Wire size.

[centenary]

Hello everybody.

Sorry if this table does not format too good. Wire sizes, as can be seen below 18 awg multi strand wire is well capable of handling up to 20 amps and even 20 awg upto 12 amps. Ive seen suggestions of using 16 and even 14 awg cable rated for 30 amps and 45 amps respectively for the bus. Seems massive overkill to me unless you have very long bus run. But even with a 10 metre run you arent going to get much of a voltage drop, in fact in the order of 1.5v and going to the next thicker wire will only reduce that to around 1.3v

AWG Area Conductor Diameter Current
20 0.518mm² 0.812mm up to 12A
18 0.82mm² 1.02mm up to 20A
16 1.31mm² 1.29mm up to 30A
14 2.08mm² 1.63mm up to 45A
12 3.31mm² 2.05mm up to 70A
10 5.26mm² 2.59mm up to 120A

So what size do you use?

[RFS]

The issue here is ensuring that the DCC signal superimposed on the waveform is in no way compromised. That's why the recommendation is for heavy duty bus wires. A 1.5v voltage loss on a DCC bus would almost certainly lead to a compromised DCC signal.

[TimberSurf]

I don't know were you got your table from, but the current values are outrageously over what they can carry!
See here as a good example
This thread covers the recent discussion viewtopic.php?f=7&t=52718

Droppers can be 22AWG or 24AWG and bus can be from 18 to 14AWG dependant on the size of layout, size of power supply and current of max loco's (consist)

Metric MM2 (Diameter) AWG Approx Stranding
0.45mm 24 awg 7/0.2
0.5mm 22 awg 16/0.2
0.75mm 20 awg 24/0.2
1mm 18 awg 32/0.2
1.5mm 16 awg 30/0.25
2.5mm 14 awg 50/0.25
4mm 12 awg 56/0.3
6mm 10 awg 84/0.3
10mm 8 awg 80/0.4

As long as the wire is rated above the max power the supply can feed it won't burn out, but bare in mind the rails will also carry some current (allowing smaller droppers). Past that, the bus will be sized to assist with volt drop (nothing to do with current carrying capacity)

[Roger (RJ)]

You also have to ensure that the resistance is kept to an absolute minimum to ensure that the controller trips out reliably in the case of a short circuit anywhere on the layout, even at the greatest distance from the controller or booster.

[TimberSurf]

The issue here is ensuring that the DCC signal superimposed on the waveform is in no way compromised. That's why the recommendation is for heavy duty bus wires. A 1.5v voltage loss on a DCC bus would almost certainly lead to a compromised DCC signal.

I have to disagree! Volt drop will have little impact on DCC, unless the PSU is 12V,as decent DCC boosters are in the order of 17V! The voltage can be set on some, all the way up to 22V! Small cables will reduce voltage and have little influence on the waveform. The reason for a copper bus in parallel to the rails is to overcome high resistance (that causes arcing) in the potentially poor electrical conductance of the fishplates and (if a very large layout) volt drop. It is also to help eliminate bundles of feed wires and complex "tree" of wires to individual (multilane) tracks by simplifying the wiring.
On very large layouts, as Roger says, there is a potential to not create enough resistance in a short to trip the controller, so as said in the other post, powerful boosters should be broken up into current limiting districts with suitable protection.

[RFS]

The issue here is ensuring that the DCC signal superimposed on the waveform is in no way compromised. That's why the recommendation is for heavy duty bus wires. A 1.5v voltage loss on a DCC bus would almost certainly lead to a compromised DCC signal.

I have to disagree! Volt drop will have little impact on DCC, unless the PSU is 12V,as decent DCC boosters are in the order of 17V! The voltage can be set on some, all the way up to 22V! Small cables will reduce voltage and have little influence on the waveform. The reason for a copper bus in parallel to the rails is to overcome high resistance (that causes arcing) in the potentially poor electrical conductance of the fishplates and (if a very large layout) volt drop. It is also to help eliminate bundles of feed wires and complex "tree" of wires to individual (multilane) tracks by simplifying the wiring.
On very large layouts, as Roger says, there is a potential to not create enough resistance in a short to trip the controller, so as said in the other post, powerful boosters should be broken up into current limiting districts with suitable protection.

The point I was making is that if your DCC bus wiring is causing a 1.5v drop then it's too thin, and one of the side-effects of this will be a compromised DCC signal. It may also lead to a command station not tripping when there's a short which can result in serious problems.

I use 2.5mm solid copper for my bus on a layout that's 30ft long. It's from twin-and-earth wire from the likes of B&Q. This wire is relatively inexpensive and you don't need a lot if it.

[Gordon H]

Wire current ratings are not just about the absolute maximum current a wire will conduct before melting.
Generally, the quoted rating for a conductor is the current which will produce a temperature rise in that conductor which is in the acceptable range of the insulation material it is coated with.
For example, you will find that a PTFE covered wire will have an apparently much higher rating than a PVC one of the same cross-sectional area (CSA).
Here, the aim of the game is to achieve optimum power transfer between the command station and the locos.
What is more important is the overall resistance of your chosen conductor(s), considered as a complete circuit path. Hence, for the long runs (i.e. the bus) you use thicker wire to minimise voltage drop along its length (very low Ohms per metre), and limit the use of thinner wires by only using them to perform the final connection between the bus wires and the running rails.
As a simple rule of thumb I would suggest the allowable voltage drop between any two points (i.e. locations) anywhere along the path should be no more than 0.5V at maximum load, preferably much less.
Unfortunately, these things are never as simple as many people would like to believe they are, and some engineering is required to produce predictable, reliable and repeatable results. Ohms Law will always win...

[Bufferstop]

If you use the cores from mains wiring cable, strip it from the cable so that it looks nothing like mains wiring, and if you need flexible connection to the control panel find something that isn't blue and brown. We were testing a cupboard full of computer bits and pieces. One item was a pair of speakers with a two core mains lead dangling from it. No plug, no problem we had one of those snap down terminal blocks on the end of the bench so my mate pushed the wires in and snapped it closed without a second thought. Very loud 50Hz hum followed by a bang. The blue and brown were the wires to the second speaker, it should have been powered by a wall wart adaptor. Yes my mate should have been more attentive, but who thought a bit of old mains lead was a good choice for a speaker lead. 230 volts up your bus could do a lot of damage.

[RFS]

I think I'm savvy enough to recognize my own DCC bus wires and not start thinking I ought to connect them to the mains. In the US, the recommended colours for your DCC bus are red and black. Why? Because these are the colours used for mains wiring there, and wire in these colours is readily available at all DIY electrical stores.

[Gordon H]

In the US, the recommended colours for your DCC bus are red and black. Why? Because these are the colours used for mains wiring there...

Don't think so.
Black is live, and white (or grey) is neutral on single phase installations, which are what most people would encounter.
Red is only used as the second live in systems with more than one phase.

[Forfarian]

Why would anyone leave loose ends on their power bus for some idiot to connect to the mains??
I have my feed at one end and at the other it is screwed into a connector box which is screwed to the base board, simples no loose ends.
Black and red for the bus with black and red droppers, cant go wrong