Spaghetti

Nothing revolutionary, just that good practise is to actually plan the arrangement of what the panel contains, giving thought to wiring runs and accessibility; and leaving space for expansion if there is any possibility of the layout growing in size or complexity and thus requiring more supplies and control gear.

Documenting the panel arrangement, and using a labelling or numbering system to identify components both on the panel and on the layout becomes progressively more useful as complexity increases.

Systematic identification schemes work well for many people. So for example everything to do with point control is prefaced 'P'; and switch P001, wires P001 and motor P001, means you can trace confidently all the components of one complete circuit. (It is possible to considerably elaborate on this type of scheme, but that's down to individual taste.)

I don't have a control panel as such, but even with DCC and computer control coming soon there is plenty of wiring to confuse. I've done the following:

(1) Define a colour/thickness coding and stick to it. For example all my track commons are black, DCC feeds are red or grey by power district and block detector feeds are brown. Point motors are thick white for left, orange for right and blue for common with thin orange, white and grey wires for the frog switch. And so on. Then if there are several wires associated with the same device it is easy to know which is which. Each point/signal/track circuit is also named using a system based on the prototype.

(2) I undercoat and paint all my boards underneath with water-based white satin paint. This is partly a precaution against warping but I've also found it makes it really easy to write on with a permanent marker. So wherever a wire goes through a frame or into a connector block I have marked what it does.

(3) Also really useful to be able to remove each board individually and stand it up on edge to do the wiring. In fact most of my board-to-board cables have enough slack to remain plugged in when I do this, so I can even test things with the board up. Wiring is much easier to do and to understand if it's easy to get to, no above your head!

I've now wired about 60 detector sections and about 40 points (no signals yet!), with only a handful of wiring problems and each of these can be traced within 5min or so.