Re: [CAD_CAM_EDM_DRO] Power Supply Safty

On Wednesday 29 June 2005 04:07 pm, Wayne C. Gramlich wrote:

> In this discussion about Variac's (auto-transformers)

That's _Variacs_ (no apostrophe).

> and isolation transformers, I think some very vital information did not come
> to the surface.
>
> The key thing to understand is that electrocution can take place with a
> remarkably small amount of current through the heart:
>
> <http://www.rfcafe.com/references/electrical/electrocution.htm>
>
> That is right, 100ma applied across your heart would probably terminate your
> life.

It actually takes quite a bit less than that.

> The reason why people are not be electrocuted all the time is because the
> human body has a high resistance, so it is takes some effort to get that
> much current flowing. However, once the current gets flowing, watch out!

Yep, as current starts to flow resistance drops.

> The only way to absolutely guarantee that your electrical system can not
> electrocute you is to ensure that the maximum current available is 10ma or
> less.

Actually, currents between 1 and 2 mA will cause ventricular fibrillation,
and currents beyond 2mA will stop a beating hard instantly. This is actual
current through the heart muscle itself, I don't know what the article
referenced is looking at because I didn't refer to it.

So the best way to guarantee that this won't happen doesn't have anything to
do with looking at the current level, it's to look at the current _path_.
If you don't have current flowing from one arm to another, say, or
otherwise taking a path that puts current through your heart it won't be a
problem.

Being a tech for roughly 40 years I've gotten a few shocks. Most of the time
the current path was within one hand, often within one finger, occasionally
within one arm, but never beyond that. The old advice about keeping one
hand in your pocket comes into play here.

> Since most of the Do-It-Yoursef CNC crowd are working with power supplies
> that in the 20 - 100 volt range with currents measured in amps, care must be
> taken to avoid electrocution.

Yep.

> It is very prudent to install your power supply into a metal box that is
> grounded to the 3rd prong of a grounded outlet.

Probably not a bad idea, though I've lived in a number of places that were
old enough to not have grounded outlets in more than one location.

<...>

> When we are working on these power supplies, the box is typically open and
> some additional precautions are in order. Whenever possible, unplug your
> power suppply for soldering, wiring, etc.

There's really *no* reason to work on anything that's plugged in unless you're
taking voltage or signal measurements.

> However, sometimes it is necessary to work on a power supply "live" and
> open. Done improperly, this can be quite dangerous. The key thing to avoid
> is establishing a ground path through your feet, rear end, etc. For example,
> do not work on a wet concrete floor in your bare feet. Conversely, wearing
> dry shoes with an insulating sole, while sitting in a non-metal
> chair will help avoid the ground path to a concrete floor. The goal here to
> avoid a current path from your power supply, through your heart, to ground.

Yep.

> I always treat the 120VAC section of my power supplies with great respect,
> if not out right fear. My latest CNC power supply uses a 24VAC transformer
> and I am extremely careful to use heat shrink tubbing to insulate the metal
> connectors as best as possible.

I have one item of equipment here (an eprom burner mounted in a metal chassis)
where I sure wish they'd done that -- there's an AC connector similar to the
type you see on the back of computer power supplies and a lot of other stuff
(do these have a name?) and it's wired to a switch, with bare connections.
It got me *once*.

> Here is a picture:
>
>
> <http://gramlich.net/projects/cnc/motion/rev_d/assembly/wiring/ac_after.jpg

Not bad, and that's an interesting method for clamping those caps there. If
I have the room I prefer to keep the AC stuff and primary wiring as far as
possible from the rest of the circuitry if I can. Looks like the box you're
using there might limit your options a bit though.

> If I really work at it, I can manage to touch some live 120VAC, but it is
> not easy. I assure you that I never try.

In the case of that one item I mentioned I'm thinking about a hunk of
plexiglass or similar, mounted on spacers, and that would likely prevent
even that, except for maybe the transformer connections. In fact I have a
salvaged transformer which uses barrier-type terminal strips mounted on it
for connection points and it does just that -- uses a strip of clear
plexiglass that has to be moved out of the way sitting over the connections.

> There seems to be a mentality that anything after the step-down transformer
> is "safe" since it is "current limited". Alas, at the power levels we work
> at, that is simply not true. In order to get every last bit of performance
> out of their CNC system, people try crank the DC voltage as high as the
> electronics will support.

This is pretty similar to the sort of power levels you see in a lot of
solid-state audio amplifiers too, which are getting up there for exactly the
same reasons -- performance. Though they tend to go a bit higher in voltage
and don't carry quite as much in terms of amperage. I have seen some 300W
amps with both positive and negative _80-volt_ rails, though!

> The problem with DC voltages is that they have a tendenacy to cause the
> person being electrocuted to grab harder and harder. For example, an 80
> volt DC system with the ability to punch 2 amps per coil kill you just as
> dead as 120VAC side of system.

Only if it goes through your heart or other vital organs, though.

> Please treat the entire power supply with the respect that it deserves and
> plan your actions around it to maximize safety.

Yep. People that don't will end up learning things the hard way. Or their
heirs will...