AH-HA...I think (boost controller issues): RESOLVED

[georryan]

So for a while i've been wondering at the high boost rates of my car. I've checked for vacuum leaks all over, and even replaced all the turbo to controller hoses and a couple others.

Hardy let me borrow a boost controller setup that he just got from Ciper to throw in my car to test things out. Before I did that, though, I ran the vrg3 scan tool on the way home from work and monitored all the settings. Amazingly, I got no timing corrrection at all. I NEVER saw it change from 0 degrees. Back to the topic, though, I never saw the controller duty cycle change. It stayed solid at 1.56%. I swapped the controller today before I left for work, and when I drove to work the only difference I saw was that when I left home I had a system voltage of 14 vs 13.76. The boost controller duty cycle was the same: Fixed at 1.56%.

I convinced my brother to drive around a little tonight, and I instantly noticed that his boost controller duty cycle was only at that value at idle. All other times it ranged from 30% to 80%.

So basically my controller is doing nothing?? I sappose it is possable to get two bad controllers, but if that isn't the case, then what else could cause this??

Bad Ecu? Could a hose that is oversized cause this? The hose that goes from the controller to the resonator was replaced because of a broken connection, and it was replaced with a hose that looks the same size, but its hard to tell.

I want to put my controller in my brother's car and see if it does anything. (my car is a 91, and his is a 92)

Any input? (sorry if this doesn't fit the turbo forum, wasn't sure if it was engine or turbo, but seems to fit turbo better.)

-Ryan

[BAC5.2]

Just run a manual boost controller, and forget about the problems?

[georryan]

If the problem doesn't reside in the controller, it won't help with a manual one.

[BAC5.2]

Why not?

The only thing controlling boost is the boost control solenoid. If it does not work, then bypass it with an MBC. If the ECU control for the Boost control solenoid, then the MBC will bypass that also (since YOU control the MBC, not the ECU).

[georryan]

what about the wastegate??

[georryan]

You are probably right, though, Phil. I just want to make sure that I test this thing, and know what else could cause the problem, before I do much else.

-Ryan

[vrg3]

1.56% is the minimum it'll ever show. That really means the solenoid isn't being driven at all. So, on your car, the ECU is sometimes deciding to quit boosting above wastegate pressure.

I don't fully understand exactly what triggers that behavior. It does seem to happen when the ECU fails to accurately control boost -- it just gives up. I've seen this happen when it tries to control boost through an intercooler, for example. It does also seem associated with excessive detonation. You'd expect to see some timing retard even at low boost though if that were the case.

It might be that hose you're talking about. When you replaced the hose, did you keep the restrictor that goes inside it? Without that restrictor the ECU's boost control maps wouldn't work (particularly at higher RPMs) and so it's easy to imagine that it would give up on boost control. And then after you drive around gently for a while it would try again and find that it works again, and then start controlling boost again.

The hose from the boost control solenoid to the resonator is supposed to be a 5.5mm ID with a 2mm restrictor in it. The OEM hose is part number 807505442, and the OEM restrictor is part number 22326AA090.

[georryan]

What?? I thought the restirictor was that reducer piece in the hose that goes from the controller to the turbo, or wastegate. I can't remember which one it is at the top of my head.

It isn't that my car quits boosting, cause it boosts no problem. It also isn't that my controller stops working, it doesn't ever work.

So that restrictor is on the hose that comes directly down out of the controller? I could be missing that.

[BAC5.2]

That's what controls boost. Pressure bleeds to the wastegate can through the boost control solenoid, allowing it to open and exhaust gas bypass the turbine.

An MBC is a restriction. Not allowing pressure to pass to the wastegate can until it overcomes the resistance of the spring in the MBC. It bypasses the boost control solenoid.

[vrg3]

Yes, the restrictor is in the hose going from the bottom of the solenoid to the side of the resonator:

http://www.graphics.cornell.edu/~v/vacuum/

I'm not saying it quits boosting. The ECU just quits trying to raise boost pressure above wastegate level.

It's not quite correct to call the solenoid the "boost controller." It's just a solenoid valve. The solenoid, the pressure sensor, and the ECU all together form the boost controller.

I thought you said when your brother drove it around you saw non-1.56% duty cycles... Wasn't boost rising above wastegate level then? Or does your brother have his own Legacy Turbo?

