I am wiring a new HO scale layout in DCC. I am planning to use a pair of Tam Valley Hex Juicers for AR polarity control of six reversing segments. The input to the hex juicer accepts 14 AWG wiring from the DCC track bus. Output is limited however to connection with 20 or 22 AWG wire.
I am aware of relationship between wire guage size and run length to voltage drop (and degredation of DCC signal). Locating two hex juicers more or less centered at opposing ends of the layout will minimize the lengths of the 22 AWG feeder lines from the juicers to the track segments. Still, at least some of the 22 AWG feeder lines could exceed 3-4 feet in length, and that is longer than the commonly recommended length for 22 AWG feeder lines.
So my question is, can I alleviate the potential for excessive voltage loss by connecting the 22 AWG wire from the output of the hex juicer to a 14 AWG segment running to another connection to a 22 AWG feeder within a few inches of the connection to the track?
First, is there any potential gain to be had from the larger guage wire between the AR device and the track feeder wire? Or is the short 22 AWG connection at the output of the AR device a controlling factore ("choke point"), such that there is no theoretical reduction in voltage loss to be gained by splicing a larger AWG segment into the feeder connection.
Second, would the two added connections themselves provide sufficient additional resistence that would likely preclude any net reduction in voltage loss?
I appreciate that, even if there could theoretically be a reduction in voltage loss from use of a larger AWG segment in the feeder line connection, there might be a trade off created by the additional connections. Under those circumstances, does anyone have an idea (or even just a guess) on how long the larger AWG segment would need to be in order to offset the detrimental effect of the two additional connections in order to yield a net reduction in voltage loss? Or would the additional connections present an insurmountable obstacle, either in terms of voltage loss or in degredation of the DCC signal.
Any guidance, even purely theoretical, would be appreciated.