Author Topic: Connecting Fluorescent ballasts on the Phase side  (Read 2500 times)
Ash
Member
*****
Offline

View Posts
View Gallery


Connecting Fluorescent ballasts on the Phase side « on: September 02, 2016, 03:01:14 AM » Author: Ash
In all Fluorescent luminaires here (switch start 36W) the ballast is connected on the Phase side of the supply

When starter is shorted, both electrodes are at Neutral potential. If they were at Phase potential instead, could this aid in creating ionization while the starter is still closed ? (assume 36W Kr filled lamp, with the FL luminaire metal being at socket height's distance from the lamp ~15mm)

If the ballast fails with short to Earth :

 - With the ballast on Phase, it can burn for long time and reach extreme temperatures (especially if the initial short is followed by winding to winding short, as result of heat from the arc in place of initial fault), possibly way above its rated dTan (stuck starter temp rise above ambient)

 - With the ballast on Neutral, it will blow the lamp cathodes and starter in short order and stop

Looks like it would make more sense to place the ballast on the Neutral. Then why most of the world (except New Zealand) place it on the Phase ?
Logged
Medved
Member
*****
Offline

Gender: Male
View Posts
View Gallery

Re: Connecting Fluorescent ballasts on the Phase side « Reply #1 on: September 02, 2016, 10:01:46 AM » Author: Medved

When starter is shorted, both electrodes are at Neutral potential. If they were at Phase potential instead, could this aid in creating ionization while the starter is still closed ? (assume 36W Kr filled lamp, with the FL luminaire metal being at socket height's distance from the lamp ~15mm)

With an inductive ballast this aid is not needed, so the safety/cost makes the ballast in phase the preferred way.
However with resistive ballasts this extra ionization was indeed utilized to aid the starting (however for Ar T12 lamps; for Kr T8's it is not enough)


- With the ballast on Neutral, it will blow the lamp cathodes and starter in short order and stop

This blow tends to end up in lamp explosion and so release of the mercury and sharp glass around. And that can be hardly treated by a common fixture designs (even with vapor tight fixture it will get released once someone opens the fixture).
On the other hand the ballast overheating could be quite easily contained by just mounting the ballast on a metal base and so separate it from the heat sensitive materials. So no over expensive sealing or so. Plus once you are using an RCD, it shuts down the power, still without any excessive heat.
 Because of that, it is considered safer (and/or cheaper) to have the ballast on the Phase side.

Logged

No more selfballasted c***

Ash
Member
*****
Offline

View Posts
View Gallery


Re: Connecting Fluorescent ballasts on the Phase side « Reply #2 on: September 02, 2016, 01:53:22 PM » Author: Ash
I have hard time to imagine the tube exploding, as its internal atmosphere is near vacuum, and there is no arc along the center of the lamp - Any gases that expand from the arcs near the electrodes would go to the cold center of the lamp and thats it. Cant imagine how any significant pressure build up inside would happen

(the starter is near vacuum too i think, but there i can see how the glow lamp can explode, as the arc fills the entire volume of the lamp)

Why would the FL explode then ?

And if it does, i doubt the 2mg of Mercury in there is any worse to health than the fumes of burning up resin in the ballast (and that is some strong smell, allthough i did the experiment where the ballast burned up outdoors) - So only the Glass concern really remains



Most FL luminaires dont have the chokes mounted to any stand off plate, just to the back wall of the luminaire enclosure



There are some places where there is no requirement for RCD in lighting circuits, when the circuit is only for lighting and "inaccessible enough". The design of luminaire (especially one that could be installed on such circuit) can not assume that there is RCD in the circuit



What are the requirements for Phase/Neutral polarity in Europe ? The switch would be on the Phase, but is there any requirement to connect "properly" the luminaires to the circuit ?
Logged
Medved
Member
*****
Offline

Gender: Male
View Posts
View Gallery

Re: Connecting Fluorescent ballasts on the Phase side « Reply #3 on: September 02, 2016, 03:11:00 PM » Author: Medved
I have hard time to imagine the tube exploding, as its internal atmosphere is near vacuum, and there is no arc along the center of the lamp - Any gases that expand from the arcs near the electrodes would go to the cold center of the lamp and thats it. Cant imagine how any significant pressure build up inside would happen

(the starter is near vacuum too i think, but there i can see how the glow lamp can explode, as the arc fills the entire volume of the lamp)

Why would the FL explode then ?

