legacycentral bbs

I have an EJ20R from a 1996 Legacy GTB in a race car. I have run many events and have been working the bugs out of it. I got to an event where I could really open it up, and started having some bogging issues. I had no gages in it but a tach and CEL. I did some testing tonight and here is what I found.

I added a boost gage to the intake opening on the left intake runner where the power brake booster hooked up. When I start the car, I get a steady vacuum reading around 18". If I accelerate quickly in first or second, and hold it to redline while in gear with wheel raised, I never get below a 5" vacuum reading. It never goes to boost.

When I shift to third and try to load up the engine by feathering the brakes, again, I never see less than 5" of vacuum, and never any boost. It runs smooth to 4500 RPM, then it starts cutting out and dying.

I think I have all the lines hooked up OK. I see the secondary turbo exhaust actuator open up, and even disconnected the vacuum port going to valve that opens the turbo to the intake so it would stay open. All that did was cause a code 66 but no boost.

Is it possible that the boost gage is in a bad area in the intake? I can't see how it would matter as it still ties to the intake.

When the car cuts out, I thought it might be running out of fuel, but I am not so sure right now. I do not get any check engine light or no codes are displayed.

If I am not running off of boost, this car will be scary fast when it does boost.

Any thoughts on this?

Thanks,
Steve

I had my gauge hooked up to the intake at first as many said it would be better.Same thing happened although i did occasionally read posotive numbers.Then i hooked it up to the outlet of the turbo and the numbers suddenly started making sense.

I've got my boost gauged hooked into the vacuum line which goes to the pressure sensor. Just T-ed off that line, since it's actually what the ECU "sees", regardless of what the turbo is doing.

BSOD2600 wrote:I've got my boost gauged hooked into the vacuum line which goes to the pressure sensor. Just T-ed off that line, since it's actually what the ECU "sees", regardless of what the turbo is doing.

Since I am new to Subarus, what do you typically see as far as vacuum at idle and when would you see boost?
Would you see it under no-load stopped condition, or must it be under load?

Thansk<
Steve

BSOD2600 wrote:I've got my boost gauged hooked into the vacuum line which goes to the pressure sensor. Just T-ed off that line, since it's actually what the ECU "sees", regardless of what the turbo is doing.

Same here.

I also see about 18 at idle.

Both my boost gauge and MAP sensor (which is what the ECU sees), see around 19-20 inHg while at idle. I'm at sea level too. As for boost.... well running a TD05 and revtronix chip so it gets up to the 16-17 PSI range.

You're not going to build boost unless the engine has load on it.

You need to drive it. If you're driving it normally, and not building boost, then you've got a problem somewhere.

Regarding the boost gauge, it needs to be hooked up to the manifold, not the turbo outlet.

How the hell do you guys see boost when hooked to the manifold? I would have a hard time showing 5-6 PSI when i had it that way and that was on a long uphill pull,the numbers made no sense.Hooked to the outlet i see the 17 i am supposed to,that would never have happened before.

gijonas wrote:How the hell do you guys see boost when hooked to the manifold? I would have a hard time showing 5-6 PSI when i had it that way and that was on a long uphill pull,the numbers made no sense.Hooked to the outlet i see the 17 i am supposed to,that would never have happened before.

Where exactly did you connect to the manifold? Any "normal" (i.e. no 1-way valves, etc.) vacuum source from the manifold shold read boost just fine. Reading from the turbo outlet won't tell you what's going on downstream of the TB, especially if the throttle's closed!

Here is an update.

I checked the boost gage location, and it was for the vacuum booster, but there was no valve and it was open to the intake.

After further applying thottle and brake, I was able to get it to read as much as 8 lbs boost. Guess I wasn't loading it hard enough.

That still leaves my main problem. The engine is from a late 96 JDM Legacy GTB with EJ20R and twin turbos running off the factory ECU with no mods. It has yellow topped fuel injectors.

