I am a T-Trak modeler from the PNW area. I just completed a set of modules and for their initial testing I noticed some odd behavior. I will explain my setup, then the behavior I have a 4 track stub ended station "yard" Entrance into the yard from east to west is as follows. Module 1 Kato crossover (1-5 for DS64 Address) then a Kato #6 switch (1-1) Module 2 has 3 Kato #6 switches 2-2, 2-3, 2-4. Both DS64s get their signals from the "RED" line, and our red line is occasionally used for DC operations. While in DC mode power to the DS64s is from a single Digitrax branded PS14. Additionally I have a control box that has momentary pushbuttons to manually activate the switches. The pushbuttons are wired with a common coming from the DS64 into the switch then each channel is routed to the respective DS64, both DS64s share a common common (I'm thinking this may be an issue but couldn't find anything against doing this.) Each leg of the yard has isolated rails that I can control power to through a set of toggle switches, this is to allow train storage when in DC mode. OK so that's the setup. The problem is when I turn power on to any of the tracks the DS64s will start cycling the switches randomly. Not just the switches on Module 2 where the leg tracks are but all of them on both modules. The tracks are physically cut after the switch so the power routing feature of the switch will not interfere with the ability to kill power to the leg. While power is off for any or all of the tracks the DS64s operate normally, they allow control from a throttle, or from the control box. I do not have a Loconet connection in use here, as my modules my be used with an NCE command station from time to time and I wanted to still have the DS64s work. Any Ideas on what to check or how to resolve this. I cannot see any shorts, I have checked for shorts with a meter and even with the PS14 supplying power to the DS64s in DCC mode this happens. This happens in both DCC and DC mode, and ONLY when power is applied to the leg tracks. Power to the leg tracks is fed similar to the pushbuttons with a common input and each switch providing individual power to a section of track. There are jumpers under the layout allowing power to each modules' leg sections so the power drop shouldn't be a cause. Also it does it with the jumpers unplugged as well. Rails are wired WBWB on all modules, my command station is a DCS210 with a locobuffer and JMRI on a Raspberry Pi, I have a PM42 providing power management and when the switches freak out the PM42 is not shorting or even reacting. I cannot see any unusual loconet traffic when this happens either. Also a loconet cable connected to one of the DS64's didn't stop this from happening either.