Is there a way to ground a fixture without using a ground wire?

I remember reading on this site it was possible to connect the fixture ground to the neutral wire but put some capacitors in series with it to make it safe but I can't seem to find the post.

Is there a way to ground a fixture without using a ground wire?

I remember reading on this site it was possible to connect the fixture ground to the neutral wire but put some capacitors in series with it to make it safe but I can't seem to find the post.

Don't do that! Doing that will cause the fixture to be a death trap, because if someone decides to reverse Neutral and Hot, the fixture will be live and anybody/anything that touches it can/will get a full voltage AC shock. Ash can explain in full detail.

I already know that you're not supposed to. That's why I'm asking if there's a way to isolate the it using capacitors but still have the grounding effect.

Essentially, prevent any current from flowing if someone were to touch it, but allow the ballast to be "grounded" to aid in starting for RS ballasts on a hot/neutral system.

I already know that you're not supposed to. That's why I'm asking if there's a way to isolate the it using capacitors but still have the grounding effect.

Essentially, prevent any current from flowing if someone were to touch it, but allow the ballast to be "grounded" to aid in starting for RS ballasts on a hot/neutral system.

I ran a rapid-start ballast with no grounding wire (just hot and neutral) and it worked perfectly, for what I can remember.
The photo showing this is "It Works!".

If you happen to have a metal pipe, metal studs, etc. nearby, you could run a separate ground wire to that. Alternatively you can install a dedicated ground rod but that's probably not worth the trouble.

Why not add an Earth wire ?

Anyway, using a capaсitor may work and be safe, if you use an Y rated capacitor of small value, same as used in line filters between line voltage and enclosures of unearthed equipment. That is, if that capacitor is safe enough to be between line conductors and metal casing of unearthed DVD player, it is safe enough to be in the exact same position in a Fluorescent luminaire..

(Whats more, according to modern standards, not Earthed appliances must be double isolated, which the old unearthed Fluorescent luminaires arent anyway, so the Y capacitor is no less safe than the rest of the luminaire as it is)



Think what would happen if line voltage or ballast output voltage (up to OCV) does get to the casing though :

 - If the luminaire is not accessible, shock risk is low, though you still must know about it if you are about to touch the luminaire

 - If the luminaire is hanging on isolating supports, it will just hang there with voltage on it and nothing else happen. However, if voltage is applied to wood then over time it will carbonize, potentially decompose around the screws and let the luminaire fall

 - If the luminaire is hanging on metal building parts, or connected to them by a wire, then current can go from there to the building parts, and then continue through building structures, pipes and whatever else - Possibly creating a shock hazard somewhere else in the building, or a fire hazard (high currents up to 15+ A through metal building parts...) - Good reason not to Earth stuff to nearby pipes - If there is trouble in the device better have it localised and not wired to elsewhere

I guess I should have added to my previous post that if the light is grounded to the pipes or building structure, then whatever it's connected to should ALSO be adequately grounded somewhere else, preferably not too far from the light.

Still very bad practice

 - The "properly Grounded" building parts may become disconnected from each other, and then they pose a shock hazard. For example : You ground it to a metal pipe. At some point in the future another section of the pipe needs replacing, you unscrew some joint of the pipe and get a shock from the 2 sides of the pipe that you just separated

 - Eventually there is a fault between Neutral and the luminaire body, and eventually your proper Ground rods (wired to the panel or meter box) rot away. Then this "proper Ground" becomes effectively the only Ground of your house : It is connected from the Neutral bar in the panel, through the Neutral wire, fault in the luminaire, luminaire body, "properly Grounded" building parts and Ground. In this path consisting of structural steel and maybe #14 Copper wire, go all the stray currents between power company's Neutral and your Ground, which may be quite ligh. If the Neutral fails, then all your Neutral currents that can be 10's Amps. Good potential for a fire, and there isnt any single breaker along this path to interrupt it cause it is all Neutral !

If you connect the lamp body to neutral with a capacitor then there won't be enough current flow to aid tube starting also it might not be very reliable better to run a separate ground wire back to your main fusebox

The tube starting is provided by capacitive coupling between the lamp and the reflector. This capacitance is on the order of 10pF. The Y capacitor used is 1000pF. So it is the capacitance between the lamp and reflector, not the Y capacitor, that limits the current flow for starting



The Y capacitor does serve an important role though :

For starting, we want max voltage between the "Phase" end of the lamp to the reflector. So the "Phase" end of the lamp is at potential on the order of ballast OCV (not necessarily equal to OCV, that depends on ballast design), and the reflector at Earth potential

There are allways parasitic capacitances, and in gear with old isolation materials also resistive leakages, between all the internal wiring and ballast coils to the metal enclosure. This capacitance will tend to bring the enclosure to some potential around half of the line voltage (maybe more, maybe less, as the capacitances are not equal, various other voltages are present on some components, and resistive leakages are a random thing altogether). But that means, that the potential difference between the "Phase" end of the lamp and the reflector is reduced, and so is the starting capability

The Y capacitor is there to be greater than all the parasitic capacitances, so that it pulls the enclosure potential to that of the Neutral more "firmly" than the other stray capacitances pull it in other directions, so that overall its potential will be closer to Neutral/Earth

Some of my older church buildings don't have grounding in the UN-rennovated areas.  In this case, I usually ground to the metal conduit or something else nearby that will suffice. 

Some of my older church buildings don't have grounding in the UN-rennovated areas.  In this case, I usually ground to the metal conduit or something else nearby that will suffice. 

That may be working, but I would see that as quite dangerous: If the fixture develops a short circuit or an excessive leakage to the schassis, you bring the line voltage onto that conduit and there it could shock someone quite far away from the damaged fixture.

In that case I would prefer not direct connection, but via some high impedance, preferably Y rated capacitors (about 2.2nF) or bleeder resistors (3MOhm total, if using the ordinary parts, assemble that from a series connection of at least 5 pieces).  These provide strong enough current path for the function of the shield as an external starting electrode, but limit the current to a safe level when a fault develops.

But then why not use simply the local Neutral ? Thats safer as it would keep any problem local and not send current to elsewhere accessible, also there isnt potential issue of current from several luminaires summing up as they fail over time

(as long as the luminaires are not touching through metal chain etc to the building anyway)

Local Neutral maybe, but still via the safety capacitor and/or resistor. Many ballasts do feature such bleeder from the factory, then you need to just connect the ballast case (or the "grounding" wire from the ballast) to the reflector...

Definitely through the Y cap. But i never seen Magnetic ballasts with a bleeder to the case