98 legacy NO SPARK see post for deetz

Alright Lads and Gents. i have a 98 legacy gt. redid all the wiring harness with a 97 gt harness. dropped a ej22 phase 2 motor in it. repined some plugs put a new clutch in and a new ignition switch cause previous owner lost his key. and i cannot seem to get spark. i took the fine toothed crank sproket off and put on the 6 tooth one. i also check and i am getting injector pulse the starter is turning the car over. all my interior jiggers and dodads work. but i cannot get spark. i had to take the 4 prong coil pack off the intake manifold and put a three prong on it. (when i say prong i mean for the harness to plug into.) i may have plugged the wires in upside down when i first put it on. if that is the case could i have blown a fuse or fried that coil pack? if so no big deal i have another but my question is why the hell am i not getting spark. also i have not repinned the IAC sensor but i did repin the TPS. let me know what you guys think. thanks.

This sounds scary.....as you've probably read the Phase 1 and 2 don't play well together - @ least electrically / harness/ ECUs. If you're following somebody else's info that it works, post it up.

 

- The '97 GT and '98 GT harness should be the same - Why swap? Did it run w/the '98 harness B4 the 2.2 swap?

 

- The Phase 2, 2.2. is looking for a different ECU and vice versa.

 

Sorry...no quick fixes.

Td

Okay so what happened was. The 98 legacy was merged with a Wrx harness. I replaced all electrical units except behind the dash harness. The bulk harness rear harness and engine harness were all changed I'm also using the same ecu for that harness. Everything works on the motor except for spark. I have repined necessary plugs but I will get the phase 1 intake if need be. I fear even at that it won't fire. I'm going to try a different coil block and see what happens tomorrow if I get spark then I will know what the issue is.

The two different computers for this swap use different crank position sensor wheels? Are you sure you got them right way round?

 

I had a problem once where my crank and cam sensors were a bit out of phase, and the computer would double fire some injectors at low RPM, some double all the time, and two wouldn't juice at all above 3450. It was also dropping spark on 1 and 4 a bunch of the time. My point is, if the computer can't sync with where the physical engine bits are in the cycle, all hell will break loose in the software including wrong or no spark.

@doublechaz how did you fix that. Yeah I'm sure o have everything timed correctly. and I do have the correct cps and crank gear.

i thought the ej25 and ej22 had different igniters in them?

okay update. i went and got an ej22 ecu out of a standard impreza out of the junkyard and i have spark now but for some reason i still cannot get it to start any ideas why it wouldnt start. this happened before and my motor was out of time. but i retimed the motor twice to make sure it was in and it is. but it still will not start im getting spark and everything is plugged in except the IACV but that shouldnt keep it from starting. any ideas whats going on i have air fuel and spark.....

I fixed mine by getting in there and finding that the cam with the position sensor was 2 teeth off. Seems that really confused the computer about when to juice and spark.

 

I'm not deep into the Subaru mysticism with engine swaps, but if you have a 2.5 with matching computer and you take out the 2.5 and put in a 2.2 don't you also have to swap the computer? Then there are harness pinout differences between the two.

 

Maybe I didn't follow, but I thought you said you went from 2.5 to 2.2 and at first didn't switch the computer.

 

I'm sure I read somewhere that several different chasis and motors use the 84 pin connector that is the same size and shape, but they use different pinouts. If you are mixing and matching those then there are a few subtle pin changes to make.

 

I know that I got a computer from an Impreza of unknown year and engine and put it in my 2.2 Legacy 95 to check if a problem I was having was from the computer. It started and ran fine. Everything was normal until I got back from around the block and I couldn't shut it down. Key to off and out of the ignition and the car was still happily running along. I had to pull battery neg to stop it.

okay so Update.

 

The car is running. tans is bolted up ect. my steps of action for anyone trying to do a phase two into a phase 1 are as follows....

 

1. when you move the motor over the trans will still bolt up. BASICALLY ANYTHING THAT DOESNT RUN OFF OF A SENSOR IS GOING TO WORK ANYWHERE YOU PUT IT. you can swap over intake manifolds but you need the adapters and to buy 2 sets of gaskets and wait for the adapters would but just as much of a pain in the ass as it would to just repin some plugs.

