Another DOHC EJ25 Head install on EJ22T block-Misfire-FIXED!

[wtdash]

Edit: As of 12/2010, I couldn't get the Forester tuned properly on the Greddy emange ultimate - 2 different tuners couldn't figure it out, so I put it back to stock and sold it. The engine is now going in my '90 Legacy, which I originally swapped w/ a stock '93 EJ22T. The main differences to the below info are a JDM IM (no EGR to mess with) and a '91 turbo Legacy harness.

Edit 2: After the engine went into the Legacy I was still having issues, so I had a Comp/LD test done and Cylinder 2 (or 4?) was low/leaking. The piston had a broken ringland and my Subaru Tech was very surprised it hadn't blown up completely. I Had the engine rebuilt w/stock pistons (budgetary restraints!), and then sold it as I'd moved on to the '04 Forester......so, it's very possible the engine was having the same issues while in the Forester and I just didn't think to check it out.

This thread may be updated w/the Legacy install on Page 2.


Yes, this has all been covered, but hopefully there's enough info in this one thread to consolidate some of the info on the board. If you see anything questionable - Speak Up - I don't want to add to the misinformation already out there.

First, I'd like to Thank Legacy777, Napphappy, Grayguy, All_talk, Matt Monson, PhyrraM, Skid542, Douglas Vincent and others on here for making this possible...I'm probably one of the least mechanically inclined members on here, but have managed to swap my '90 w/an EJ22T and now working on another Subie project all because of this Board.

My current project is a putting an EJ22T block w/the '96-'99 EJ25D DOHC heads into a '98 Forester, so the wiring in the pics is relevant for OBD2 cars (roughly '95-99) only, but all the plumbing should be the same (except for crossover pipe).

I'm also running JDM EJ20K turbo camshafts. They fit just like stock in the '97-99 USDM DOHC shim over bucket lifters (not sure about compatibility w/the '96 DOHC hydraulic lifters - HLA ).


I wanted to run a thinner headgasket to keep the CR up around 8-8.5:1. Based on a few threads over on NASIOC I read that others had used an EJ22e (non-turbo) HG. But when I got ready to install it I saw this:


And decided too much of the passages were covered, so I went ahead and just installed the EJ22T HG (pic below). For more info check this THREAD on NASIOC. Due to the risks of running too thin of a HG I'm NOT recommending this option, but if you do your research you can make up your own mind.


I decided to use the EJ25's water cross over since the sensors on the intake manifold would match the EJ25's harness (I'm not sure if this will be a good idea as it puts the 2 sensors in a very tight area where the turbo, etc. are all installed). This is ONLY for the EJ25 harness. Re-use your EJ22T pipe w/its matching harness connections or extend the wiring, if necessary (EJ22T on top):



I kept the pipe from the EJ22T that runs from the water pump to the Turbo's coolant tank, which I hope will will mount in the same location as my '90...we'll see. I did use a '92-94 EJ22T water pump. Others have used the non-turbo pump, but I don't know how it was plumbed:


To get the other heat/coolant line I modded the EJ22T's line. I cut and bent the tab, drilled the hole bigger and got lucky to reuse the oil galley bolt. Hopefully, it'll be close enough when installed (that wire loom cover was removed - just there temporarily).


Cut/Bent tab:


Here's a comparison of the coolant pipes that come from the water pump. The bigger diameter pipe is the EJ25 and although it follows the same path as the turbo's, it's for the heater hose - not the turbo coolant tank. As mentioned above, I ran the EJ22T's heater hose pipe from the water pump, under the head, like stock on the EJ22T.
EJ25 on bottom:


EJ25 on top:

IM installed:



STi engine mounts and exhaust manifold installed:


EDIT: The PCV and Breather lines are two separate systems from what I've since learned....do not mix/match - keep 'em separate or you might create engine pressure issues.

