Sometimes only getting 5 lb of Boost

[Primeza]

I have a stock 92 TW and just installed a boost gauge. I find that sometimes I am getting 9-10 lb of boost, and sometimes I am getting only 5 lb of boost. This is pretty common and happens more than 25% of the time and pretty much every time I drive. The wastegate spring of the VF11 is 5lb, so I'm thinking that either there is a problem with the factory boost control solenoid or perhaps the knock sensor or something else telling the ECU to limit the boost. Does anyone have suggestions on what I can do to see why I am not always getting 9-10 lb of boost?

I have not seen the CEL go on. I am thinking of making a cable for the Select Monitor to see what I can find out from the ECU. I think the next thing I will try is to see if there are any stored codes, reset the ECU and see if I get any new ones.

[ericem]

Are you running premium 91 octane? Maybe try a better fuel station for starters, and look at your knock sensor and see if it is cracked. Just get a flash light and look on the side of the manifold on the driver side of it. You should see the plastic sensor below with a single wire coming from it, and should be very obvious if it is cracked.

[Primeza]

I do run premium 91 octane from the same station I fill my 04 WRX. Let me check the knock sensor again, I didn't see any obvious cracks. I did have a problem with the wiring of the knock sensor and had to replace the connector for the knock sensor. Let me check that the sensor is properly grounded. When the sensor was disconnected it did obviously trigger a CEL. I wouldn't mind replacing the knock sensor if I knew it was bad. If this was happening on my WRX, I would connect the Cobb ODB cable and check the values for Knock Correction, IAM and wastegate duty cycle.

[ericem]

Most likely it is the knock sensor. Funny thing is my knock sensor on my 94 SS did not have a obvious crack at all, was VERYY difficult to see even when removed lol, yet tripped a code and little to no boost. On my 93 wagon n/a was split with a good 2mm gap and did not trip a code. Made a huge difference after replacing the knock sensor though, always makes 8-9 psi pretty instantly.

[Bdub]

It may be your boost control solenoid. That seems to be a very common problem as the car ages.
If you were to connect the wastegate and the turbo inlet nipples together, with a vac. hose, overriding the solenoid, you may notice a change. I don't recommend doing this as a fix, but only to see how the car will react with the solenoid bypassed.
Don't unplug the harness on the solenoid though, but you can leave the vac. hose nipples on it alone, no need to cap them off.

[Imprezive]

I've had this same problem. The vf11 as a waste gate with an spring resistance that will allow for approximately 5psi of boost. Your waste gate solenoid opens and closes to restrict air pressure from opening the wastegate until your MAP sees a certain pressure at which point it will progressively open the wastegate to maintain a solid 9-ish psi.

your waste gate solenoid is almost certainly getting stuck open, at that point the waste gate solenoid does nothing and the system is now basically a loop from the compressor housing to your wastegate actuator. Or as if you have simply bypassed the waste gate solenoid.

Before assuming its your knock sensor or anything like that I would think its a faulty wastegate.

[Primeza]

Thank you all for your advice. It's about $100 for a new knock sensor and don't want to replace if I don't have to. I think this is what I'm going to do:

As soon as I have time, I'm going to read any stored codes. Then reset the ECU and read the codes again later.

Next I'm going to make the cable to get Vikash's Select Monitor scan tool to run. Hopefully there is a parameter that will tell me the duty cycle of the boost control solenoid or if the ECU is detecting knock.

If I do find that I have a faulty boost control solenoid, then I will replace with a manual boost controler.

[Primeza]

I connected a voltmeter to the boost control solenoid. It looks like the BCS is fine and the ECU is forcing the car to run on wastegate boost. I'm still trying to get the select monitor cable to work. I pulled the knock sensor and did not find any obvious cracks. I would think that the problem is either the knock sensor or the manifold pressure sensor or the engine is really knocking. I wouldn't think that it is knock because of the altitude and I am using 91 octane.

When I only see 5lb boost I see the voltage on the BCS is always ~0.18v. When the ignition is turned off the voltage is zero.

When boost control is working, I see ~10lb boost. Under vacuum, the voltage is about 13v and decreases as boost is applied and decreases to 5 to 6 volts. Now, when stopped under full vacuum the voltage goes to 13v then drops to ~0.18v.

I initially thought that the voltage at 0.18v could be caused by a short in the BCS. But then I noticed that the voltage drops to 0.18v when operating normally. So it seems to me that the BCS is working fine and the ECU has disabled boost control and is forcing the turbo to run on just wastegate boost.

What I can't wrap my head around is that the voltage to the BCS goes from 0.18 under full vacuum to 13 then tapers down to 5 to 6 volts under full boost. Can someone explain how or why the solenoid operates like this? The solenoid is closed when the voltage is 0 or 13v and opens when the voltage decreases. I would think that increasing voltage would cause the solenoid to open. When the voltage goes from 0.18v to 13v would this cause the solenoid to momentarily open?

[BXSS]

I had the same problem - knock sensor was the cause.

Not sure if there is a u-pull-it in your area but a knock sensor from any Subaru will work, you just have to re-pin the connector on the updated sensor off of later model cars - total u-pull-it cost should be $5.

[Primeza]

I measured the resistance of the knock sensor and mine was 556k ohm. I disconnected the knock sensor and connected the resistor instead and grounded the otherside. After a quick test drive, I now always get >5lb boost. Of course I did not drive around a lot with the knock sensor bypassed and avoided full boost. I have a new knock sensor on order. I guess I could have saved a lot of headache by replacing the knock sensor at the start, but I wanted to make sure this was the problem before ordering a new sensor. Thanks everyone for the advice.

[Primeza]

Well, I did replace the knock sensor with a brand new OEM one and still had the problem. At that point I thought I was having knock, possibly due to carbon build up or something else. At the same time I installed a new Bosch O2 sensor and later I installed new NGK BKR6E11 plugs (OEM and turns out the same as what it was running). I was still having the problem.

Later I noticed that the harness wires from the knock, crank position and cam angle sensors were not protected by wire looming and the insulation was cracking on the wires. I didn't see that the wires were shorting, but as a preventative measure, I wrapped the wires in electrical tape to prevent shorting and bundled the 3 wires in looming to protect them from the heat of the turbo/engine. After I did this (and I replaced the speedo cable that was causing a diff fluid leak) I was getting a solid 9 LB of boost. So it seems to me that the problem was a wiring issue.

Wierd that this would be the fix because there was no obvious shorting and I would think that replacing the knock sensor with a resistor I would have still had the problem. Perhaps the exposed wires were shorting from engine oil/fluids or perhaps the knock sensor is more sensitive to wiring problems than a resistor. Anyhow, I resolved my problem and though I would put it out there in case anyone else is having the same problem.

[ericem]

Alot of my wiring harness is exposed as well and the car doesn't seem to be performing very consistent. I think I better take the same measure. Ill get the wire loom but I might not tape it.