If you're missing the restrictor in that hose it seems possible that that would make the ECU give up on boost control, and that replacing it and resetting the ECU could resolve the problem.

Actually, just resetting the ECU ought to resolve the problem for at least a short time, like maybe one full-throttle acceleration run.

[georryan]

No, when I saw greater boost cycles was in my brother's car, not my car.

Yes, my brother has his own turbo legacy. That's the one we did all the tranny work in and changed an auto to a manual with 4.111 gears.

Thanks for clearing this up guys, I think I'm finally understanding this whole thing better. Is that restrictor easy to find? I think I'm going to go look for it in Hardy's stock hose. Is it possable to pull out, or will I just have to order it and stick it in my current hose?

-Ryan

[georryan]

My current setup takes hose 'J' and t's it into hose 'B' or 'Q' (I'll have to check which one), and the resonator hole is plugged up. I'm betting 'J' doesn't have the restrictor.

[vrg3]

Ah, okay. Well, I think you might want to try just resetting your ECU to see if it starts driving the solenoid a little. If it works, and then stops after a few cycles of cruising-and-boosting, that's a good indication that failure to control boost is the problem.

I would expect any dealer to be able to order you the restrictor for 4 or 5 bucks. You can push it out of a hose if you force a rod through the hose. A 3/16" brass rod worked for me. And you'll want the rod to use when installing the restrictor too.

Or you could just get some 1/4" brass or aluminum rod and make your own restrictor. Just cut about a centimeter's length of rod and drill a 2mm hole down the center. Then shove it into the hose a some smaller rod.

Okay... If it's teed into B, then that would be terribly terribly wrong.

Teeing it into Q might work okay, but I'd feel funny about venting pressurized air into the hose that's meant to carry fumes from the charcoal canister. I mean, sure, the partial vacuum in the resonator will probably mean it'll all go into the compressor, but just the possibility of pressurizing the canister sounds unpleasant.

But why not just use the resonator hole? If the nipple broke off it's a simple matter to drill it out to 11/32", cut a few 1/8" pipe threads in, and force a nylon 1/8" MIP to 1/4" barb fitting hard enough that it seals.

[georryan]

It's not t'd into B. I was talking to someone else when I typed that and said the wrong hose. It's actually t'd into Q.

I took the restrictor out of the other hose i have and put it in my current hose. I'm not sure if the ECU got reset or not. I had the batter disconnected all night. The boost control solenoid still read 1.56%. The fact that there is a signal there says that there is some level of communication between the two right?

I may have to go make myself a MBC and see what difference it makes.

-Ryan

[vrg3]

There aren't "levels of communication" between the ECU and the boost control solenoid. The solenoid always does what the ECU tells it to do, which is either to open or not to open.

The scan tool reporting a duty cycle of 1.56% means that the ECU is still not telling the solenoid to open. I don't think it got reset. A freshly reset ECU should drive the solenoid for at least a little while.

Try disconnecting the battery and then pressing the brake pedal for 30 seconds to discharge any capacitors that might be charged. Then start it up with the scan tool connected and just blip the throttle. You should see the duty cycle change when the RPMs rise.

An MBC would certainly control your boost level. And it would either mask or render moot this problem. But it might be good to try and figure out why the ECU thinks your engine shouldn't be under more than 6 psi of boost.

[Brat4by4]

I assume your check engine light isn't on? Problems that trigger that light also initiate wastegate boost.

[georryan]

Well, my purge solenoid has flipped the check engine light. That is the only code, though.

I agree vrg3, I do want to see what is up between my ecu and my boost controller solenoid. That's why I've put off getting a MBC. I was wondering if maybe the wiring could be bad for whatever reason, but I wouldn't be seeing seeing that percentage if for any reason the ecu couldn't see the solenoid, right? That's all I was getting at.

I'll try resetting the battery again tonight. I pressed the brake pedal several times but didn't hold it.

-Ryan

[vrg3]

Right, if the ECU didn't detect the solenoid connected to it, it would throw a CEL about it and actually cut fuel any time you started to go on boost. So your ECU does actually see the solenoid.

[Brat4by4]

Check Engine light on = wastegate boost.

You gotta fix that before anything else happens.
The ECU thinks something is wrong and the engineers at Subaru told it to not let it boost if something might be wrong.