 

2. you need the wiring harness for the engine that matches to the car you are putting the phase 2 in. ex:(i used the 98 2.5 legacy harness for my ej22, so the computer would know how to read my plugs and such)

 

3. Your fuel injector plugs will be different so you will have to repin them ex:(take the phase 2 harness fuel injector plugs unpin them, take the phase 1 harness and do the same. take phase 2 fuel pump plugs and put the wires from the phase one harness into the plug. then plug in.

 

4. your throttle body will also be different. there is an "IACV" on top of your intake manifold. if you take the phase 1 throttle body and put it on the phase 2 intake you wont have any issues with the IACV.

 

5. on the brake master cylinder i had a plug that was different it called for a skinny 2 prong plug and the one i had looked like the pigtail on the wheel speed sensor. luckily on the 98 legacy and the 00 impreza the plugs were the same, its only 2 wires so i sniped and rewired them. i don't know why i didn't just do a pin job on them so dont ask lol.

 

6. If you find that your running a 2.2 into what should be a 2.5 you're gonna run into some sensor issues and ecu issues. i couldn't get spark. at all when i first dropped it in. but i changed the crankshaft sproket from a thin pronged to a 6 thick prong (will provide pictures soon) after that was done still no spark. so i took the 2.5 ecu out because i figured it was looking for a dual cam 2.5 camshaft sprocket. got the 2.2 ecu out of a standard, subaru impreza, and plugged it in. HEY! theres spark. but no fuel. so i changed the camshaft sprocket from the phase 2 2.2 to the phase 1 2.2 cam sprocket. check my fuel lines realized they were back wards swaped them around to the right spots. turned on fuel pump cranked it ... ... ... boom started right up.

 

6. I should have put this at step 2 but the hell with it. for some reason i hadn't been able to get my starter and relay to properly come on when i turned the ignition... so. what i did was wired a toggle switch into the plug. how to do that if you pull the green plug behind the dash bolted to the side of the car plugged into a relay and you look at it, there are 4 holes. the 2 on the right hand side are the switch on. that is the circuit that you need to complete to make that fuel pump turn on the other 2 wires are useless.

to reply to double chaz im not sure what the issue is in your subaru. but i turned mine on and off multiple times and i didnt have any issues with it shutting down.

That was pretty weird. My computer is marked:

2D

22611 AB414

JA18000 RT8 4Y21

 

The other computer of unknown origin that didn't work is:

Z7

22611 AC150

A18-000 D03 4Y11

 

Anyway, the computer wasn't my problem I was just testing. The actual problem is all fixed and I've been tearing around town again.

That's awesome to hear! I didn't think this build was gonna work!

I'm having another issue. Why when I start up my car does it Rev to 2k stay there and when I touch the accelerator it bogs down to almost a shutoff.

Congrats on getting it to work.....hopefully you can figure out the rest of the 'bugs' as they occur.

 

 

The other computer of unknown origin that didn't work is:

Z7

22611 AC150

A18-000 D03 4Y11

 

 

Sorry to HiJack, but for future reference Google found THIS.

 

It's a 1995 IMPREZA EJ22 / 2.2 ECU. Which means it's an OBD2.

Thank you WTdash! Appreciate the positivity. If you know any leads on the idling and acceleration issue please input. Thanks!

apologies if this has alread been said.

 

1/ the ecu MUST match the wiring harness in trhe car.

or you will not get the sensor info to the correct pin input on the computer.

the engine May run fine, may not, but there WILL be an error code.

 

2/ the pahse 2 ej22 will not run in a phase 1 car UNLESS you swap on the phase 1 cam and crank sensors that match the ECU.

and even then you will have an error code, for something.

 

i have not read all of the posts above so i may be giving advice that has already been given.

but if you have not done these things listed, the car will not run.

 

it is a mistake to ever swap the computer unless it is dead.

it is HUGE mistake to swap the wiring harness unless you are installing a turbo.

it is mistake to swap a phase 2 engine into a phase 1 car, EVER.

 

good luck.

If you know any leads on the idling and acceleration issue please input.

 

this sounds like a vacuum leak.

check to see that all hoses are connected.

any error codes.??

Alright update. I put the 2.5 ecu on it and it seems to be idoling just fine now. But for some reason. I still cannot press on the gas pedal. When I do I hold it in and it starts backfiring horribly and it sounds like it's only firing or something on 2 cylinders. Help appreciated thank you.

Johnegg yes I did all of these things so far. I know I should not have touched any of it but I did. Haha I wanted the car back to stock and I got it to run and stuff. And everything except powerlocks on the doors work. But I can't get this acceleration issue figured out.