Unexciting as it is, I got the PCV/breather hoses where I think they'll work- I'll clean it up after installed in the car. The PCV and breather hoses will all go thru the Helix catch can, which I've read is sub-optimum. {Don't do this!!>>>>}. The PCV and breather hoses should each have their own catch cans, if not going back into the intake {YES - 2 cans!}:


I ran the hose off the PCV out thru the pass. side of the IM....wait and see if that'll work w/the PS lines, et al:


Installed my 1st DOHC Timing belt. Used clamps to hold the belt in place on the pass. side and to hold the sprockets on the driver's side. Wasn't as big a deal as I thought it'd be. Used the marks on the timing covers and then counted the teeth - 3x - to verify. I had another 'gotcha' as the EJ22T's Tbelt tensioner pulley's bolt stripped out. Thankfully I had another EJ22e sitting in the corner. Since the bracket that the pulleys/tensioner attach to is replaceable, I just swapped in the other one. I could have also used the newer style tensioner and pulled the bracket from the EJ25:



Here's the plumbing for the turbo. I bought all the WRX piping from NASIOC members' listings. Oil drain/return, oil inlet, and coolant pipes (HERE is another other option and HERE is a good comparison of NA vs turbo heads). Since the NA heads don't have the oil return or coolant pipes threaded (just the bosses (?) for the cooland and return) I had a local Subie shop drill and tap those for me. I just removed the bolt from the head for the WRX's oil inlet line. The oil return is a TD04 coolant line bent and cut, w/the OEM rubber return hose connected:


Macgyver'd oil return line:


I used the WRX's turbo mounting brackets, as the EJ22T's didn't fit due to the DOHC oil inlet's location (it's on the end of the SOHC EJ22T's head). It wouldn't take much to modify it, but since I knew ahead of time (research!), I bought the plug-n-play option. I also used a WRX up pipe as I may need its EGT down the road (it's plugged for now):

EJ22T:

WRX 2-piece version:


If you've got any recommendations or need pics, speak up quick as I hope to get this installed this weekend!

Thanks,
Td

[Legacy777]

Nice write-up Todd!

[Mattheww044]

very nice write up. BTW, you are definately more mechanically inclined than me, but I am learning!!

[douglas vincent]

Looking good!

[skid542]

Awesome! You're going to have another mean machine once you're done. I'll be anxious to see it in action . Definately keep me posted on progress and if you need any help with it.



BTW - you're still underestimating your abilities.

[Grayguy]

how hard was it to find the EJ20K turbo cams? I should really get a set of them for my DOHC's

[wtdash]

how hard was it to find the EJ20K turbo cams? I should really get a set of them for my DOHC's

I got 'em off NASIOC. There were a few sets for sale when I bought mine.

Td

[wtdash]

BTW - you're still underestimating your abilities.

Let me clarify....you all UNDERSTAND HOW this all works ...I'm just putting together my Legos and hoping it RUNS!

[wtdash]

very nice write up. BTW, you are definately more mechanically inclined than me, but I am learning!!

Just keep @ it!

[wtdash]

Although this is a for a Forester, it's got the EJ22T block.

And this site is the main reason I attempted this engine configuration, in the 1st place.

AND I need some input, please.


Here's the good news from yesterday's visit to Maxwell Power in Arlington, WA.

This for my '98 Forester w/the stock ECU (OBD2) running a Greddy Emanage Ultimate piggyback. It's the MAF-based system, using Doug V's modified MAF .

Running the 90° elbow TD05-16G and 550cc injectors.

1. About 242/239 @ 13 psi DYNO SHEET

My goal was 250/250 @ 18psi - maybe even 20psi- which we almost reached @ just 13, so there's a possibility I can see 275/275.


The bad news:
1. Couldn't duplicate it consistenly. Most of the time the car would hit about 4800 and misfire and fall off.
2. We replaced plugs , gapped to .030 - No Change- (I had put 300 miles on mine getting there w/out the correct tune so we thought 'maybe' that was the issue)
3. We disconnected the EGR vac lines - Dom (tuner) thought there might be an issue - No Change
4. We changed my vac lines as the pressure sensor (the one that connects to the stock map sensor) was on the same vac line as the stock Fuel Press. Reg.. - No Change.

After over 4 hours on the dyno, we gave up.

The car runs/idles better (not perfect), but whenever I get to about 4800 RPM, it starts to stumble and if I keep my foot in the thottle it'll throw the CEL for 'misfire' on all 4 cylinders....plus occasionally I get the P0420 Catalyst System Efficiency Below Threshold which is probably just a secondary code due to the Misfire. Since my car's ECU is old, Dom couldn't 'read it' real time to see exactly what, if anything, the ECU was doing to cause the misfire. His scan tool is only '99+.

Dom says it's most likely electrical/electronic vs. mechanical. The AFR's were fine right up to the point of failure...it's like something in the ECU says "No More" and causes the misfire, but Dom really didn't know what to check.

