Electrical hacking (was Re: [Techtalk] [OT] charging USB devicewithout a PC)

[Maria Blackmore]

On Sun, 13 Apr 2003, Alvin Goats wrote:

> It is actually the amount of electrical energy in joules (aka
> 'watt-seconds') that can kill you.

ie, the graphs of current versus time

> Too many people are injured each year due to the "ground loop", an
> essentially zero voltage many ampere current handling lead.

Ground loops are where a big loop (or sometimes not so big) loop acts as
an antenna and/or coil and picks up the general electronic crud that is
flying all around us, either through magnetic induction or as an
antenna.  This means that you get a voltage between two different ground
connections.  The highest I've ever met this difference being is 4V, which
was at a university with it's own power distribution system and power
station, it caused immense quantities of havoc with all the audio
equipment and we had to go to great lengths to sort it all out.

nevertheless, as mentioned it can be a very large source of current,
sufficient in fact to melt wires in some cases.  I've never heard of it
directly killing, I must admit, because the low voltage is usually
insufficient to overcome the resistance of the skin, and the current that
passes is equal to voltage divided by the resistance.  It's perfectly
possible though, and responsible and correct grounding is always
neccesary.

Grounding is always a problem in the audio industry, and much more so in
america than europe.  In america, if audio equipment is improperly
grounded there's a risk that you'll get "bitten" by a microphone or guitar
or something, because there's 110V between it and the local version of
ground, because something isn't grounded correctly.  There are recorded
instances of people dying from it :(  It's something that is usually
avoided, most simply on the basis that it's really bad business practise
to kill your clientele :)

<snip stuff about capacitors>

> DC power supplies taken from AC lines have one large filter capacitor to
> smooth the ripple from the AC line.

Or multiple capacitors, the one behind me has 6x 20 mF capacitors.

> Rectifiers take the sine wave AC and 'flips' the bottom half of the
> sine wave up on top, so you have repeating 'hills'. These hills are
> still AC, it just doesn't go negative anymore. To smooth the ripples
> out, you add a capacitor which takes time to charge and discharge. The
> amount of time should be very long and can be lengthened by adding a
> coil (inductor or choke, aliases for the same thing). The smoother the
> ripple pattern, the larger the capacitance of the capacitor.

Hmm, it's usually easier to use a higher AC frequency, so that the time
between the peaks is shorter, so the capacitor doesn't need to last so
long.  Which means you get less ripple, and a smaller PSU.
Good things all round :)

> FYI: the picture tube of monitors and televisions are a large plate
> capacitor. Being under vacuum, there is very little current leakage
> through the dielectric, so they hold their charge a LONG time.

They usually have a high value resistor to discharge this over time when
turned off, for safety, without adversely affecting the operation when
turned on.

It's recommended to use something insulated to short it before commencing
work.

<snip stuff about electrons and torture>

> Isolation transformer function is to create an AC path that is not
> directly coupled to any power line main. Typically, they are 1:1 turns
> ratio, but not necessarily so. Any short that can cause damage to the
> power line main requires an isolation transformer so that the
> transformer is damaged and not the power line grid. It doesn't always
> work as sometimes the primary winding gets shorted as well and will take
> the grid down anyway.

This is due to it melting under extreme overcurrent situations, and
forming a short circuit

<snip stuff about getting power to awkward places>

> 230 V center tapped isolation transformers are set up to deliver 230
> volt and two 110 volt lines. Using the two ends yields the 230 VAC,
> using the center tap and one leg gets 110 VAC, the center tap and the
> other leg gives the other 110 VAC. To 'load balance' the transformer so
> that you don't burn out the secondary windings on the transformer, you
> would use both lines. Often, the AC electrical outlet that is attached
> to an isolation transformer used by construction has one socket wired to
> one leg and the other socket wired to the opposite leg just to force
> load balancing (current sharing). 

Hmm, deja vu


> My worst experiences:
> 
> 100 VDC battery rated at 4amp-hour: accidentally touched both ends with
> the same hand. Hand clenched into a fist, curled my wrist and folded my
> arm at the elbow for about 10 minutes. Hurt like heck.

I knew someone with "yelnatS" on the palm of his hand.

Famous (almost) last words "but it's only 24 volts".  Indeed.  Why
do I always get to say "I told you so" ?

> 15 kVolt AC, 60 cycle: hurt like made and jumped about 4 inches to get
> me.

awww, it likes you.  It only wants to be your fweeend.

> Watch out for the humidity in the air, it reduces the dielectric of
> air and lengthens the distance for high voltage to jump.

dielectric strength, or breakdown voltage.
nasty.
 
> 400 Volt AC, 400 cycle:

Hmm, you used to work in the army?

> I deal in energy, lots of it on occaision. Highest current: 100,000 Amps
> at 230 Volts.

Chunky.

Fry me an egg sometime :)

> Highest Voltage: 500,000 Volts, 0.25 Amps.

mmmmmm, tesla coil, mmmmmm

> Highest Energy:  150,000 Volts at 10,000 Amps.

my my, what a big generating station you have.

> Largest ground loop current I've encountered: 150 Amps

Made life interesting, then?

> No matter how good you are or how careful you are, you're still going to
> get 'bit' by electricity.

I try to avoid it.

It hurts.

I'm allergic to pain, you see.

Maria