I get a constant CEL that flashes for 10-20secs after giving it a little stab of gas once in a while. Under normal driving conditions, it's not flashing. To give you an example, if I give it a bit of welly to get sideways in the snow, the CEL goes from constant to flashing right after for 10-20secs and then returns to constant/solid lit.
I'm a bit confused by this as I know that P0304 is number four cylinder and I have all new NGK iridium plugs w .030 gap. I pulled the plugs for cyls 2 and 4 and switched them to no avail. So then I got a new coil pack and tried that in number four w no success either.

This CEL first appeared at the end of tuning last week and has been w me ever since.

Here's a quick breakdown of my motor and related parts..

Ecutek tuned by Four Star
ej257 w ported wrx heads
ported intake mani
stock intake
ARP studs
Four Star giga motor mounts
Group N tranny mount and pitch rod
Spearco TMIC
STi red Y-pipe
Exedy stg 1 clutch
JDM STi Spec C flywheel-maybe weighs 17LBs
New STi pink injectors
New VF-39
Stock (ported a bit) headers
HKS Up-Pipe
Bosal divorced Down-Pipe
stock third cat
Perrin Dual Tube catback

Any of you fellas got any suggestions as to what I should check next?

Thanks in advance,

Simon

Just to add something. Not sure if it matters or not. But on Sunday, I was driving back from out of town and I played tag w an S6 and an RX8 for 100kilometers and I did a bunch of 120 to 150 klm squirts w zero flashing CEL. This seems to only happen at low speeds. Like when I spin the wheels in the snow..

sound like u are running too much timing in there is my guess u have to run very very low timing in the hybrid motor

sound like u are running too much timing in there is my guess u have to run very very low timing in the hybrid motor

why is that??

I'll have to take a trip back to Four Star. Thanks for the quick response.

I was planning the same upgrade but think I am going to go with the Crawford S2 block. Can make wayy more power. What kind of numbers did you put down on 4-star's dyno?

Also, flashing CEL usually means missfire, no? I blew my engine the other day and had no cel. I think it was partially due to det too.

The Dyno Dynamics that Four Star uses only spewed out about 240/270 wheel. Frank said he can pull more power out of the motor, but we ran out of time that day. I asked him how much and he said like 10-15 wheel more.

The reason I went this route was partly time constraints and partly money constaints. Going w a Crawford or Axis or whatever, I would no doubt have gone w pistons, rods, cams, bigger turbo, injectors, etc. Too much money for me to rationalize to my Wife, and I think this would have taken considerably longer to put together.

Yeah, I have the stock STI block in hand right now but I think I'm gonna go Crawford anyway. Pat Richard tells me its the way to go and he has never been wrong in anything that he has told me. I have not heeded his advice in the past, much to my own demise.

Gotta listen to that man. He knows whats up. I see you listing your sti sb on TSc. Good luck w the sale.

Thanks. I think it will sell. There are lots of 06/07 WRX/FXT owners looking for these. Not to mention WRX owners.

Ok, so I still have the flashing CEL but it's only at idle. It's been doing this for a couple weeks. Went back to Four Star last week for fine tuning and the car is as dialed as its going to get. There's tons of work into this map and its as safe as it could be. Boost is at 18Lbs max.
But under normal cruising type driving conditions theres a tiny bit of hesitation at 2200, 2400, 2600, and 2800 rpms and that CEL is constant except at idle. Every single time it idles, it flashes. As soon as I pull away or rev the motor up a bit, it goes solid.
Nobody can seem to tell me why this would be happening.

(Frank at Four Star moved my stereo power cables away from the block and that made no difference. I have a PLX Devices wideband and that was causing timing issues being wired into the ECU but we disconnected that from the ECU. It still has power and the wires are still probably laying next to the ECU until I rip it out entirely tomorrow. It couldn't be anything to do with that now, could it?))

Can anyone shed some light?

Is the p0304 the only code that you are getting?

Hey man, a flashing cel is usually a bad misfire. You are already pulling misfire codes.

I am positive its your timing belt and that it is on wrong. I had this problem with my hybrid. It is probably off by a tooth. My symptoms were exactly like yours. and especially since its adjacent cylinders that is your most likely cause. Timeing is fairly easy to do, if you have some mechanical no how u can easily do it yourself.

