Hey John,
Quick question: Were you able to regain mount communication before power-cycling the mount when you found it in the inverted position? If so, where did the mount "think" it was pointed at that time?
In my mind, the mount would only go past the limit if it didn't think it was at the limit. If there was a power surge/interruption/glitch of some sort and the mount experienced a cold-start when the mount was close to the meridian, it might power back on and think it was CWD even though it was at/near the safety limit. Then it would continue to track until it was counterweight up... thinking that new position was the safety limit given it's relative position (~90 degrees) from the meridian.
This could possibly be caused by a failing power supply, poor connections that contract as the equipment cools, failing battery, or possibly corrupted files caused by long term low-level power issues. Perhaps your mount started a flip, but the AMP draw overloaded a connection, and your Gemini 2 cold-started... woke up... and continued to track from what it "thought" was CWD.
?
If the mount does not reconnect, you could at least look at the logs and see when the last connection was known... If you know what time that was and what target you were looking at you should be able to figure out what look angle the mount had at the connection loss. If the angle between that position and the found position is equal to CWD and your Safety Limit (or less) that may provide a clue.
I recently had a handful of evenings where the mount wouldn't flip and would instead just run up against the limit. I cleaned my power contacts, replaced my power cables and power supply and the problem has (for now) not reappeared.
Ryan Noonan