The other thing that needs to be borne in mind is switch on surge. If the mains is switched on at mains zero crossing, you get doubly hit with magnetizing current for the transformer core, and charging current in the capacitors (although the magnetizing surge is much more than capacitor charging). Worse with toroids as compared with open frame because of low leakage inductance.
You can easily get an initial surge when the core has no field determined by the mains voltage and the primary resistance. I've just measured a 300VA toroid and the primary resistance is 4.6 ohms, so the maximum surge can 52A (240/4.6) until the field in the core is established (1/4 of a cycle - so 5ms) - then the primary inductance controls the current. But a 50-odd amp spike can trip the breakers in your house even if it does not blow the equipment fuse.
There are various schemes to manage this surge to do with cheap and cheerful passive using thermistors, or usually more sophisticated relay based delay. So on startup, a power resistor is in series with the transformer primary (10 or 20 ohms, 20-50W - that sort of range) and then after a delay of around half a second delay, relay contacts shorts out the power resistor.
Great fun and games!
Craig
PS This is far too serious stuff for the topic of the thread