Yeah. When Frank looked at the codes in memory, there was 0302 as well and two that were "unknown" or something like that.

I think because 0302 came up as well, he maybe thought it had something to do w my stereo power cables that run along the driver side of the motor to the battery. But he said he moved them around and that didn't work.

The funny thing is I ran the motor for 2,000 klms without any CEL and then right at the end of tuning the light popped on.

The funny thing is I ran the motor for 2,000 klms without any CEL and then right at the end of tuning the light popped on.

I could have always skipped a tooth. Very likely if you reused the tensioner and timing belt. It depends how many miles u had on ur original tensioner, they do go bad if you compress them incorrectly. but i doubt it.

But seriously check out that timing belt it will take approx 2 hours to do. most of that is to get to the timing belt. the crank pulley is a bitch i will say that. Its expensive if you go have a shop do it. You will need to take out the radiator.

If you need me to send you instructions from the subaru manual let me know.


GROUP OF #2 AND #4 CYLINDERS
Are there faults in #2 and #4
cylinders?

Yes
Repair or replace faulty parts.
NOTE: Check the following items.
-Spark plugs,
-Fuel injectors,
-Compression ratio,
-Skipping timing belt teeth

No

Go to DTC PO171
and P0172. <Ref.to EN(D0HC TURBO)-1 52, DTC PO171 -FUEL TRIM MALFUNCTION (NF Diagnostic Procedure with Diagnostic Trouble Code (DTC).>

I agree with NITROS. Check that belt. Did you reuse your tensioner? Did you reuse the belts? The tensioner has to be compressed REALLY slowly and in a vertical direction.

Thanks a lot for the help. I think it was a new belt. But maybe an old tensioner.
I'll get Four Star to do it since they warranty the motor. :)

God, I hope thats it..

Also look into your injectors and see if they could be the issue. Did you use new or used ones?

Also, what head gasket did you use?

New injectors, new plugs, old coil packs. New timing belt...I think. Probably old tensioner.
Not sure about the gasket. I'd have to check my invoice at home. Nothing unusual like Cometic or anything..
I sent Four Star an email to see if they'll look at the timing belt for me. I'll get an answer tomorrow and then its just going there. So I should know either way in a few days..

Yeah, I have the stock STI block in hand right now but I think I'm gonna go Crawford anyway. Pat Richard tells me its the way to go and he has never been wrong in anything that he has told me. I have not heeded his advice in the past, much to my own demise.

You will like the S2. I was running the S2 on Wrx heads, but due to a timing belt problem, I had to upgrade the heads so I went with the JDM Large Port. I am having Bill from I-Speed tune it over the internet through Ecutek Easy ECU, which you might want to look in to as well.

What timing belt problem would make you have to upgrade the heads?

The Bolt that holds the belt tensioner to the block sheared off, which allowed the timing to jump. Ultimately allowing the valves to try and make their own valve reliefs in the top of my CP pistons. So the valves got bent. The block was fine though. I decided that I would not just rebuild the WRX heads back to stock, so I have a set of JDM Ver7 Large Ports with JDM Ver7 ECU running AVCS. I wound up getting a better deal on the Heads and ECU then doing a small build on the WRX Heads.

Wow! That's crazy! I can't believe that bolt sheared off. Did you take that up with anybody because clearly a sheared bolt is not your fault?

Who's fault would it be? It's not Crawfords as they did not supply the bolt nor did they do any of the work. It's not Subaru's fault that a bolt that had been on a car for 80K niles worth of wear and tear broke. It's not my fault that it broke, as I did not leave anything in a bind when I put the motor together. The setup had been in my car for about 4-6K miles when this happened, so it's just one of those things that happen from time to time when you build high performance cars. It comes down to the point that nobody was at fault, things just happen.

I suppose but man, but it really sucks when something so stupid costs you so much money. I'm crossing my fingers that all of my bolts stand up after 130k kms.

be careful when you gap iridium plugs. they are super sensitive and can break if you are too hard on them (ie, tapping the ground strap on the concrete to narrow gap, etc). Most iridium plugs come gapped correctly because of this.

