Okay, bear with me here. Do you see the resistor in the a/c relay in parallel with the coil?

When in the run position, 12v is applied to pin 86. Current flows through the coil of the starter motor, to pin 85. The resistor is a much higher resistance than the starter motor, so most of the voltage is dropped across the a/c relay energizing the relay. When turned to the start position, 12v is applied to the starter motor (it turns and starts the car) and is also now felt at pin 85, deenergizing the relay (12v on both pins 85 and 86). This causes the a/c to be deenergized during cranking lightening the load on the engine.

Thank you mreid. It makes sense to me now how it should theoretically work. I'm scratching my head as to why the starter engages with the ignition in the ON position, unless relay terminals 85 and 86 are fused closed without any resistance. I presume that relay SR9826 operates in the standard Bosch fashion so I'll try one and see where it takes me.
Johan

The wrong a/c relay (one without the resistor) will cause exactly that problem.

Can't find anything on the SR9826 specks. Any idea what the resistance should be between 85 and 86?

Pull the relay and see if the problem goes away. You don't need it unless you run a/c and the car will run without it for this test. Trying to read resistance is tough as the relay coil will cause a very low resistance reading.

The starter behaves when I pull the relay. The relay measures 40.8 ohms vs the standard Bosch variety measuring 36 or so. Not much difference I think.
I crossed terminal 30 to 87 and both the starter and a/c fan behaves. I'll try grounding 86 next.
I'm toying with the idea to use the tec3r engine management system to control the a/ c fan. I already use it to disengage the a/c compressor on heavy descelleration to avoid stalling.
Johan

Get a new a/c relay and your problem should be fixed.

Thanks mreid. Sometimes one needs someone else to point out the obvious.http://forums.pelicanparts.com/support/smileys/wat5.gif
Cheers,
Johan