I think the servotronic can only increase or decrease the power assist within a range, not eliminate it.

Do you mean pins 1 and 2 at the SVT module connector in the glovebox?

I saw it in the linked photo, thanks. I guess testing it probably requires removal. I'll crawl under this weekend to see if I can get hands on it.

In normal operation, that makes sense to me. But what if that valve #5 is completely closed or clogged? I wonder if that valve handles just a control bypass in parallel with the rack or full flow from the pump?

Yes. You should be able to measure some sort of impedance. Then if not, make sure you have continuity on the harness at least. It's just a solenoid like for a transmission. This thing runs on a PWM signal based upon the calculated duty cycle from the SVT module. There is a functional description that explains this that I would have to find. I'd have to figure out if the solenoid is normally open or closed. If open, it would make sense that it would divert all line pressure back to the reservoir and no power assist. Not sure exactly...

So I did some more reading and found the hydraulic architecture for the servotronic PS.

The servo is normally open when no power is supplied to it. Therefore, this would be considered minimum support from the PS pump and divert the majority of assist back to the reservoir (expansion tank).

Supplying the Servotronic valve with power

The Servotronic valve is supplied with power in response to the car's road speed. The power-assisted steering characteristic is memorised in a characteristic curve. The characteristic curve specifies the power based on the car's road speed.
The characteristic curve is written using 16 reference points.

The Servotronic valve is no longer supplied with power when the engine is OFF (over the PT-CAN). Minimum steering force support is set (high road speed) when the Servotronic valve is no longer supplied with power.

Equally, the Servotronic valve is also no longer supplied with power if there is a fault-memory entry.

This leads me to believe there is a fault on the SVT and its effectively turning off the powersteering, aka, no steering force by means of all the DSC issues you mentioned on the previous post.

Where is the pump in that diagram? If steering assist is at minimum (for highway speed) with loss of the servotronic then you would think if that were the problem it wouldn't have cut the assist enough to almost miss the off-ramp?

Haha, OK I see the pump. My mistake.

From what the technical description is saying, it would seem when the servo is off that nearly all pressure is diverted back to the reservoir, thus, making it high effort at the steering wheel.

Sorry for my slow follow-up. With the holidays I haven't had time to even pull the glovebox apart for the multi-meter test just yet. I did read over on the M5board that a bad servotronic valve almost always throws a code. Otherwise, the valve seems easy enough to replace with just removal of the drivers front wheel and a couple of torx bolts to free it from the rack (assuming they don't strip). They also said 4 ohms is good but even a few ohms different is bad, FWIW. So hoping to finally test the valve resistance soon.

Since I've still got no relevant code, though, it does seem at the moment to more likely be an actual pump or rack issue in my case. Not wanting to deal with the rack itself, I can get a 2007 pump (since our X3 is 1/2008 production, but I think the pumps changed in 3/2008) and will probably see if that helps after confirming the correct valve resistance. Hopefully no pump difference for servotronic - I'm assuming just the rack with servotroinic valve would be specified differently with versus without the option. Next I just need to figure out how to get at the pump - hopefully from the top.

If the valve resistance seems okay and the pump replacement doesn't help, then I'll of course swap out the valve, code or no code, before contemplating replacement of the whole rack. R&R of that rack does not look like any fun! More soon...