There are diodes inside the Arduino chip which try to prevent the input exceeding the supply voltage or going below zero.
I say "try" because they are very small diodes with very little ability to carry current. 2mA is the max. So a large-value resistor can be used to limit the current while the diode limits the voltage.
BillHelmo:
Yes, exactly, how do you suddenly drop 1MW from the grid? Switch on a load of fat resistors and dissipate it as heat.
Or use it to store momentum or potential in some form like spin a mass or pump water uphill. You might only get 20-30% back but it's more than nothing and would eat surges.
Super capacitors are getting better, they may hold energy efficiently on large scales.
There is or was another energy storage developed using molten salt of some kind, runs at very high temperature.
MorganS:
There are diodes inside the Arduino chip which try to prevent the input exceeding the supply voltage or going below zero.I say "try" because they are very small diodes with very little ability to carry current. 2mA is the max. So a large-value resistor can be used to limit the current while the diode limits the voltage.
I thought diodes have a max reverse voltage. Does the greatly reduced current kind of Ohm that down?
Yes, exactly, how do you suddenly drop 1MW from the grid? Switch on a load of fat resistors and dissipate it as heat.
Heres how the UK fixes that problem - no brown outs here guys dinorwig
Mark
I have a UPS here. My PC and monitor get clean 60Hz square wave 120VAC.
Even a good line conditioner is enough, even in an electroplating company in a busy industrial park 
Something like that, I feed my Unos regulated 5V since unregulated power may vary more than .5V.
BillHelmo:
But I'm also interested to see the variations geographically, for instance does it vary from city to city, north to south, coast to mainland etc.
I have often wondered about that myself - but I had no idea you were planning a network of data collection sites with VERY tightly controlled time keeping so that you could compare different places at the same instant.
...R
Robin2:
I have often wondered about that myself - but I had no idea you were planning a network of data collection sites with VERY tightly controlled time keeping so that you could compare different places at the same instant....R
I wonder if GPS time signals could be used to do that?
Our national grid is tied in with european grids - you'll have to persuade them as well!
Allan
GoForSmoke:
I thought diodes have a max reverse voltage. Does the greatly reduced current kind of Ohm that down?
Think about the two diodes on an individual input. One is connected to GND, the other to 5V. Now apply -1000V to the input. The 5V diode sees a reverse voltage of 1005V. That's beyond the capability of any diode, yes.
But the GND diode is conducting at that point. It's trying to keep the voltage near 0V. So long as it's functioning correctly then 5V diode won't see any more than -5.6V, which is within the capability of regular diodes.
-1000V isn't unusual for static electricity. That's the level that you probably won't even see or feel a spark. But the energy in 1000V of static is very small, so the protection diodes can effectively dissipate that without burning out.
allanhurst:
Our national grid is tied in with european grids - you'll have to persuade them as well!Allan
I think that is a DC connection.
...R
MorganS:
Think about the two diodes on an individual input. One is connected to GND, the other to 5V. Now apply -1000V to the input. The 5V diode sees a reverse voltage of 1005V. That's beyond the capability of any diode, yes.But the GND diode is conducting at that point. It's trying to keep the voltage near 0V. So long as it's functioning correctly then 5V diode won't see any more than -5.6V, which is within the capability of regular diodes.
-1000V isn't unusual for static electricity. That's the level that you probably won't even see or feel a spark. But the energy in 1000V of static is very small, so the protection diodes can effectively dissipate that without burning out.
Cool for just how it works!
A digital read is supposed to eat 1 microAmp (AVR datasheet), perhaps it is through a clamping diode?
And it can save on parts count.
And it helps me understand why I can feed a piezo through a diode to a BJT gate and not hurt the diode or BJT, those spikes can get pretty high (piezo smacked HARD with a screwdriver handle) but are current weak but I think to not try it with a FET!
If you split the mains wires apart yet still draw current through them, wouldn't a Hall sensor aligned with and very close to one of the wires see a constantly reversing though weak field? Perhaps a coil would be needed?
If the mains wires are not separated, field of one will cancel field of the other.