In first or second gear, even with the brakes nearly locked and boost at 7 lbs, it powers right up to redline and I can hold it there (7500), When I go to third or fourth gear and apply load, if I increase gradually, I can get it to redline. If I apply heavy load and try to squeeze throttle, the car just sputters and dies. When it is dying, it feels like it runs out of gas. It is usually at 4000 - 4500 RPM when the car shows boost. If I leave off completely, and car drops down significantly in RPM, I can go right back up. If I try to lift slightly, or mash it to the floor, it continues to die like if I have no fuel. If I keep on it, it will drop right ot 2000 RPM and start sputtering. Sometimes, I can go almost to the floor and it will pick up for a few seconds then die again. I do not have any codes or CEL while any of this is happening.

I am going to try to test fuel pressure under load next to see what happens. If I disconnect MAF or anything to put it in open loop with CEL on, it does not change.

Any other thoughts on this?

Just checked fuel pressure.

It sits at 46 PSI with engine off.

Starting up, it runs at 38 PSI at idle.
At High RPM, it runs up as high as 50 PSI.
Under mild boost, about 46 PSI.
When it is acting up, it never drops under 44 PSI.
Those numbers are right to what the liberty manual says they should be.

So I am guessing that it is not fuel related, but something driven by the computer that does not create a Code.

Is anyone using a datalog system with this typw and model ECU?

Thanks,
Steve

The ECU should be OBD2 compliant, and you can find some some programs out there to interface with the ECU.

Legacy777 wrote:The ECU should be OBD2 compliant, and you can find some some programs out there to interface with the ECU.

Unfortunately, this is not OBD2. It is early subaru.
I think EVOScan will work with it nut am verifying now.

I had this problem or similar to this problem a few months ago, found out that my fuel pump was going bad. when i eased up on throttle i could get up there, but if i just floored it, the car would pop, sputter, and die.. i tried everything to figure out what was goin on, even had others look at it. then one night on my way home from work, pump just quit completely.

cj91legss wrote:I had this problem or similar to this problem a few months ago, found out that my fuel pump was going bad. when i eased up on throttle i could get up there, but if i just floored it, the car would pop, sputter, and die.. i tried everything to figure out what was goin on, even had others look at it. then one night on my way home from work, pump just quit completely.

Thanks for the reply.
I was thinking fuel pump, but the fact that the test showed I am not losing pressure makes me wonder if that is it or not. I am using an aftermarket fuel pump from Summit made for EFI:
http://www.summitracing.com/parts/SUM-G3138/

Since I do not have the Subaru MAF on it, though I was told that was the same unit, I wonder if it is an air volume under load issue in the MAF output curve?

I think I need to try to find a MAF from an EJ20R somewhere.

This thread seems a little out of my league, but I'm gonna ask a stupid question that you probably already checked.

Is there a fuel cutoff on that engine? I had a vacuum(or intake?) leak that was making me put out too much boost and it would die at full throttle like it had no fuel, and it was just because I was above 13lbs.

Apex3 wrote:This thread seems a little out of my league, but I'm gonna ask a stupid question that you probably already checked.

Is there a fuel cutoff on that engine? I had a vacuum(or intake?) leak that was making me put out too much boost and it would die at full throttle like it had no fuel, and it was just because I was above 13lbs.

I am not sure, but it probably does have it.
I really am not sure if I am getting anywhere near that boost, but it is possible. When it has a fuel cut, does it have to drop RPM or Boost to come back on?

Thanks,
Steve

OK,


Back at it last night. Re-checked many of the hoses, cleaned all solenoids and checked them for operation, traced all lines to components and they were where they should be. I think solenoid box is OK now.


Took off MAP sensor hose, replaced sensor with gage and only getting about 5 PSI with little load, but I can still shut it down, so now I am thinking it may not be over boost. I read on some boards, and the solenoid box schematic seems to show that map sensor is after a solenoid in this twin turbo system, and maybe on all.


Tonight I am going to try to rig a test light to the solenoid before the MAP, and pu a meter on the MAP sensor. What I am hoping it may show is sthat if it switches to MAP on secondaries, that I can see the solenoid open up and then watch the MAP readings. I have not checked MAP output up to this point, so it is possible that is bad sensor, wire, or ground.


After that I am going to add 5 gallons of 107 CAM 2 Race gas to take away any thoughts about octane causing this.


I did notice now when I snap throttle open, I get a pop in the intake manifold almost like a backfire. Does that give any new clues? I had not heard this before.

post removed,
didn't mean to hijack the mic, i mean thread