Serious Avionics Problem

As I had subtly mentioned yesterday, I discovered a weird connection between the Autopilot switch and the function of the ADAHRS. For the Dynon to be able to see the ADAHRS, the Autopilot switch has to be in the on position. This is clearly not correct. Bill H. replied on VAF that his system runs fine with only the backup battery with full GPS and ADAHRS functionality.
I tried that last night and found that even without any external power, I still have to have the Autopilot switch in on position for the ADAHRS to work. Removing the Autopilot fuse or disconnecting the harness leading to the switch module disables the ADAHRS function.
I contacted Van's builder support this morning but they were at a loss and need to research this problem before they can call me back. I can't say that their response made me feel more comfortable about this. It appeared they had never heard of something like this.

Update
No response from Van's builder support. But I found out much more about the problem after I got home. First, there should have been 5 devices being detected on the SkyView network: Two SV-32 servos, the SV display itself, the ADAHRS and the EMS. The latter did not get detected despite the little red LED blinking happily. With just the DB9 connector in the EMS it is supposed to be at least detected by the SkyView which it did not. I unplugged the ADAHRS from the system, I turned the AP switch off which made the servos disappear, but the EMS did still not get detected. So I used the provided test cable from Dynon and plugged this into the EMS and into the second SV network port. Still no detection. Then I removed the DB9 from the EFIS harness that went into the the lower SK network port and left only the Dynon SV network cable plugged into the upper port (it still connected the EMS to the SV display).
That was finally it. SkyView detected the EMS. Nothing but the EMS now, as there was nothing else connected to the bus anymore. However, this shows that with the correct wiring everything is working correctly.
Eliminating other harnesses contributing to the problem by removing them from the Control Module one by one revealed that the problem only shows when the Optional and Autopilot plugs are both in. This disables the ADAHRS and EMS detection. When just one of them is not plugged in, ADAHRS and EMS are working fine. I will stop for tonight and continue tomorrow to debug the wiring harness, starting with the Autopilot plug.