DaSeitz wrote:To cut a long story to almost 0:
I need an EMP proof storage solution, or at least a EMP proof Backup.
I would love to tell you the field stength, but there is no chance of predicting it.
Greetings from Bavaria
I've done some radio field installs, and they're designed for lightning. Used quite a bit of PolyPhaser gear, seems to work as advertised, though not sure what they're curent offerings are. They had some Ethernet supressors last time I checked (long time ago). They have lightning-rated parts and EMP rated parts, it's all in the clamping effectiveness for the risetime of the pulse, EMP being much 'faster' than lightning. And they have a mean-assed looking simulator in their labs.
The basic trick is to use supression gear for every external cable
at a common entry point bonded to a common panel, thus all of the equipment in the rack can never stray any more than the rated surge voltage away from the grounding panel. The entire rack rises above ground potential and falls again after the event. Of course, this does nothing good for the external gear, but the stuff in the rack should be good.
So, with entry supressors which should allow firewire, Ethernet or USB, or Fibre Channel signals to pass unfettered, referenced against the input grounding panel, and with the AC input also referenced against this ground. If a pulse comes in through a cable, there will be a sudden voltage differential between 'outside' and 'inside', and the supressor clamps would maintain safe voltages for the equipment inside the rack, instead shunting the entire voltage differential to equipment outside the rack, more than likely destroying it.
I've personally seen a radio site in Jamaica up in the hills where the electrical panel almost got blown off the wall, and upon plugging the rack back into a live circuit, it came right up without any problems except the text on the monitor in the rack was now tilted about 30 degrees... man that had to have been some whack. The GaAsFET preamp ahead of the receiver multicoupler wasn't even damaged in the slightest, nor was the attached battery backup bank and charger.
How thick does the metal in the cabinet have to be? It would seem to me so long as the entire cabinet itself was RF tight and made of reasonable-gauge metal, it should never have to carry enough current to destroy it. It just has to be able to handle the induced current due to the propagating pulse, and if it's a flat sheet it should be able to handle a lot without failure. If the cabinet (and contents) are 'riding the pulse' so-to-speak, even less so. With a good metal cabinet, with good metal drive arrays and such all bolted together, I can't see enough of an EMP pulse getting inside to do much damage, at least compared to a direct entry via a cable. Cover the inside with copper screening, that'll likely get another 20 dB or more isolation.
Of course if you're instrumenting a blast site, then there may be issues, likely and magnificiently subsumed by what needs to be done to protect from the mechanical blast wave.
Wonder how WiFi works through 40-50 dB of isolation? Then you could just run AC into the rack, have an internal server/WiFi, and connect from outside, no wired data at all. Could be as simple as that. Paint 'er up like a British telephone booth with the time you save.
The real deal, though, is stuff like flash detectors that mechanically clamp inputs ahead of the EMP pulse.
But, no matter how you slice it, whatever's outside of the cabinet should be considered disposable in the plan, deferring to data safety instead. Otherwise the problem can't really be effectively dealt with.