Thoughts on Tesla coil grounds...

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "Terry Fritz" <twftesla-at-qwest-dot-net>

Hi All,

There has been a lot of talk about grounding lately...  Us engineers and RF
guys seem to disagree a bit but "we" seem to still get it right.  However,
I think we are confusing everyone else a bit...  So I'll add some more to
the mess :-))

My house has a big water pipe coming into the basement right near my play
area.  So that is a natural RF ground.  The AC house wiring is also
grounded to this pipe as well as being grounded "somewhere" buy the neutral
wire in the two phase (+-120 VAC) coming to the house.  So I have an AC and
RF ground given that the house was built in 1927...  My RF ground (the
water pipe) is connected to the "stuff" through 25 feet of RG-8 coax from
Radio Shack.  I measured the inductance of its shield and inner conductor:

	RG-8 25 feet long:
	outside braid: 7uH  0.148 ohms
	inside conductor: 11uH  0.156 ohms

I then measured the inductance of the AC wire (A long (25 foot heavy-duty
extension cord).  I have jumpered this outlet's ground to the water pipe
independent of this outlet's normal AC ground too.

	AC line ground (eventually ties back to water pipe): 426uH
	Resistance:  0.247 ohms

So...  There is a giant difference in the inductance but not a great
difference in resistance...

I use the braid of the RG-8 coax as a safety ground sometimes just because
I know exactly were it is and goes, but we'll ignore it now.

So lets say I plug a AC lamp drawing 2 amps at 60Hz into the AC line.
Let's also suppose it is defective and pushes the 2 amps into the ground.
With the AC line ground I get:

	2 x (0.247 + 0.000426 x 60 x 2 x 3.14159) = 3.7 VAC

-note that I am not using the true vector functions, but it is still real
close since the inductance is the big player...

So I get about 4 volts in the AC ground.  I am pretty safe :-)  It is
interesting to note that most of the voltage on the ground at the lamp is
due to the inductance even at 60Hz!

Now...

Lets ground my 250kHz small coil into that AC ground noting that it pushes
about 4 amps peak into the secondary base lead.

	4 x (0.247 + 0.000426 x 250000 x 2 x 3.14159) = 2680 V

Yips!!  I now have 2700 volts floating on the "ground".  The inductance of
the AC line wiring ground is so high that there is a big RF voltage across
it.  2700 volts is just about where the motor and other stuff may arc over.
 I even have the AC line ground jumpered at the outlet to the water pipe to
be pretty low inductance compared to most...  I can suppose that the AC
ground does attenuate the signal so the VCR and computer only sees maybe
500 volts of that :-((

So... lets connect the small coil to the RF ground and see what we get...

	4 x (0.156 + 0.000011 x 250000 x 2 x 3.14159) = 70 VAC

So the RF ground has low inductance and we only get a voltage rise of 70
volts.  Still not real low but far better than 2700 volts.

So to make a long story short, the typical AC line ground is great for 60
Hz stuff but terrible for RF stuff.  The AC line ground is robust and
reliable for saving your life with the 60Hz faults, but RF will dance
through it like crazy.  If you are going to conduct RF to ground, you
really need something with vastly LOWER INDUCTANCE!  A short, straight, and
wide path directly to a ground rod...

"I" don't like to use RF grounds for 60Hz safety stuff.  The "real" AC
ground "always" has low resistance where sometimes the more fragile RF
ground (vice grips, spare wire, etc. :-)) can fail.  Thus, I ground the
power controller (variac and AC switches) to the regular AC line ground.
However, anything that has a chance of conducting RF goes to the RF ground
since the line AC ground is useless at RF...  I always do keep the RF and
AC grounds separate since I don't like the thought of my coil sharing the
ground with the VCR and computers...

I also have MOVs all over the place so that if the RF (or AC) grounds start
to go too high, the MOVs will conduct current to the other ground if
needed.  When thinking about grounds, think about where the streamers might
strike too.  Streamers have to connect to RF grounds and not AC grounds!

Hopefully, this will clear up my thinking on this subject or invite further
thoughts...

Cheers,

	Terry