912 ULS with 460 hours built in 2008 1000 RPM drop during Mag test

Hi Jeff.

Thanks for that.  I see that I did not completely follow the mag switch setup guidelines.  Today I will pull the two mag pins and replace them with a shielded length of wire for testing.  Maybe 10".   Ground the shielding on both ends and when the engine is running just touch one to ground at a time to the engine block.

Do you think that this would be a sufficient test?

I do think that would be a good test. And will totally eliminate or validate any problem with the ignition switch wiring. And with test wire leads that short grounding just one end of the shield would be just fine.  Careful out there around that spinning prop! If you have a helper in the aircraft they won’t be able to shut the engine down with the mag switches disconnected.  

Ok, another day another test.
Today I completely bypassed the mag switches being that I used unshielded wire in the original install.
I used two foot pieces of shielded 18ga wire.  Each end was grounded directly to the intake manifold.  This fulfills the installation spec as far as I know.  Then started it and tested it out.  I touched the exposed ends of the test mag wires to the engine.  The exact same behavior occurred.

I have nothing to test right now.  I'm hoping some Rotax 912 expert chimes in and gets me moving again.

I tested the stator power wires (red), the pickup coil harness, also purchased and install a new one,  I eliminated the mag switches.
What's next?  I firmly believe that it cannot be something from the 6-pin to the ignitor on down to the sparkplugs.  If there was any problem downstream from the 6-pin plugs then the problem should have jumped to the other mag switch when swapped the 6-pin connectors.

All this thrashing.  I just want to fly.

Michael Stated, " ...  AND a few seconds later after it was established at whatever rpm it would just suddenly start running faster and smoother like a switch was thrown.  I think that the problem is getting worse. "

Something Changed!!!  A Clue!!!
What you noticed is that the Soft-Start circuit has started working as it is supposed to.
Apparently, it was not working prior to this moment or you would Not have mentioned the change.

Your ignition drawing is from the Edition 1, Revision 3 (E1R3) Maintenance Manual.
The E1R6 manual is current and there is a minor but possibly significant difference.
You will note that on your E1R3 version, there are only 5 pins in use on the upper 6-pin connector.
On The E1R6 Version, all 6 pins are now used including a new Violet Wire for the soft-start function.

These VIOLET Wires connect to the Starter Solenoid to signal the Modules to retard the spark momentarily during the start.
During START, the Violet wires will have 12 volts on them.
After START, The Violet wires will be essentially grounded (0 volts) through the Low resistance of the Start Solenoids Coil.

If one or both of these wires has a bad connection It/they will be in an indeterminate floating state, subject to electrical noise, and all bets are Off.  Your timing could vary wildly from cycle to cycle.

Check to see if you have the extra VIOLET Wires, and if so, confirm that they are un-damaged and otherwise securely connected to the Start Solenoid.

Hmm.  I did not think that this engine had a soft start installed.  The engine is used.  I guess that I will have to check for that.  As far as I know, you can have the 6-pin plugs without having a soft start system.  
I will take a look.
Thanks