I have a neon light at work (a display hanging in a window). So I was doing some work next to it and felt a little shock on my neck. I went and got my multimeter and grounded it on what my hand was on at the time and put the other end on the metal chassis that the tube is mounted on. Sure enough, it read 25V. Which means that there is some sort of a short to ground. It’s not a “full short” (that is, the wires arn’t touching the metal) since it’s running off a transformer (mounted remotly) with a 9000V secondary, and I’m still walking. This light is three colors and has several different sections connected by wires (and heavy boots over the connections). I’ve checked all the areas where bare wire could be exposed (near the connections) and there doesn’t seem to be anything obviosly wrong. Is there some way that the voltage going thru the tubes could be leaking out to the metal chassis? Has anyone ever seen this happen? Any ideas??
It sounds like something is leaking current to the chassis, probably the high voltage transformer.
Is the chassis tied to earth ground? Unplug it and do a continuity check w/ the power cable’s third pin.
BTW: The voltage is probably higher than 25 V, since the source impedance and voltmeter’s input impedance form a voltage divider.
It might not be a short to ground but an induced voltage caused by current flow next to the object in question, i.e. there might be some small eddy currents in the metal chassis because it’s surrounded by the magnetic field generated by the high voltage.
If the chassis isn’t grounded you might try grounding it.
b.
“Voltage” in a circuit spreads all through the space around the circuit, although we rarely deal with this effect. It’s otherwise known as “capacitive coupling.” If you place any metal object in the voltage-field surrounding a high voltage circuit, you’ll measure a voltage appearing on that metal object. The cure: connect all adjacent metal parts to ground. If there is a metal bracket as part of your neon sign, there should be a wire connected to the bracket which eventually leads to the 3rd prong of your electrical outlet. If not, you’ve got yourself a potential electrocution hazard.
On the other hand, if YOU are immersing yourself in the voltage field of the neon sign, then you can receive tiny shocks whenever you lightly touch a grounded metal object, or touch a fairly large metal object which isn’t grounded. This isn’t dangerous except possibly to people wearing medical implants. The AC voltage might be high, yet the currents involved are down in the microampere range. It’s like walking near a VandeGraaff machine: everything you touch will give you tiny sparks.
Yes, there were tiny sparks when I moved the multimeter probe along the chasis. And as I was writing the OP I realized that the unit is NOT grounded. The only thing that connects the transformer to the tube is the two lead wires and since the transformer is mounted away from the tube it’s not grounded by direct contact. So I suppose I could just run a wire from the chassis to the transformer and thus it would all be grounded. One thing though…Is is safe to constantly have 25V leaking through the ground wire and through all the conduit in the building until it finds the earth?
You won’t have 25V leaking to ground. The whole purpose of grounding things is to make them the same potential as the ground. You will have a terribly small amount of current however through that ground path. I would guess about 1000 times as much current flows to ground from your computer power supply when from this tube. UL, or CE mark if you are in Europe, allows 3.5mA in the ground conductor.
Really, there is no danger. If it bothers you, move the light. I can’t recommend making modifications to a lamp that is already safe.
Interesting thing about those high voltage lamps, the corona(sparks) around the tube can generate ozone, which we all know is bad for you. How much ozone, I don’t know.
You won’t have 25V leaking to ground. The whole purpose of grounding things is to make them the same potential as the ground. You will have a terribly small amount of current however through that ground path. I would guess about 1000 times as much current flows to ground from your computer power supply than from this tube. UL, or CE mark if you are in Europe, allows 3.5mA in the ground conductor.
Really, there is no danger. If it bothers you, move the light. I can’t recommend making modifications to a lamp that is already safe.
Interesting thing about those high voltage lamps, the corona(sparks) around the tube can generate ozone, which we all know is bad for you. How much ozone, I don’t know.
Sorry about the double post. I got impatient.
Sure are a lot of billys in this thread. Have we passed the “billy” limit? is there one? should there be? is it strange that a bunch of people named “bill” or “Billy” would respond to a question about Neon? does this have some special meaning? Is it all part of the secret Cabal?
Naaaah.
b.
The 25 volts is probably wrong.
Normal DVMs have a 10-megohm input impedance. When measuring the capacitive coupling voltages on small metal parts, 10 megs acts like a very good conductor (like almost a direct short.) The true voltage on the metal is probably a few thousand volts. If you’re seeing sparks, then it MUST be higher than 700 volts.
In other words, the meter’s 10M resistance is way too small in this instance, and it’s dragging down the signal.
The voltage on the metal parts is being communicated from the plasma within the glass tubes. The voltage-field reaches out and lands on the metal. The currents involved are extremely tiny; probably way less than microamperes: it’s all voltage, no current. To properly measure such things you need n “quadrant lectrometer” or “electrostatic voltmeter” having impedance not in the megohm range, but instead in the thousands of gigohm range. (I see these devices on ebay occasionally. They also can accurately measure your body voltage when you scuff across a rug in the winter!)
See:
Measuring high voltages at zero current
http://www.amasci.com/emotor/voltmeas.html
Definitely ground the chassis:
- Make sure the receptacle is the “grounded” type.
- Make sure the receptacle’s ground socket is properly connected to earth ground via the breaker/fuse panel.
- Is the existing power cord a 2-conductor (hot/neutral) or 3-conductor (hot/neutral/ground) type? If it’s the former, you’ll need to either: a) install a 3-conductor power cord, or b) install a 3-conductor plug and run the ground wire separately.
While you’re at it, why not also install a GFCI outlet? They’re only $8. And they trip when the sense a ground current between 4 mA and 6 mA.