[georryan]

Well boosting isn't the problem. I'm boosting just fine. I'm getting 10-13 psi consistently. I can't get above 10 in second, but with a load in third or fourth I will get fuel cut if I push it.

The ecu knows how much boost I'm getting. The solenoid isn't doing anything to stop it.

I'm assuming by wastegate boost you're referring to the 6psi cap that the ecu would normally stop me at if the solenoid wasn't working correctly, right?

[vrg3]

Woah, wait -- you're getting 10 to 13 psi with the ECU reporting a constant boost control duty cycle of 1.56%? That's not right.

Do you have a leak in one of your boost control hoses or something?

Or maybe a faulty wastegate actuator? What boost level do you get if you just connect the wastegate actuator directly to the compressor outlet nipple with a short piece of 1/4" hose?

[georryan]

I'll check on connecting the wastegate to the compressor, but yeah, this isn't right. I thought I'd made that clear earlier, maybe I didn't explain it right it right.

So..quick recap:

My boost is NOT limited in any way. My controller solenoid reads 1.56% activity 100% of the time.

I've replaced all the wastegate/compressor hoses that go to the controller solenoid (3 of them, and two are joined with the plastic union piece which I thought was the restrictor everyone talks about). My atmospheric numbers are correct as well.

The Line J was spliced into line Q a while back but didn't have the restrictor in there. I have put it back in since.

The ecu is seeing accurate boost readings as well.

How easy should it be for me to force open my wastegate by pusing on that rod above the turbo, cause I can do it without tooo much effort.

Edit: Also, I'm seeing boost ranges of 5-6 in first, 9-10 in second, and if I have a load on the engine, I'll see it raise from 11-12+ in third and fourth, although I never leave it there long enough to see how far it will go. I'm too scared of detination or hurting my engine to push it like that. I figure with my exhaust 10 isnt so bad, but I don't want to do more than that, so I monitor it pretty carefully.

[vrg3]

I guess I just didn't understand what "high boost rates" meant and didn't bother to ask. Anyway, I see what you're saying now.

It shouldn't be too hard to open the wastegate by hand. I'd say it should take maybe 15 or 20 pounds of force. You could calculate how much force it should take by figuring out the area of the diaphragm in the wastegate actuator and multiplying by 6 psi.

Actually, that's an important thing to check... you want to make sure the wastegate can open freely.

Basically, my current assessment of the situation is this:

Something is causing your boost levels to go way too high. Obviously in such a state the ECU wouldn't see what it expects. So the ECU stops cycling the boost control solenoid. In such a state you should be seeing about 6 psi of boost. Since you're not, obviously something is wrong (see the beginning of this paragraph).

The constant 1.56% is a symptom.

But you're not seeing unlimited boost. Without any boost control at all, you'd rocket up to 20 psi in the blink of a second even in first gear.

Potential causes that I can think of:

- A leak in one of the boost control hoses. You replaced the hoses, though, so this seems kind of unlikely... are they clamped down and everything? It's possible to have hoses or joints that seal normally but leak under pressure.

- A leak in the wastegate actuator housing. If you have a source of low-pressure compressed air, you might try applying it to the nipple on the actuator and see if you can hear air hissing out or something.

- Some kind of damage to the wastegate actuator that increased preload on the spring. I suppose if the actuator were crushed along its axis this could happen. Seems doubtful though.

- An internal leak in the boost control solenoid. This also seems doubtful, because the solenoid was designed in such a way that it should be almost impossible for a leak like this to happen. And, of course, you've swapped out the solenoid with no effect.

- Some kind of blockage of the exhaust gas coming out of the wastegate. You have a CES exhaust, right? Is it possible you have the wrong BEP and it's interfering with the wastegate actuator or anything like that?

- Boost creep due to insufficient wastegate flow. You have the stock turbo, right? I'd have a hard time believing that the stock turbo's compressor could ever make too much flow for the stock turbo's wastegate.

- Umm, that's all I can think of right now.

[dzx]

Maybe the wastegate door is messed up. On my stock turbo, the wastegate wouldnt open until about 11psi. When i took it off, i noticed i could move the arm a bit before it opened the wastegate.

[georryan]

Its all good. I should have been more clear. When I connect the two turbo lines together (wastegate/compressor) I'll let you know what happens.

I don't think there is any obstruction with the BEP cause I was pretty careful about checking that out. I'll look into that as well.

-Ryan