Forester XT Turbo question (USA)

I believe it's always spinning to some degree from exhaust pressure. But the turbo could be working at reasonable highway speeds too.

More or less.

Not exactly. The compression ratio of the cylinders has to be reduced to make up for the increased pressure when the turbo adds boost. The turbo itself puts a bit of restriction on the exhaust even there's no boost.

This is a pretty easy to read primer:

Thank you very much for the info and links.

The computer controls it based on the demands you put on the engine. There is a wastegate that bleeds off pressure based on the computer control. During normal driving the turbo is always spinning to some degree. A "turbo" supercharger depends on exhaust gas pressure and velocity so more "turbo power" is available at higher RPM. The computer/wastegate will allow a maximum pressure at high rpms so as not to overpressurize the engine.

An after market boost gauge would tell you the whole story. The turbo does spin constantly. However, as RPM increases so does boost. At a pre-determined point, you go from vacuum to positive boost. This is the point at which the engine is working with mechanical aspiration as opposed to normal aspiration. Unless you put your foot in it, the Subie will drive just like a non-turbo car and the mileage will agree as well. Put your foot in it and get the boost going and the power pours on...as does the gas with a resulting decrease in mileage.

But it's not purely a function of RPM, although higher revs usually results in higher exhaust pressure.

I believe the potential problem is more than just overpressurizing the engine, but that the turbo can only spin so fast before its bearings self destruct.

Yes, some early high boost turbos had problems caused by the limitations of the lubricant which would "cook" the turbo bearings. Remember when it was recommended by car writers that before turning off a turbo engine, one should run the engine at idle for 30 seconds, so as to permit a cool down with fresh oil spash? Intercoolers and improved sensors and computer control have made that recommendation unnecessary.

From what I have observed, the amount of boost and the point at which boost begins are both computer controlled. On my way back from Arizona last Wednesday I noticed that at 80 mph I was pulling 15 lbs of vacuum. If I pressed the accelerator more than a half inch, I got 10 lbs positive boost. Three quarters of the way to the floor and I got 18 lbs boost. Flooring it permitted up to 20 lbs boost. I could cruise at 85-90 and still be pulling vacuum as long as the throttle position was less than 1/2" depressed.

Prior to the dyno-tune, the car would boost up to 22 lbs if floored. It did this even though it would begin to ping at 19 lbs. Watching it on the dyno, it was clear that the boost would continue and the computer would retard the spark at the first sign of detonation. With the better gas and a fresh dyno run, the boost was limited to 20 lbs max and the spark was allowed to advance considerably farther. With a complete cat-back system, 3 " piping, an injun intake and a front mounted intercooler and an after-market blow-off valve, the car develops a regular 275HP with occasional jumps to 285 HP after soaking the intercooler with water spray. This is at the wheels HP measured on the dyno. With the bad gas, the best we got was 255 HP.

Turbo "coking" on shutdown is a totally different matter than what I was talking about. It used to be a function of reduced lubrication (creating more heat) and lack of heat transfer from the bearings when oil flow was shut off. The excessive heat produced would cause the "boundary layer" of oil left on the bearings to cook - also called "coking". Allowing the engine to idle before shutdown would allow the turbo to slow down to the point where there wouldn't be as much heat generated after the oil flow stopped.

The real reason why turbo cooldown isn't as critical these days is because they're water cooled with a mechanism that allows the coolant to circulate after the engine is shut off. As long as the heat is drawn away by the coolant, the little bit of oil left in the bearings after shutdown shouldn't coke in most conditions.

I use Mobil 1 5W-30 in my WRX, just to be on the safe side. Even with the Mobil 1 and water cooling, I still let it spin for a few seconds before shutting it down.

I was talking about excessive pressure spinning the turbo beyond its ability to survive even with oil flow. It's got its limits before the bearings get fried. I understand that turbos can coke at excessive revs even with oil flow, depending on various factors (like oil type).