Transformer Safety
I would just make one comment on the PH-180 having a very good breaker, as mentioned above. That is, such a statement is misleading. The device actually inside is a supervised relay, not a breaker. It is very fast, being intended to interrupt small overloads that would cause its output to exceed 180-watts, as mandated. I understand it operates at 10.5 amps being exceeded. Its internal voltage drop at this amperage would reduce the output voltage a bit below its nominal (no load) voltage of 18 volts.
If the track is shorted close to the PH-180 transformer, the amperage output may rise to slightly over 100 amps (I have calculated 110 amps, but it may be slightly different). This would be in a worst circumstance of course, such as loosening of the screw that mounts a Pulmor motor to its power truck. The screw goes down the third rail like hard chalk on a blackboard, trailing a shower of sparks. Truly spectacular. (Large Z transformer on very long, small (#16) leads.)
If the internal relay in the PH-180 interrupts such events, its contacts will be damaged, and eventually the relay will burn open, or weld closed (permanently in either case, and welding can go unnoticed for a long time). This relay case is only about the size of a sugar cube.
And I'm not sure that typical thermal breakers, with their curves for 120-volt systems, have any quickness at less than 100 amps. That's where the special output plug on the PH-180 comes into play. At a minimum, a line-side shed relay should be used to connect the output plug to the track. The contacts in this puppy are much more substantial than those in the PH-180, yet I think still light enough that sub-cycle operation is possible (i.e. < 1/60 second) .
To make this arrangement, called cascade protection, work, the relay in the PH-180 obviously must delay somewhat, possibly two or three cycles at maximum expected amperage. This is called coordination (with the downstream heavier breaker). Often a circuit in which a capacitor is charged thru a resistor is used, to oversimplify. Such circuits would be the supervisor of the relay and the downstream breaker. The alternate choices, variable-voltage downstream devices, are quite reliable and use supervised MOSFET metal-oxide field-effect transistors in pairs, which are very effective at handling heavy shorts; of necessity these are always their very own very quick protection.
Some factoids, in no particular order:
1. Thermal breakers for 120v (household supply) include those often mentioned here, such as the P&B (now Cisco? or is that changed again?) X27 or X28 series (I can't recall exactly) have curves (amps vs time to open) similar to your household breakers, at least in the larger 15 and 20-amp sizes.
2. A recent study of household shorts (extension cords burned off at the outlet) showed that most such had a value between 1000 and 100 amps. The most significant factor was how far the outlet was from the panel. In a detached house, the limit is considered about 100 feet (200 feet counting out and back wiring. And, the wires used for this are copper, #14 for for 15-amps; and #12 for 20 amps. The resistance of these wires limits the current to the range noted, for the most part. There is a small amount of impedance added by the breakers. These figures may be applied to the toy train layout, adjusting for amps, wire size, and distance.
3. The amps and wire size arrangement above may be extended to smaller wires for toy trains as follows, assuming the distances to be a bit smaller (just my opinion here & also note that regular wire is 60-Centigrade, while most home stores sell 75-C neoprene-jacketed wire as that is often required in rewiring; also the small Lionel wiring in locomotives has been very high-- i05-C, with motors this or higher):
3a. #16-- 10 amps, 12 if higher-temperature wire.
3b. #18-- 7 amps, 10 if higher-temperature wire (PH-180 has higher temperature wire).
4. Smaller wires used by accessories and switches should be protected. Fuses are usually wanted, as breakers below 15 amps are not very accurate unless individually calibrated (as postwar Lionel breakers 5-amp to early 15-amp breakers were).
5. Do not use a single common return wire for two separately fed hot wires in the track supply circuits. As #18 is about the largest stranded wire practical in the Fahnestock clips of a lockon, consider the use of double lock-ons when appropriate, at drops from feeders, or direct from larger transformers.
6. In power circuits, the current flow is for all practical matters not compressible. That is, it is the same for all points in the circuit. So, assuming that the only likely shorts are derailments, the order of cascaded devices or whether they are in the hot or common wire does not matter. However, they are customarily placed close to the power supply to also protect against wiring shorts. Note that a line-side (120v) breaker cannot protect the load side of a transformer. Any such device is a protection against an internal short only, in the transformer. Generally external breakers or fuses are avoided in the common returns. See following.
7. All transformer commons should be tied together, preferably as close to the transformers as possible, and all transformers phased. Note that different outlets may be on separate phases as two phases are usual in residences. In dining rooms and kitchens the two receptacles in a single outlet may be on separate phases, depending on the jurisdiction (usually these are 20-amp circuits). So phasing should be checked carefully. The purpose is to prevent the "touch voltage" from exceeding 30 volts. More than 30 volts is unsafe; without proper phasing, voltages approaching 60 could be encountered. (The purpose of the ZW was to make a single transformer adequate for most people... a single transformer cannot be mis-phased, all of its common return are permanently tied together, and its outputs are all permanently in phase.
9. Do not tie the various hot outputs together, as this may lead to the 120-volt plug prongs of a disconnected transformer becoming energized by another active transformer. Rolling stock with two rollers standing across an insulating 3rd rail pin may connect somewhat different voltages, which can create circulating amperages and damage transformer(s) if prolonged (implied requirement is 1 minute resistance; but not sure UL tests this). Try marking such points.
10. The goal of all the above is to avoid two things-- dangerous touch voltages, and melted thermoplastic wire insulation. Both are serious matters, as the hazard is serious injury. So, do not work under an energized layout. And, it will also help to avoid building one's own transformers, for most people.
--Frank