Transformers isolate the load completely from the primary power source. Does that mean that I can connect two or more transformers one after the other and expect my Linear Power Supply to have better safety than having the isolation of one transformer?
For example, lets assume I need 12V AC out. The situation I am trying to imagine here is having 230V/50V and 50V/12V transformers.
Would this make my system more immune to lightning?
Are there better ways of doing the same?
Tranformers can protect against lightning in two ways- first they have galvanic isolation from primary to secondary that is good for hundreds, and more likely, thousands of volts of transient voltage. That's good for common-mode voltage (voltage on the input leads wrt earth). Secondly, they saturate and will prevent high normal mode (across the input leads) transients which may be induced by a lightning discharge from reaching the circuit on other side (they might let double their rated voltage through, but not 10x or 100x).
The problem with the galvanic isolation is that it may not be enough for a direct lighting strike. It's thus better to find some way to conduct the discharge to ground through a lightning arrestor (basically a spark gap) so that the voltage doesn't rise too high. For transients on telecom lines, very small spiral gas discharge tubes as "Enemy Of the State Machine" mentions are available.
The way to deal with such a huge amount (potentially) of energy is to divert, then soak it up in some series impedance, divert again, and some more series impedance until what's left is harmless. If you try to divert or absorb it all with some small device, the end result will be a couple of disembodied smoking leads, and no small device any more. In many cases there will be a natural spark gap in the creepage and clearance distances in your input plug, switch, terminal blocks etc. but it's not always wise to depend on those, as they may break over after your transformer does. If your "first in line" transformer breaks over to ground internally, it will probably be ruined (assuming it 'tracks' through the insulation), so your second transformer may save the circuit but a service call is still required to replace the transformer.
The insulation on a standard, consumer-grade transformer, would break down and arc over. This would allow those high voltages to jump right through them. A popular and effective way to take care of surge protection of such high levels are Gas Discharge Tubes. They are readily available and come in small packages these days. You can put them in before your transformers and help save them too.
The company Littelfuse makes them in all shapes and sizes: http://www.littelfuse.com/products/gas-discharge-tubes.aspx
There is info on their site about the operation of this type of device.
The isolation transformer frame should be tied to the earth ground (the preferred discharge path for the predominant energy dump you are going to get from a nearby or direct strike. The MOV and GDT technology devices go across the primary and each leg of the primary to that ground.
I was thinking that adding surge protection to the frame ground on the secondary side would be good, but stray capacitance coupling effects with the devices could "defeat the purpose" of the isolation transformer for its emi/rfi isolation.
My personal experience is based upon having personally witnessed TWO direct strkes to my ham radio tower, while I was passing digital traffic. No damage to the equipment, but the hockey puck sized GE MOVs blew like a shotgun both times. The events were 6 months apart. There was an 1800VA isolation transformer between the 240 from the panel, and the 110 to the various radios, IBM 80's vintage pc, etc..