His suggestion, as the EMU is just a piggyback, is to go Standalone, but that's not in the budget, plus I'd likely lose some of the car's convenience features, such as Cruise and whatever else the standalone can't handle....plus the wiring would likely be too much 4me....and then I'd still need to get it tuned - again. He also mentioned returning to the stock injectors, adding a 7:1 Rising Rate FRP (RRFPR), and running w/the Greddy EMU that way....since I'd still need to control timing, but then I'd still have to get it tuned.

Neither of those is appealing, so I'll likely spend some (more) time online looking for issues, and try replacing a few sensors on the car that could be a factor.

I had an issue on my old Legacy, but w/out the CEL. When it got warm out (above 50) the intake hose from the MAF to the turbo's compressor inlet got soft and collapsed under boost...I didn't figure it out 'til I re-did my intake and got rid of the long hose and substituted w/a shorter one. But the issue on the Legacy never thru a code and was purely mechanical.

If you have ANY ideas, please feel free to throw 'em out there.

Thanks,
Todd

[Justins91ss]

Doms An awesome turner he did my wrx/sti... And well respected on the Impreza forums.. Too bad he cant tune my 91. Oh well am in the market for a 2nd subie right now. Gl on the build

[Legacy777]

Todd,

As you pointed out, your AFR's got way too lean. Personally, I think you're a little too rich....9.5-10 AFR is pretty rich, unless you're compensating for something else not being on the money. Around 11 AFR is good.

You mentioned you're running the stock ECU. Is this a non-turbo ECU? I have heard of issues where they just don't like to see boost, and will just cause all sorts of issues when you try and put positive manifold pressure to them.

I know you may not want to hear it, but I'd probably suggest looking for another suitable turbo ECU. You may want to do a little more research on the positive manifold issue with a non-turbo OBD2 ECU.....but I seem to recall reading/hearing something about that for the imprezas RS's of the similar vintage putting turbo swaps on their cars.

[wtdash]

Josh,
I'm running the stock ECU (Non-turbo), but w/the Greddy Emanage Ultimate...supposedly one of the better piggy-backs...albeit hard to tune...from what I've read.

There is no direct turbo ECU I could swap in, that wouldn't include a full harnes swap, etc.....not something I'm willing to do. The JDM/ECM/Aussie Turbo ECU's won't work w/the USDM harness (or plug in), as far as I know.

I bought the Greddy from a guy who ran a very similar setup on his '97 LGT w/the stock non-turbo ECU: EJ20 w/EJ25 intake mani/550's/VF28. He was right @ 250HP @ 18 psi, and it ran very well- great DD (He's since done a WRX swap into his '97). So I know it's doable...but he spent a LOT of time trouble-shooting and tuning it himself.

The rich AFR's leads me to believe there was still something 'not quite right' even before the car hit 4800, but since the tuner couldn't get past the 4.8K 'limiter' we never got a proper tune.

Thanks for the lead on the Impreza 'positive mani issue'. That's the 1st I've heard of it, but willing to look into it.

Thanks for the reply,
Todd

[Legacy777]

I wouldn't get hung up on the HP numbers so much right now. Looking at the AFR's, you've got something else going on, either tuning or something else.

What parameters can you monitor from the OBD2 port?

[wtdash]

What parameters can you monitor from the OBD2 port?

NADA...I don't have any software that let's me connect. I posted on USMB, and the Subaru Select Monitor was mentioned, but don't you have to be a dealer to use that?

I know there a couple of options such as the ScanGaugeII, but not sure if it's not more than just a 'code reader'. EDIT: It apparently can @ least do some data logging: "the specs indicate that it does good things like take a snapshot of engine data at the moment a code is thrown."


Ideally, I'd want to see real-time timing/fuel trims/etc. I would think? The guy I bought the Emanage from had a scan tool....I'm waiting to hear from him what it was he used.

Td

[skid542]

Todd,

Good numbers to start off with! Sounds like you're on your way albeit with a couple of problems.

The issue that you're having - it sounds like this is an every time you hit 4800 rpm? Can you get past 4800 if you're not at full throttle? What does the car do at half throttle, light throttle?

Which injectors are you currently running?

Do you have a link handy of the mod on the MAF you did?

I'm not sure what to think of Josh's thought about the positive pressure since you're making positive pressure well before 4.8k... hmm...

You're running the LC-1 WB correct? Does it allow data logging of other parameters? Does it have any user inputs you can feed into it and log?

Are you having any other issues with the car?

