I'm in the process of wiring up my Advance AoA (Sport) at the moment and noticed that there is only a single ground [Com] wire to the CPU. It is up to the builder to provide a return path for the flap switch and the two push buttons. I ran a single ground wire to both push buttons, so I got to wondering if a loose connection on that ground could be responsible for both the uncommanded dimming of the display and initiation of a test cycle. The manual states, "you cannot dim the display during the self test..." so that would seem to rule out my loose ground theory. Today, I checked the two push button switches and they're both Normally Open switches, so a faulty ground couldn't be responsible for what Angier saw.