Grounding problems / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

Grounding problems

1 2 next

Move Thread

LAN_403

lhl_henrylui

Fri Feb 26 2010, 11:50AM

Registered Member #1498 Joined: Thu May 22 2008, 07:08AM
Location: Hong Kong
Posts: 72

I've read a lot of materials about tesla coils but I am really confused about the grounding of a TC.

Some threads in this forum mentioned that the TC cannot be grounded to the mains ground, otherwise it is going to destroy the precious equipments and it is said that a TC should be grounded to a metal rod which is hammered into the earth.

This website says that a TC should be grounded to a RF ground. I only know that it stands for "Radio Frequency Ground" but I know nothing more about it and I don't know how
it differs from the earth ground that uses a rod hammered into the earth.

I have got Daniel McCauley's MiniBrute Reference Design Book, it says that a copper strip attached to the base of the secondary coil is the RF ground, this allows the current to circulate within the secondary coil.

However, those measures above even make me more confused. I'm wondering which type of ground I should choose, the RF one? A rod immersed in the earth? Or use both?

Nevertheless, I should never connect together the primary circuit and the secondary circuit, otherwise I'm seeking for death, right?

Kipmans

Fri Feb 26 2010, 01:39PM

Registered Member #91 Joined: Thu Feb 09 2006, 03:03PM
Location: The Netherlands
Posts: 45

"RF ground" is just the name for the ground point you use for your TC. So that could either be a rod hammered into the earth, a copper strip attached to the base of the secondary or even the mains ground.

Personally, I haven't experienced any problems when using the mains ground for a small SSTC, but still I wouldn't advise using it as I think it is better to be safe than sorry...

lhl_henrylui

Fri Feb 26 2010, 04:29PM

Registered Member #1498 Joined: Thu May 22 2008, 07:08AM
Location: Hong Kong
Posts: 72

So which one is better? The copper strip one or the rod hammered into the earth?

radiotech

Fri Feb 26 2010, 07:57PM

Registered Member #2463 Joined: Wed Nov 11 2009, 03:49AM
Location:
Posts: 1546

I think a RF ground (system) is one which has low impedance to the frequencies that are dealing with. The reason is you dont want voltages *across* the lengh of the ground wire.

I test my ground wires with a RF Ammeter like this one available from MFJ. (model 854 RF current meter.)

Little Tesla coils dont cause death generally; How they annoy others is a different matter.

Its now how you do it, its how you its

Dr. Slack

Fri Feb 26 2010, 11:34PM

Registered Member #72 Joined: Thu Feb 09 2006, 08:29AM
Location: UK St. Albans
Posts: 1659

TC grounding

Rule #1 - current doesn't doesn't flow from point to point, it flows in closed circuits
Rule #2 - just calling a conductor "ground" doesn't magically make it equi-potential 0v, especially at frequencies of several hundred kHz

Using the simplest model for a TC, an inductive secondary resonating with a capactive top load, we see that current flows in and out of the bottom coil connection, and displacement current is flowing in the dielectric (air) between the topload and all the conductive objects it can "see". Some of these objects have a conductive path to "ground" (the wiring in the walls, the pipework in the floor), some don't (you, the metal-framed desk in your room). Current flows via these isolated things because they are themselves capacitively coupled to more remote grounds.

Somehow, the current flowing in the bottom of the coil has to make some sort of a closed circuit with *all* the return paths for the topload displacement current.

Mains ground *can* be a very good choice to make the connection. If the power cord is short, if the ground at the socket is connected within a very short distance to "protective multiple earth" connections to your water pipes, electrical conduit, the building steel frame, a copper grounding spike, then the circulating RF currents are going to be very well contained locally.

On the other hand, mains ground can be very bad. If the socket ground wire goes wandering off through the walls a long way right back to your breakers, then the RF current flow in it will induce voltages into all your wiring. If you're are unlucky with configuration and you're in a duplex, then into your neighbours' too. You might kill their TV, you might mess with their reception and get an FCC visit.

What you need to do is to "be" the topload. Put your eye there, look around, and see what the displacement current is going to flow into. Then connect the all the electrodes that you can see (electric fields can see through insulators like plasterboard walls remember) by the shortest reasonable path back to the bottom of your coil.

The easiest situation is when you are operating the coil outdoors, on conductive ground. Wet the ground, bang a spike into the wet patch. Job done.

If you are outside on hard concrete, use a number of radial wires layed out away from the coil. These will capacitively couple to the ground beneath, the topload, and do a good job of returning the current. Wires are good, foil or chicken mesh is better.