[Legacy777]

Lee,

He's running the 550's.

The MAF he's running is likely one Doug machined out to be a larger bore, which increases its range.

[Legacy777]

Todd,

Since it's OBD2, you will be able to read some of the ECU parameters. That vintage ECU is not as good as the newer ones, but you should be able to get some data. You need a scan gauge or an OBD2 reader that interprets the ECU signal and converts it to data that can be read by the scanner programs on your computer. The interpreters are called ELM. I saw an ELM327 blue tooth earlier. That'd work, and then you can run whatever software you want. Some of the free stuff is ok, but the stuff you pay for will give you more features.

Here's one
http://www.carpartslights.com/elm327-bl ... -p-28.html

The drawback with the scan tool is you can only log/view a certain amount of parameters without the refresh rate going to crap.

What I'd be looking for are clues in any of the various parameters that seem outside of what they should be. That's unfortunately part of the troubleshooting process.....until you see, you may not know what you're looking for.

[wtdash]

Todd,

Good numbers to start off with! Sounds like you're on your way albeit with a couple of problems.

The issue that you're having - it sounds like this is an every time you hit 4800 rpm? Can you get past 4800 if you're not at full throttle? What does the car do at half throttle, light throttle?

Which injectors are you currently running?

Do you have a link handy of the mod on the MAF you did?

I'm not sure what to think of Josh's thought about the positive pressure since you're making positive pressure well before 4.8k... hmm...

You're running the LC-1 WB correct? Does it allow data logging of other parameters? Does it have any user inputs you can feed into it and log?

Are you having any other issues with the car?

Lee,

Yes, it's Doug V's enlarged MAF. He's said it's good to 400HP...plenty for my needs! And STi 550's (side-feeds).

The car drives 'just OK' before 4800. I have a slight flat spot after shifting to 2nd, but not in 3rd or 4th. Idle varies a little. I've not tried to get past 4800 w/out boost, but the car Revs past it in Neutral. Otherwise, the car easily/readily boosts to 13psi via the MBC - as long as I shift before 4500 RPM....I'll be good @ 'short-shifting' before long!

Maybe 1 out of 5 tries I can get past 4800? I've not tried it regularly, as it's a pretty violent shake when it misfires...can't be good on the engine, I wouldn't think.

It starts harder than my Legacy (which fires almost immediately) and cranks slowly...which might be related to the battery or ground.

I haven't hooked up the LC-1 (didn't have an extra bung 'til just before getting the tune) as I was expecting the tuner to handle it on the Dyno.

Thanks

[wtdash]

Todd,

Since it's OBD2, you will be able to read some of the ECU parameters. That vintage ECU is not as good as the newer ones, but you should be able to get some data. You need a scan gauge or an OBD2 reader that interprets the ECU signal and converts it to data that can be read by the scanner programs on your computer. The interpreters are called ELM. I saw an ELM327 blue tooth earlier. That'd work, and then you can run whatever software you want. Some of the free stuff is ok, but the stuff you pay for will give you more features.

Here's one
http://www.carpartslights.com/elm327-bl ... -p-28.html

The drawback with the scan tool is you can only log/view a certain amount of parameters without the refresh rate going to crap.

What I'd be looking for are clues in any of the various parameters that seem outside of what they should be. That's unfortunately part of the troubleshooting process.....until you see, you may not know what you're looking for.

Good info, Josh...thanks for the links....Looks like I have 'homework'

Also, I just looked @ the Revscan software that Revtronix put out b4 going AWOL. It has an "SSM-OBD2" setting...I'll see if my cable will connect....maybe get lucky?

Td

[skid542]

Hmm... okay.

Do you have any plots or correlations on the AF's to indicate if it's going too lean or way too rich when it misses? It would be helpful to know if it's a fuel cut or an overload to sort out which system is acting up.

The MAF's are picky and I'm wondering if there was any damage to it during the modification or if there some strange thing going on in the air stream that's upsetting it. A bad/intermittent spike in the MAF signal could really screw things up. You are pulling a significant amount of air through them as you get up there in RPM. I don't know, but a thought.


Have you checked your coil pack to see if it's in shape? If your coil pack is on its last legs it may not help the slow start problem you're experiencing.

Did you try changing your ignitor?


Ultimately there's still only three main systems that you have to worry about - fuel, spark, and air. Air I don't believe is your problem. Fuel could be a number of things but knowing if it's too lean or too rich will help.

