The starter motor was relatively simple - with small complications:

* it's hard to reach the starter. mounted on the engine, engine mounted in the engine bay. They should have connected it before mounting the engine

* it's especially hard to reach from above, so we better connect it while the car is on pit (I don't have my own workshop)

* some original assumptions about wires turned out to be wrong

Not too bad altogether, it went smooth.

----

Pinout

The Etzold book (page 250, and bottom-left of schematic I) lists a massive set of 3 connections:

* node31 - GND through the engine body.

* node30 - (black? red in my case) always hot (battery +). no relay, fuse or anything (this is very clear and consistent in every sources I have seen). Just a thick 10..16 mm2 short wire from battery. I found a thick red wire that is good for node30: +12V (direct VBATT+ connection) all the time

* node50 - (black/red?) +10..14V supply on this starts cranking. It mates the gears and also starts to spin the motor.

My surprises were:

* node50 (without power on node30) only drew 11mA when powered from 12.2V (with no power on node30, as I didn't want to rotate the engine with no oil). I expected 400..2000mA draw and a click sound for when it aligns the wheel (or maybe it is mated already?)

* the wire close to the above thick red, with blue-white color insulation (appr. 4mm thick with insulation) has +12V at ignition_on and cranking. to connect to node50, it should only be hot when cranking, not when just ignition_on) ... hmmm ....

Maybe we connect a new wire and install a pushbutton ?

Why ?

* It might be possible that node50 is somehow effected by factory ECM (it doesn't make sense, but...), which was not installed during the experiment. Also, the schematic is not too clear on this part. There are variations (optional parts) on the schematic depending on engine, and things that are not clear for me or conflicting with my experience.

* I must also check the groundings

----

See also

* MembersPage/MarcellGal/EngineSwap