No joy today. Went back to Four Star and they checked the timing belt and it's all perfect. They removed the crank pulley and that was perfectly aligned too.
I moved the stereo cables a bit further away from the block since I was standing there anyway. And I removed the PLX-Devices wideband and all it's wiring completely from the car as well.
The last thing I asked them to do was pull the codes and reset it. Codes were 0304 and 0302.
The only difference is now the CEL flashes at idle and disappears completely while driving.

So weird.

Do you have the TGV deletes?

What do your plugs look like when you took them out?

No deletes and the plugs are all new.

Hmmm. Has Frank looked into the timing tables for the idle? Also, look for a VAC leak on your intake manifold... you could be pulling in air somewhere, making you run too lean under vacuum and causing misfires.

Hook an aftermarket wideband up. Bad o2 sensor?

I switched from iridium to copper on my hybrid. It cured my misfires. They wear out much quicker though.

What kind of Engine Management do yo have? If you have a UTEC you need to load in the STi parameters from Turboxs's website. They have different knock settings than the 2.0 engine. Also MAKE sure your throttle position sensor is calibrated to the UTEC. Mine wasn't for awhile after I loaded in some new software and the car was running like crap. Also running without a tune you can be throwing all kinds of bad things at that poor engine it could be getting too much air or fuel or timing at different RPM's. I would take it easy untill you get it tuned.

I'm using Ecutek. Frank at Four Star Motorsports tuned it. (he also built this motor for those of you who don't want to read from the beginning.)
He's gone up and down the tables and he went so far as to send the map to Ecutek as well. Even they were impressed remarking on how much work Frank put into it. So the map is well dialled in.

Next, I may try another set of plugs just to be doubly sure.

Just a few comments.

First, cegpcola, I must say I respect you a lot for how you answered thejean in post 25.

Second, two years ago I had the exact same issue as described by LowUFO. It was in fact my timing belt and replacing the belt and the tensioner fixed the issue immediately. I spent a long time looking for other issues before I decided to take the timing belt apart. Lucky for me, it was only enough for the motor to act up without killing my valves.

I only wish that was my problem. We inspected everything yesterday. Timing belt, tensioner, and crank pulley. Everything was perfectly lined up. We even pulled the idle control valve..

Nothing.

Pretty soon, I'm just gonna get the electrical tape out for the CEL. lol

Dexterous, your belt was off by a tooth?

Could you do a compression test to see if the timing is off if so it could read more compression just a guess.

if you guys think a compression test will help, I'll suggest that next.

K, so new plugs today and no difference. So I'm going back to Four Star and suggest they look for a vac leak on the i.m. and that they do a compression test.
Anything else I should ask them to consider?

Update?

Well, I suggested to Frank that we raise the misfire detection threshold. So we raised it to four from one and that made no difference.
I suggested going further, but he says that would be just ignoring the problem so I'm dropping the car off today and leaving it w him until he figures it out.
Clearly, tweaking the map isn't helping. The map is really good.

I still get exactly the same symptoms. Slight hesitation at 2200, 2400, 2600, 2800, 3000 rpms, exhaust smell in cabin on startup if I have the heat going, and the evil CEL flashing at idle but not while the vehicle is moving.

All those symptoms started at the same time. Right at the end of a day on the dyno. I want to believe that its something simple like a vac leak/ruptured hose or IM leak. But fortunately, I got the motor, software, and tuning all from Four Star so they're standing behind their product and trying to figure it out.
I'll just have to be patient and hope it doesn't take forever to find. Because I don't want the car back til its fixed.

Oh, and I did ask about a comp test and he dismissed that. Maybe because he built the motor. But I'll suggest that again today now that he's requested that I leave the car with him. And I'll mention about how these hybrids require very low timing.
I believe this is his first hybrid build because he had to create the map from scratch.

I agree with Frank about not upping the knock threshold unless you are certain it's artificial knock. These 2.5 engines are noisier I've been told so they may require even higher knock thresholds. Delicately suggest to Frank to talk to Phil at Element or even Pat Richard... they've done a fair amount of hybrid's and know what the knock thresholds should be. Also, the 2.5 with a 9:1 CR will require a LOT of timing to be pulled.

