Boost creep on WG spring pressure STI

Hey guys,

I'm currently tuning a 2002 Subaru STI on open source and noticed something a bit odd. With the engine on only waste gate spring pressure, the boost comes on to around 6psi at 3400rpm and then gradually increases to 10psi at redline. This was on the dyno...out on the road in real world conditions its even worse 10-13psi.

Car is on the stock turbo with the only mod being a 3 port boost solenoid

Any ideas?

Your wastegate actuator has a boost leak. Repair or replace.

I recently ran into a similar problem on a 944 Turbo with a standalone ECU. At the rated spring pressure, the wastegate would not fully open due to the leak. This allowed the boost pressure to rise, the wastegate would not fully open until more than our overboost limit. I thought it was my boost control programming, but by closing that off, we determined it wouldn't limit the boost even with just spring pressure.

Dave, could it be the preload on the actuator ?

I would think the preload would only control the initial pressure required to start opening the wastegate. But, if you change it, does it make the boost creep better or worse?

It may be the preload contributing - it won't just affect the force (pressure) required to lift the valve off it's seat, but continue to apply that preload all the way through.

ie. if the spring rate is 100 lbs/inch, and the preload is 10 lbs (1/10th inch), it will require 30 lbs force (initial 10 plus 20) to open it 2/10ths of an inch. With 20 lbs seat force it will require 40 lbs total force (20 plus 20) to open it the same 2/10ths of an inch. Of course, because of the bleed off in this application, the actual result will be somewhere in between for both opening and force as they find a balance.

It is normal for any issues like this to show up more on the road, rather than the dyno (especially inertia types) because of the way the vehicle's engine is being loaded.

In your position, I'd follow David's suggestion of adjusting the pre-load first, then following up with checking for leaks as he also suggests.

Thanks for the response guys. How do I go about checking for a boost leak on the actuator?

If necessary, remove the actuator/wastegate so you can actually see it move. Then apply a known pressure (use a regulator connected to an air compressor) to apply a known pressure to bottom of the wastegate. Adjust the pressure and observe the wastegate movement. Observe what pressure it takes to start opening, and when it fully opens.

Listen for sounds of the compressed air escaping -- if it comes out the top of the actuator, you make have a leaking seal.

Dave, in my case its an OEM internal waste gate with a single port. I'm assuming in my instance, I would simply be connecting the compressed air feed to the port?

Is there any procedure for testing the wastegate in the Factory Service Manual? Since I.ve never looked at or tried to understand that design, I’m afraid I can’t offer any useful advice.

When checking a wastegate for leaks , simply hook a vacuum gun that can do both vacuum and pressure and test out if it holds pressure, if it doesn’t you probably have ruptured diaphragm, which is quite common with internal gates .