If you are indoors, then use the fact that your walls and floor are full of mains wiring. Run a wire from the earth pin of a plug into at least one socket of every outlet plate in your room back to the coil bottom. Run wires to water pipes and gas pipes. If there's anything conductive in the room, steel shelves, desk, bench, put a wire onto each of them and run each back to the coil bottom. A number of radial wires in addition will improve things.

Do you worry about the lighting wiring? You could, but is it worth it? You'll probably get away with it just coupling to the rest of the house supply. If you do want to connect to it, then if it is run with 3-core grounded cable, run a wire to the ground of your light switch (hint, metal screws on such switches are usually grounded to the lighting ground, no need to increase risks by removing the switch). If it's only run in 2-core, then a mains rated (X or Y) capacitor to the not "hot" conductor for mains isolation would allow the RF to flow into the wiring. If you are sending RF all over the house and upsetting people and are not including the lighting wiring, then including it in the collective ground may be a good thing to do.

If you are in a workshop or garage with a steel-frame roof, then that's a certain connection. Drop a wire from each corners of it down to the coil bottom.

Steve Conner

Sat Feb 27 2010, 09:46AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

"You must be the topload" - I like it - a very Zen approach to coiling

I recently did a series of demos in a classroom at the university using my DRSSTC. I had to fill out a risk assessment for it, so I wanted to do it according to best practice.

We built a Faraday cage out of 8ft x 6ft wire mesh fencing panels as used on construction sites. The coil's RF ground was strapped to its mains ground, which went back along the power cord to an EMI filter installed at the boundary of the cage. The filter only filtered live and neutral: ground was solidly wired through its casing. The filter's casing was bonded to the cage with an earth strap, and this completed the circuit for displacement current. The whole setup was grounded through the power cable that ran from the wall outlet to the EMI filter.

The coil also had a 9-core screened cable for control. So as not to break the EMI barrier, the screen of this cable was also bonded to the casing of the EMI filter, however it was also grounded inside the coil. So to prevent RF current returning down it, 10 turns were taken through the biggest ferrite ring we could find.

All of this worked fine, even though the estates department only gave us 4 panels, so the Faraday cage had no floor or ceiling.

I used one of our prototype spectrum analyzers to look at the emissions, and while there were some, they were well below what we've seen with the same instrument in substations and switchyards... However I hope to get a better cage in future.

radiotech

Sat Feb 27 2010, 07:38PM

Registered Member #2463 Joined: Wed Nov 11 2009, 03:49AM
Location:
Posts: 1546

As to annoyance factor, in a classroom lab years ago involving scopes, picture rasters, lots of cables between benches issues arose that were coming from some devils in the room directly below with a TC. Their mirth an the possibe vengeance caused that coil to become invisible for a long time afterwards.

lhl_henrylui

Sun Feb 28 2010, 02:27PM

Registered Member #1498 Joined: Thu May 22 2008, 07:08AM
Location: Hong Kong
Posts: 72

Kipmans wrote ...

"RF ground" is just the name for the ground point you use for your TC. So that could either be a rod hammered into the earth, a copper strip attached to the base of the secondary or even the mains ground.

If a copper strip attached to the base of the secondary is already a suitable RF ground, why people should spend so much effort to hammer a rod deep into the earth to be the ground?
If a copper strip is already enough, I just have to connect the nearby metal objects (the metal bench, water pipes, etc., and the earthing rod on my hand for drawing arcs) to the copper strip, right?

radiotech

Sun Feb 28 2010, 04:38PM

Registered Member #2463 Joined: Wed Nov 11 2009, 03:49AM
Location:
Posts: 1546

The ground rod is to provide a low resistance path to the earth.
you can measure that resistance by driving 3 roda, A, B, C. and solving RAB, RBC,RAC, for R. Or you can buy an "earth megger" to do it for you.

Dr. Slack

Sun Feb 28 2010, 06:13PM

Registered Member #72 Joined: Thu Feb 09 2006, 08:29AM
Location: UK St. Albans
Posts: 1659

If a copper strip attached to the base of the secondary is already a suitable RF ground, why people should spend so much effort to hammer a rod deep into the earth to be the ground?

Because that's what you need for a mains safety ground, and most people don't understand the difference between that and a TC RF ground. The penalty for not understanding the difference is rarely a problem to the perpetrator, so the lesson never needs to be learned.

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob