up to this point - every wire's soldiered and heat shrinked, and I have both FSM's to reference - I double and triplechecked all my connections!

Oh, and putting the car into D check mode (green test connectors) does not cycle the fuel pump relay. There's got to be some sort of power issue there, I know the wires that go to the ECU are correct, and I didn't touch the power. Arggg, it's so frustrating, being so close!!!

Does anyone have any ideas? I don't really know what else to check. Also, for troubleshooting purposes, in Dcheck, are the injectors supposed to be powered up and clicking? I'm not sure if they do or not. It doesn't really matter, but it would be easier to check that way.

If anyone is really interested in checking things over, I have all diagrams available in PDF. I actually went to the dealer and shelled out the money for the Impreza FSM at the beginning of this project!

[Legacy777]

I can't speak for the Impreza, but the Legacy has 12v power to the injectors all the time. I would assume the impreza should be the same.

I would suggest you check your fuses, fusible links & SBF's.

There may be some other issues, but that would be a good recommended starting point.

[n2x4]

Thanks Josh, you're just about the only person that's offered any insight on my electrical issues related to my swap

I did manage to fix my problem, my self shutoff switch was crossed over to my fuel pump relay. As soon and I fixed the wires, it fired right up.

Now I can't get the damn thing to turn off! It stays running when the key's out of the car. I imagine it's related to the self shutoff control, but I don't know what kind of signal the ECU is expecting, 12v 0v? any ideas?

Thanks!

[Legacy777]

The self shutoff control can be a little tricky to understand.

I had to stare at the wiring diagrams for a little bit while I was figuring out what to wire to the Link.

There is the main power control setup that the ECU gets, and then the selft shutoff control goes to a separate input on the ignition switch to hold power on to the ECU so it can return things like the IAC valve (stepper motor) back to its start position.

It sounds like either the wiring is different on the impreza or you still have some wires mixed up.

[n2x4]

The self shutoff control can be a little tricky to understand.

I had to stare at the wiring diagrams for a little bit while I was figuring out what to wire to the Link.

There is the main power control setup that the ECU gets, and then the selft shutoff control goes to a separate input on the ignition switch to hold power on to the ECU so it can return things like the IAC valve (stepper motor) back to its start position.

It sounds like either the wiring is different on the impreza or you still have some wires mixed up.

That makes sense.

What about this mysterious diode trick I see mentioned, know anything about it? It looks like it's required for swapping into an Impreza. It was used to help give power to the shutoff switch from another wire. I read some info on NASIOC and here about it, but nobody really is sure of the function.

[Legacy777]

Took a quick look at the wiring diagrams. Your problem with the self shutoff control is probably related to not having the MPFI diode installed.

[n2x4]

Took a quick look at the wiring diagrams. Your problem with the self shutoff control is probably related to not having the MPFI diode installed.

I did add the diode like the little bit of information I found said to do, but it made no difference. I guess I'm unclear as to it's function. I found this pic searching, and i wired it the exact same way shown, taking into account the cathode and anode end positions:

[Legacy777]

Hmmm...

I'm not sure what to tell you. If it were me, I'd just go through both sets of diagrams and see what I can find....

[n2x4]

After about 10 hours of diagram reading, and wire switching, I threw in the towel. I can't get the self shutoff control to work.

However, after reading the wire manual and playing a little bit, I figured out what it does.

The self shutoff wire runs from the main ignition relay, to the ECU. The ECU sends out a signal from that pin (so yes, this is a powered pin at the ECU, not a ground). This signal allows the main relay to close, and sends power to your injectors. The old ECU's couldn't provide enough voltage off that one pin, so using a switched ignition source they wired in a diode that bumped up the voltage on that line by 2V. This allowed the relay to close. Without the extra voltage, there wouldn't be enough juice to close the relay. When the key is off, the ECU sends the voltage for 3 seconds, enough time for all sensors (such as the IAC) to close.

This is not happening in my case. When I connect the ECU to the relay, power at the ECU on the self shutoff pin NEVER goes away. The relay is stuck on, and the car will never shut off. I do know my ecu does work correctly, because if I test the voltage at the ECU on the pin without connecting it to the relay, it says powered for 3 seconds, and then drops to 0v. However when it connects to the relay, power remains constant. Disconnecting the wire does shut the car off.

So I just got one of those red aircraft toggle switches, so that when I turn the car off I flip the switch, it cuts the connection between the relay and the shutoff wire. I hate to do anything that's not factory, but at this point I'm done messing. I gave in. At least the switch looks nice, and it's wired cleanly.

Thanks for all your help Josh!

[Legacy777]

Hmm.....I wonder if the ECU is not getting the right ignition switch information fed back to the ECU.

I'll be curious to see how my link works, as I've wired in a self shutoff control.

[beatersubi]

Is the main ignition relay's ground switched?

[n2x4]

Hmm.....I wonder if the ECU is not getting the right ignition switch information fed back to the ECU.

I'll be curious to see how my link works, as I've wired in a self shutoff control.

I thought that too, but all I can tell from reading the ECU with the scan tool is the ignition switch wire that runs to the ECU controls that switching info. When I don't have my self shutoff wire connected, cycling the key will allow the ECU to see that input turn on and off. I can measure voltage on that wire as well, key on 12V. Key off, 0V. As soon as the self shutoff connects to the relay, there's a constant 12V regardless of key position and the ECU never changes the ignition switch signal on the scan tool. Just a constant on.

Is the main ignition relay's ground switched?

No. On the Legacy the main relay has it's own chassis ground where as the Impreza main relay shares a ground with the O2 Sensor. I made sure my main relay got it's own ground.

[log1call]

"This is not happening in my case. When I connect the ECU to the relay, power at the ECU on the self shutoff pin NEVER goes away. The relay is stuck on, and the car will never shut off. I do know my ecu does work correctly, because if I test the voltage at the ECU on the pin without connecting it to the relay, it says powered for 3 seconds, and then drops to 0v. However when it connects to the relay, power remains constant. Disconnecting the wire does shut the car off."

With the key on you have power going from your ignition, through the main relay and then trying to earth to a ecu pin that has power to it when the ignition is on. So the relay doesn't work because it has twelve volts on both sides of it.
When you turn the ignition off the ecu keeps supplying twelve volts to it's side of the relay for three seconds but the other side, which had also been twelve volts, is now a potential earth through anything fed off ignition, so the relay turns on for three seconds and your fuel pump goes.

Well that was before you rewired the rewire.

You have a mismatch of ecu and body wiring. More specificaly, you need to rewire the main relay.

I was looking at several different diagrams for legacy and imprezas the right years, I noticed that there are at least three variations of main relay feed. Some main relays were fed ignition directly and earthed via the ecu.
Others relays had a diode and two power feeds(like yours). They were fed power by the ignition switch and the ecu, and then only the by ecu after the ignition was turned off.
Then there were others that were fed power from two different places on the ecu... one of them through a diode and the other not. Both these last two systems had an earth for the relay.

Also... You say you just cut the ecu's plugs and fitted the new ones. I assume the engine/body plugs were the same shape and fitted straight on to the new motor. Are you sure the pins were the same positions in the old and new engine plugs?

[vrg3]

Let's take some of the mystery out of this diode. It's actually very simple.

The EJ22T ECU is designed to keep the engine management system powered on (via the ignition relay) for several seconds after you turn the key off. It uses this time to save parameters that it's learned and stuff like that.

The way it does is is via the self-shutoff control pin. After the ECU notices the ignition switch is turned off, it stops running the engine and also starts sending +12v out the self-shutoff pin.

So, the wiring has to be set up so that the ignition relay is active when either the ignition switch is on or the self-shutoff pin has 12 volts on it. This is done using the MPFI diode. The relay's coil is connected to the self-shutoff pin, and then a diode also allows the ignition switch to power the coil without the ignition circuit getting fed by the self-shutoff signal.

You could just hook the relay's coil to both the ignition switch and the self-shutoff pin, but then you'd have the problem that the self-shutoff pin would feed into the whole ignition circuit and then a) everything else ignition-powered would stay on and b) the ECU would think the ignition switch was turned back on.

That last part sounds like what's happening in this case, doesn't it? If you have it all wired correctly, a bad diode (fused so that it was simply a short in both directions) would cause exactly this to happen.

So, I think you may have a bad diode. I'm sure I have a spare stock MPFI diode, so if you can wait a couple days email me your address and I'll drop one in the mail. Otherwise, you should be able to pick up a power rectifier from Radio Shack; they sell a 1N5404 diode that is quite sturdy.

But, I'm concerned it might not be wired correctly. If you simply remove the MPFI diode, the ECU shouldn't power up at all! Did you wire the ignition relay's coil just to ignition-switched power then?

The ignition relay has a six-pin connector:

Code: Select all

 __   __
| 1|_|2 | 
| 3   4 | 
| 5   6 |
|_______|

Pins 1 and 2 are the relay's coil.
Pins 3 and 5 are one of the pairs of contacts.
Pins 4 and 6 are the other pair of contacts.

One side of the coil should be connected to ground. Doesn't matter what ground as long as it's a good solid ground.

The other side of the coil should be connected to two things, and two things only: 1) the ECU's self-shutoff pin, and 2) the cathode of the diode.

The cathode of the diode should be connected to nothing else. The ECU's self-shutoff pin should be connected to nothing else.

The anode of the diode should be connected to an ignition-switched source of power. Since the current draw is very low, and since you're already in the area, what I recommend is connecting it to the ECU's ignition switch pin. Do not connect the anode of the diode to anything else.

I don't know if addressing this problem will help with the problems we're discussing in your other thread, but it certainly can't hurt. This isn't complicated; we should be able to figure it out.

[log1call]

Hmm. the ecu has power coming out of it until three seconds after ignition off if it's not connected to the relay, which is as it should be. That's been verified.

When he connects it to the relay there is power to the self-cut pin all the time... that confuses me. If it is only when the key is on then I suspect it is because the car is wired to earth the relay via the ecu and the ignition is feeding the relay which is feeding the ecu's self-cut pin. If it is really on the whole time, key on or not, then there is some other problem altogether in the wiring.

I think if the diode was shorted the key should still turn off the feed to the ecu.

We need to check where the relay gets power from all the time, (which is what seems to be implied by the ecu's self-cut pin being always live when connected to the relay).

I'm assuming in all this that the "L" model is non-turbo.

I have found before that there can be a little over a years difference in models being released from one continent to an other.
I don't have a 91 diagram but I do have 90 and 92 diagrams and I notice that from one to the next they change the system of relay operation. I'm guessing that the car in question is wired like the older model.
I think I need to see exactly the wiring diagrams(confirmed to be applicable to these two vehicles), before I can comment further.

[n2x4]

Thanks for the feedback guys - I'm going to need to digest this info and revisit my wiring.

log1call- I have my specific FSM diagrams, let me dig them up and I'll get them posted.

Vikash - The diode isn't bad. I actually picked up a few of the radioshack ones before I found the factory diode in my harness.

[vrg3]

Well, do let us know what you find when you revisit your wiring. Like I said, if you have the correct EJ22T wiring but no diode, theECU shouldn't power up at all.

[log1call]

Yes please N2X4, pdf manuals are always handy. Where are you going to post it and I will keep my eyes open since this thread looks to be finishing. Or you could message me where it is. Either way, thanks in advance.