This process has been documented in the following post, but I thought I would add my experience as a step-by-step.
Aiflow meter pot source - Bosch CIS
Anyway, this morning I took the first step to fix the sporadic idle-rough-and-die problem in my 300E - replacing the airflow position sensor in the throttle body. I was discussing the problem with my indie, Herr Fuchs, and he mentioned that he had a sensor that he had ordered and not used, so I traded him a half-case of Heineken for it. We both thought we got the better deal.
Before removing the old unit, I scribed a line on the airflow unit along the top of the sensor body, so as to be able to position the new one. The unit rotates a few degrees in its screw holes. Then I checked the output voltage (Pin 2) and found it to be on-spec at .72v.
Moved the idle valve and the fuel regulator out of the way, and removed the old sensor. Installed the new one along the scribe lines, leaving the screws just loose enough to allow the unit to be rotated. Hooked up the idle valve and fuel reguator.
Started the engine and found that it would barely idle. Measured the voltage Pin 3 to ground and found it to be 1.24V. Turned the sensor body the tiniest bit CCW. The idle surged for about three seconds and then settled down. Remeasured and found .69v, which I declared to be success.
Unhooked the fuel regulator, moved the idle valve a bit, and tightened the four screws. Re-installed the regulator and idle valve. Installed air cleaner and took a test drive.
Immediately notice a smoother idle - no more flutter. But even driving at low RPM, I can feel more power. Turn on to an arterial, and run it up through the gears and I've got Permagrin. Definitely more pep - the old guy hasn't felt like this since I've been the owner.
Of course, only time will tell if this fixes the stalling problem, but at the moment things are looking good.
I offer the following as observations:
The pins on the sensor are as follows:
Pin 1 - brown - ground (to shield of O2 sensor - why?)
Pin 2 -blue/black - output to CIS control unit
Pin 3 - blue/green - from altitude sensor, loop to control unit.
I was getting a steady 4.92v on pin 3.
Adjusting the small screw (the "trim pot") on the outside of the sensor had no effect on Pin 2's output voltage or the idle speed.
The biggest hassle in this job is having to temporarily connect the fuel regulator to allow the adjustment, then disconnect it so the lower left screw can be tightened, then re-connect it. Gas of course leaks out each time.
Thanks to Steve Bourg and the others for their contributions.