When my ECU was losing sync with the physical motor bits it ran pretty good between 1500 and 3450 RPM, but below that it had no torque and above that the motor virtually shut off. If your ECU doesn't like something about cam and crank sensors it may be losing sync as well. I'm amazed mine would even run, and nearly code free. Just an occasional cyl2 misfire.

 

Do you have any kinda advanced tools to look at what the computer is doing? I limped through diags with a multimeter, a timing light, and some LEDs. I didn't get it figured out until I got an oscilloscope.

 

One useful bit was trying the timing light on each plug wire. I had solid steady spark on 2 and 3, and some skipped sparks on 1 and 4.

 

Another useful bit was using some computer ethernet wire to connect an LED to each injector. It was pretty hard to see the individual flashes, but I could see that two were brighter than the other two, and above 3450 two didn't light at all.

 

With the oscope I could look at the shape of the signals and saw that my injector impulses were all whacked and out of rhythm so I knew I was losing sync.

Newest updates. I have the car running and moving and idleing sort of normal. Issue is its got a hell of a miss. When I press on the throttle it seems to be pre-deting. I'm wondering if it's a timing issue or if it's the crank sprocket ? Can post pics of the sprockets I have on car and ones I have off of car. What do I do to fix this plz help. 1 I don't have any o2 sensors in. Nor is my vehicle speed sensor in.

Hm. Do you have a timing light that you can use? Check each plug wire for even steady spark. If you had 4 timing lights you could do all four at once and check their sequence relative to each other, and with added audio you could host an EDM party.

 

If you could try the LEDs on the injectors that would be good as well.

 

I'm pretty far from an expert, but it sure sounds like your computer is losing sync. All manner of crazy things can happen then. Sparks at the wrong time, wrong fueling.

Why would the computer lose sync? What causes that?

The computer uses the crank position sensor and cam position sensor to know when to squirt the injectors and when to fire the spark. In my case the cam with the sensor had jumped 2 or 3 teeth on the belt. From the point of view of the computer, because it is so fast, the engine turns in super slow motion. It senses the pattern of pulses from the crank position sensor to know where the crank is in 360 degrees, and the real time delay between them to figure out the RPM and then the cam position sensor tells it if it is in the even or odd 360 degrees (because of the four stroke engine thing). Now it knows when each spark and injector impulse should happen. If either of these sensors is not working, or if the magnetic bit in the cam sprocket, or the teeth on the crank sprocket are out of line (or they are the wrong sprocket compared to the computer that is connected) the computer will get confused about where in the 720 degrees of rotation it is. The result is mis-timed sparks and injector impulses.

 

In my case it was something like this:

Pretend the injector impulses should be at 0 degrees, 180, 360, 540. Not real numbers, but just run with it.

3 teeth late on my cam means something like 25 crank degrees late.

As the crank is turning and the computer thinks to itself, 'right, just coming up to 0 degrees fire the first injector NOW. Then it knows that things are still turning until suddenly the late running cam sensor ticks and the computer says, 'Oh, crap! I was confused, just *now* I'm at 0 degrees so fire that first injector again. Next, as you noticed there are lots of teeth on the crank sprocket, it thinks, 'ok, wait I was just at 0 says the cam sensor and now the crank sensor says I'm at 90 degrees X milliseconds later, so I'm turning at 3800 RPM. That means I need to advance the spark by Y degrees and it's time for that so fire! Then the next crank sensor tooth comes along and the computer says, 'WTF? the time delay since last tooth means that now I'm only turning at 1400 RPM. Change spark advance again.

 

As far as the computer can figure out it is time travelling forward and backward. I was getting 6 injector impulses per cycle instead of 4. I don't know what I was getting for spark, but it was doing fuel cut at 3450 RPM which because of the sync loss the computer thought was high enough RPM to be hitting the rev limiter.

 

TLDR;

If the crank or cam sprocket you are using is a different design from the one that your computer expects, or if your timing belt is installed out of time then the computer can't figure out when things should happen, but it will still try like hell and the result is it kinda runs but is really screwy.

Okay so what about putting a 2.5d computer on a single cam motor? I have an ej2.2 computer but. I was told the 2.5d computer will work. That being said I changed the cps and the cam sprocket to accommodate. Dont get me wrong I didn't put a dohc cam on that part but I did put the phase 1.2.2 sprocket on it.