We had a similar problem at the club with the draw bridge and it's protection system. The problem was solved by making the wires the same lenght. By adding the switch in the hot side, the outside rail connection is different. If the signal to the engine, and from the engine has a different time period, the signal value drops.
That explains to me why I got the results on my outside layout that others said should not matter. I rewired it to gain control of large consists. The mainline tracks outside are way longer than most inside layouts. Although the rails in G scale are much thicker, there still were issues with signal when the layout grew to the current size and multiple engines were in a consist. I almost added TIU channels until I solved the issue by completely dividing the blocks. At first, I thought it might be a DC transmission problem. Trying AC did not solve anything. Although the transmission signal thing is over my head, I still know just enough to troubleshoot when layouts go south. I had to add in mother nature's wrath and look at things that never happened indoors. I also remember studying the MTH patents in the beginning to help learn what was actually going on and how to solve issues.
So over the years I've learned to keep an open mind and eliminate all possible issues by looking at everything. Having guidance from members here like Marty F. ,Barry, G, etc. makes looking so much easier. I use their posts and apply them to my issue. I'll be jealous to see that you get direct at home service from a pro!
