Please see diagram. I'd like to use the method of having a power and ground BUS and supplying to layout via feeder wires. Do I basically have the right idea/connections? Also, if anyone has any tips or things to look out for I would appreciate your input. Thanks. Marklin Wire Diagram.pdf (104kb) downloaded 262 time(s).





Marklin Wire Diagram.pdf (104kb) downloaded 262 time(s).

That looks right to me, the loop ensures you get the lowest resistance path to any feeder. As Ray indicated, take feeders every once in a while.

For the taps on the bus, I'd use screw terminal strips rather than soldering as its more configurable and avoids the issues of soldering above your head. Another option is the Wago friction/clap connectors.

I'm only skeptical about the Beer... as I don't know what your using for that part of the solution. I'm reminded of the Monty Python "Bruces" sketch where they describe American beer as being very much like "making love in a canoe".

looks right. this is how I suggest you connect the track feeder wires to the bus wires.
http://layout.mixmox.com/1/20020224Wiretaps

bus wires should be insulated and not bare.

I assume that this has been discussed before but I have not yet found any threads on this topic.

I have been wondering if cutting in the signal at multiple locations in a circuit might cause problems with the digital signal. The only advice that I have seen from Märklin is to have a bus with connections to the track every 1-2 meters, but nothing about the architecture of the bus beyond that. It would be great to nail down the specifics before possibly constructing a power bus in the wrong configuration.

Its a conceptual complication to think of the idea as a big loop as being a signal conducting problem, on the scale we are dealing with.

If you lay out the power/signal bus in a big loop on your layout, then take the point in the middle of that loop and consider that as the 'far endpoint', and where the controller feeds the power/signal in as the 'near endpoint', and then just pull the far endpoint away from the near endpoint, until the big loop just becomes two straight runs that are parallel from the near to the far endpoint. That is exactly the same electrical configuration (two long parallel wires) as the same run arranged as a big loop, as far as picking up power/signal from the wire.

Yes, in a theoretical sense, for a point that is NOT on the midpoint from the control injection (trackbox/central station) injection point, it can see the digital signal arrive first on the shorter path, then again on the longer path around the loop. This is the traditional multi path problem. However, that loop would have to be a LOT larger for that to cause a problem (or the frequencies have to be a lot higher than the 8-40kHz space we are in) I think.