I've seen this question start very heated arguments elsewhere, so I hope we can be more civilized here! I've also seen one retailer claiming different suitabilities for the same wire depending on how you were buying it, so maybe there just isn't a good answer. But I'm going to ask anyway:
So obviously you need something more substantial than the 1A stuff used for (DC) track and accessories. But is 3A sufficient? What if you've got a CDU driving multiple points?
Is it the wire from the CDU to the solenoids or the wire from the power supply to the CDU? (Not that it matters much, having said that!)
For recharging the CDU, there should be a relatively small trickle of current so the wire current carrying capacity need not be high.
For the current from the CDU (or stud and probe or passing contact switch) the current will flow for just a short time, so the wire is unlikely to heat up and fairly thin wire could possibly be used.
However... more important is the length of wire between the CDU/switch and the solenoid as a long thin wire will drop more voltage than a short fat one and there may not be enough "oomph" to change the point.
The same goes pretty much for most layout wiring - it is volts drop, not current carrying capacity that is important. It doesn't really matter if you lose a few volts in a mains extension cable carrying 230 V, but is more of an issue if there is just 12 V to start with. For a given current the volts drop will be the same whether you start with 230 or 12.
If you look at the wires in the solenoid itself it is probably quite thin, much thinner than the wire connected to it. Clearly it is capable of carry the current for short period of time the solenoid is energised. However it cannot sustain that current for any length of time without overheating as it is so thin. It is also bunched in a winding so the heat cannot escape (which is why you should alway unwind an extension reel fully if it is powering anything "hungry").
Ordinary "hookup" wire should do fine.
To add to what Graeme has put:
All wire has resistance [measured in ohms, usually per metre] The material and cross sectional area of the wire determines its resistance. The greater the cross sectional area the greater the cost (and weight) AND the lower the resistance. (there are other influences but generally we don't need to consider them here). This resistance causes the voltage drop so we want minimise its effects.
The 'usual' wire I see for layouts is copper, 7/0.2 (about 0.22mmsq) or 16/0.2 (about 0.5mmsq). 16/0.2 is about twice the cost of 7/0.2 so it make sense to me to use the smallest stuff that will do the job.
Usually 7/0.2 will do.
As an example:
I have a Lenz LS100 stationary decoder driving 4 Seep solenoids, 3 of them are close to the decoder (within 500mm), the fourth is about 1200mm away. 7/0.2 was fine for the first three but the forth would only budge intermitently. Knowing why I substituted 16/0.2 and away it went, every time.
The voltage should be the same on all outputs, the thinner wire obviously has too greater resistance [because its longer] to allow the required current to flow in the forth motor, substitution with the heavier wire lowered the resistance, allowed the current to flow and do the work.
Making sure your point motors operate freely is, of course vital. If its stuck a large belt may shift it but for how long?
There is nothing inherently wrong with 'doubling up' on 7/0.2 to give 14/0.2 (about 0.44mmsq) almost as good as 16/0.2 albeit probably slightly higher cost per metre but it may save you buying the heavier size, but likely to be more hassle to install and maintain.
Ordinarily I now use 16/0.2 for connections from CDU to solenoids, but then again I try to avoid solenoids all together. IMHO slow acting motors are Slow much better (if you'll pardon the pun).
For multiple solenoids firing simultaneously you need a lot of amps and a chunky firing button (else it'll burn off its contacts) but the wiring requirements will be the same if you wire them all individually from the solenoid. Multiple coils in parallel may well need heavier wiring.
HTH
Andy
I have used 16/0.2mm 3A wire for all the wiring on Westdorf that has solenoid point motors with CDU
If required, there are a number of online cable calculators available. Unfortunately most are of American origin so the cable sizing is not convenient for UK and European use. However, this one lists metric wire equivalents further down the page which suggest (for example) that 16/0.2 wire (16 strands, each 0.2 mm diameter) is equivalent to 20 AWG.
http://www.calculator.net/voltage-drop-calculator.html?material=copper&wiresize=33.31&voltage=12&phase=dc&noofconductor=1&distance=30&distanceunit=feet&eres=0.3&x=66&y=20 (http://www.calculator.net/voltage-drop-calculator.html?material=copper&wiresize=33.31&voltage=12&phase=dc&noofconductor=1&distance=30&distanceunit=feet&eres=0.3&x=66&y=20)
So, say for 30 feet between a controller set to 12 V and a track feed point the voltage drop between controller and track would be 0.18 V for a loco drawing 300 mA, nothing of consequence but say the wire had to carry 6 A (suggested as the maximum current for the wire gauge) the voltage drop would be just under 3.7 V.
I have used 16/0.2 on my layout for distributing track power and for the point motors.
Thanks everyone. I'm not expecting to need to run more than a couple of metres (6' or 7' in old money), so the large pile of 7/0.2 to hand will cover all track, LED lighting and non-solenoid accessories requirements. What I need to work out now is whether all the points should be wired up individually with 16/0.2 or to do passing loops in pairs with something a bit heftier.
I would do them individually with 16/0.2, for me it makes testing and / or fault finding easier.
Why not try the most remote point motor with 7/0.2 and see if its satisfactory?
Andy
I would suggest you wire up all the point motor individually; pairin them up only at the control box - it may help with fault finding if you have problems at a later date. (Also - keep a wiring diagram!)