And then there's spark. If you had a problem with a cam or crank sensor you'd have problems throughout the RPM range. So that pretty much leaves the ECU, ignitor, coil, wires, and plugs. You already changed your plugs and if your wires were a problem you'd see it across the RPM band. I don't think you'd have an issue in the ECU that would be upsetting your spark unless it was some how waay pulling timing.


Some good logging will help. I don't think your existing cable will work but the software may if you have a compatible cable. Not knowing the Revtronix software I can't really say how feasible it will be to get a compatible cable or not. But your LC-1 may have some options for you as well to log specific signals like MAF voltage and such. Knowing injector duty cycle through the event would be helpful but only a scanner could give you that.

I'm throwing some things out there and could be off base but it will help narrow things down.

Finally though, I do not know why running the stock injectors with a RRFPR would be helpful over your current side feeds. Your modification to the MAF is not directly proportional to a straight scaled fuel pressure/flow increase. You still need to scale your MAF signal to the ECU with the stock injectors and a RRFPR. As long as you know the injectors are good then I wouldn't think they're your problem.

[wtdash]

Here's one
http://www.carpartslights.com/elm327-bl ... -p-28.html

JC,
This looks like a very good option...based on the website info:

Display current sensor data, including:
Engine RPM
Calculated Load Value
Coolant Temperature
Fuel System Status
Vehicle Speed
Short Term Fuel Trim
Long Term Fuel Trim
Intake Manifold Pressure
Timing Advance
Intake Air Temperature
Air Flow Rate
Absolute Throttle Position
Oxygen sensor voltages/associated short term fuel trims
Fuel System status
Fuel Pressure
Many others...

Thanks,
Td

[jaybirde508]

Todd,
dont know if your issue is timming related ,But one solution would be a stic -super igntition timming controler the first "standalones" were compriised of a stic & a safc one for timming one for fuel many have done it this way with non-turbo ecu too compesate

Just my .02

Jay

[wtdash]

Do you have any plots or correlations on the AF's to indicate if it's going too lean or way too rich when it misses? It would be helpful to know if it's a fuel cut or an overload to sort out which system is acting up.

Just the Dyno chart above...It is running Rich (even for a turbo) when it Misfires.

The MAF's are picky and I'm wondering if there was any damage to it during the modification or if there some strange thing going on in the air stream that's upsetting it. A bad/intermittent spike in the MAF signal could really screw things up. You are pulling a significant amount of air through them as you get up there in RPM. I don't know, but a thought.

I've got the stock MAF and a WRX MAF (which is what worked on the other guy's LGT)...just need to wire it in.

Have you checked your coil pack to see if it's in shape? If your coil pack is on its last legs it may not help the slow start problem you're experiencing.

I'll try the one from Legacy, as it seems fine.

Did you try changing your ignitor?

No...but I'll add it to the list

Ultimately there's still only three main systems that you have to worry about - fuel, spark, and air. Air I don't believe is your problem. Fuel could be a number of things but knowing if it's too lean or too rich will help.

And then there's spark. If you had a problem with a cam or crank sensor you'd have problems throughout the RPM range. So that pretty much leaves the ECU, ignitor, coil, wires, and plugs. You already changed your plugs and if your wires were a problem you'd see it across the RPM band. I don't think you'd have an issue in the ECU that would be upsetting your spark unless it was some how waay pulling timing.


Some good logging will help. I don't think your existing cable will work but the software may if you have a compatible cable. Not knowing the Revtronix software I can't really say how feasible it will be to get a compatible cable or not. But your LC-1 may have some options for you as well to log specific signals like MAF voltage and such. Knowing injector duty cycle through the event would be helpful but only a scanner could give you that.

The LC1, as far as I know, only logs AFRs. It can be connected to the Greddy, which the Greddy can use in its 'Auto Tune' function, which is OK for getting close but not for fine-tuning.

I'm throwing some things out there and could be off base but it will help narrow things down.

Finally though, I do not know why running the stock injectors with a RRFPR would be helpful over your current side feeds. Your modification to the MAF is not directly proportional to a straight scaled fuel pressure/flow increase. You still need to scale your MAF signal to the ECU with the stock injectors and a RRFPR. As long as you know the injectors are good then I wouldn't think they're your problem.

100% agree.

Thanks Lee!

[wtdash]

Just wanted to add - The car revs freely to redline in neutral AND as long as I don't put a load on it....so it's boost-related...if that helps the diagnosis Drs. out there.