Cool, thanks. I'll do that. I'm going out there in an hour or so. Fortunately, the weather is crap here so I don't mind being car-less for now..lol

Having a knocklite might not hurt your cause either. :)

How do you know you are getting detonation? I can't even tell if your knock sensor is tripping. You need some more logging that a OBDII code.

Your description of "hesitation" sounds like misfires, which could be lean or rich fueling.

The fact that it started after tuning would point to fuel or spark tuning......or you are really getting detonation.

A knock headset could clear up the detonation worry, but finding out what is actually going on requires more data from the sensor while the engine is running than you have described so far.

When its being tuned, Frank uses a knock headset and he takes the car out onto the road and watches the advanced learning, timing, knock, etc. on his laptop.
And he always plugs in his own Motec AF gauge.
If the hesitation is misfires wouldn't the CEL flash? It stays solid while its happening. It only becomes flashing at idle.
Frank has tweaked the hell out of the fuel and timing.
I dropped the car off and mentioned the timing has to be super low and he said that it is. I mentioned the CR should be 9:1 and he confirmed that too.
He said he's going to systematically start to take it apart to find something mechanically. Starting w the AC condensor and go from there.

We'll see what turns up.

Really eh? Geeze man, sorry to hear it has to be ripped apart. He really should look into what the knock thresholds need to be before he does that though. If these engines really are much noisier then it would make sense that under no load when the stock 2.0 ecu thinks eveything is supposed to be "quiet", it thinks it is seeing knock when in fact it is not. Just a theory and I am sure Frank knows this better than us anyway. I am at a loss though to figure out what mechanically could cause such a thing? What boost are you running because the rpms you mention correlate with the onset of boost. I heard the OEM hybrids sometimes can't run over 14 psi on pump 91 due to the high CR. It's funny because I just did the exact swap with the only difference is I used 8.2:1 CR pistons. I have no issues so far but I'm not pushing boost yet (breaking in the engine) and I'm still on the stock map. How much did you deck the heads... if it was a lot, your 9:1 CR could be 9.2:1, exacerbating knock issues. Whatever you do, get it sorted out because if you knock on those stock hypereutectic pistons, they will break. Also check the timing belt... not sure if skipping a tooth or two would cause this? One "possible" solution, which would avoid pulling the heads off and redecking causing even higher CR, is to use longer duration cams (like 264's), which will lower the dynamic compression, which should help a lot with knock at these rpms. JC

Boost is 18, maybe 19psi tops. But the longer I go between ecu resets, the more it learns to pull boost. When I drove out yesterday, I couldn't get past 14psi. So i guess it pulls timing?

Not sure how much work was done on the heads.

We already checked the timing belt and crank pulley. All good there.

I dunno. I'm not qualified to guess. I'm gonna leave it up to Frank.

(I just sent Frank an email w your ideas, though. Thanks JC.)

They know me pretty well there. I'm actually helping Dan out with one of their Calgary customers.

My 2,5 hybrid also had knock problems. One way to rule out high compression is to run a dyno pass on race fuel, or add 4 gals of toluene. If knock goes away it is probably compression.

JC, you going to lean on some deadbeat payer for Dan? lol


Johnnyblueeyes, were you getting knock low down in the rpms?

JC, you going to lean on some deadbeat payer for Dan? lol

haha... no, nothing like that. Just going to help some poor chap install some tricky headlights.

JC, you going to lean on some deadbeat payer for Dan? lol


Johnnyblueeyes, were you getting knock low down in the rpms?
Knocks were mid to high rpm's. Trouble turned out to be the timing belt.
Found that out after I cracked my stage 2 Axis SB. But The car made no more power on race fuel so they ruled out compression.

I'd say there is a mechanical problem with the timing. Of course that covers a lot of ground. Very hard to diagnose where the problem is.
It doesn't take too many detonations to destroy the SB by a 200 minor knocks. That was my lesson learned anyway.

Good luck

Thx brother. I'll update this when I hear from Frank.

have you tried throwing on another set of knock sensors ??

Not so far but I'll suggest that to be added onto the list of possibles to consider.

Good luck with the fix. I know how it feels to have hybrid problems.:(

Good luck with the fix. I know how it feels to have hybrid problems.:(

What kind of problems are you having?

Update: Picked up the car and its great. After all this grief, it turned out to be the TPS..!!!

Tps??? I'm an idjet.