I didn't mean a real explosion per se, but rather the glass around the tube end smashed from a thermal shock from the arc between the lead wires (that is the mechanism I assume will happen; I have made that experience long time ago, when I didn't know what the choke is for in a fluorescent circuit - due to some reason I assumed it is just an RF filter, so not needed for the working; Obviously I was wrong, because the tube ends disintegrated and flied out; I was about 11 years old at that time).
The consequence of glass pieces flying down into the fixture canopy is then the same...


And if it does, i doubt the 2mg of Mercury in there is any worse to health than the fumes of burning up resin in the ballast (and that is some strong smell, allthough i did the experiment where the ballast burned up outdoors) - So only the Glass concern really remains

Most FL luminaires dont have the chokes mounted to any stand off plate, just to the back wall of the luminaire enclosure

It depends, if the fixture is rated to be mounted on a flammable surface (marked with an "F" in a kind of triangle), then there must be something to limit the contact surface temperature and the radiated heat. Usually that is in the form of the fixture body sheet profile shaping (the mechanical integrity needs the metal to be shaped as well, so it serves double purpose). This is usually enough to keep the flamable surface cold enough even when the choke "goes away". Or some (mainly the plastic body) fixtures employ a thermal cut off fuse tied to the ballast. So when it exceeds the programmed temperature, the fuse disconnects the supply. This became quite common with the modern, plastic body fixtures...

If the fixture is not "F" rated, it is supposed to be mounted on something, that can withstand the high heat even from the burning ballast.

Many designs just keep the ballast body insulated (it is OK, if it is covered even when the user is replacing lamps or so), as well as wire the sockets according to Class-II rules, so the short to Ground could be practically excluded (that fixture uses grounding connection just for shielding the eventual RF emissions from the lamp).

And normally even when directly across the mains, normally the ballast should withstand that for some time (the users are supposed to not operate visibly damaged fixtures, nor keep them unsupervised; if that is needed, reghular inspections are mandatory and these should "catch" the nonworking fixture in time before the ballast really fail)

There are some places where there is no requirement for RCD in lighting circuits, when the circuit is only for lighting and "inaccessible enough". The design of luminaire (especially one that could be installed on such circuit) can not assume that there is RCD in the circuit

Yes, that is true, RCD's are mandatory only for sockets accessible by people without any electrical training (so practically all cases except dedicated lab desk or some special industrial installations). So lighting circuit may be without.

What are the requirements for Phase/Neutral polarity in Europe ? The switch would be on the Phase, but is there any requirement to connect "properly" the luminaires to the circuit ?

It should follow makers instruction. And it is the maker's task to ensure, those instructions lead to a safe installation.
So no strict rules per se. The thing is, the European Code does not distinguish between Phase and Neutral - they are both just "live working load bearing" conductors, so they should be treated as the dangerous voltage is present on both of them.
Of course, an exception is the PEN in a TN-C installation, where the the PE and N functionalities share a common wire. There the PE functionality dictates the PEN to be the grounded wire. But it has yellow/green color, so from that perspective it is distinguished from any live working conductor (any other color)
Logged

No more selfballasted c***

Ash
Member
*****
Offline

View Posts
View Gallery


Re: Connecting Fluorescent ballasts on the Phase side « Reply #4 on: September 02, 2016, 04:01:54 PM » Author: Ash
Here are some of our luminaires made to be installed on flammable surfaces, and marked with the F mark :

Gaash 5054 with the chokes on "bumps" in the metal :
http://www.lighting-gallery.net/gallery/displayimage.php?album=2158&pos=16&pid=108745

Gaash 1600/2800 with the chokes flat on the back wall of the luminaire :
http://www.lighting-gallery.net/gallery/displayimage.php?album=2158&pos=29&pid=108530

Shaltiel MS A2 with the chokes flat on the back wall of the luminaire :
(The A2 have another design flaw, where the tab to hich the capacitor is mounted is folded out of the back wall, so that the capacitor is exactly under the hole, and those capacitors do fail nasty sometimes. In this pic its luminaires where i moved the capacitors around to correct this problem)
http://www.lighting-gallery.net/gallery/displayimage.php?album=2158&pos=50&pid=107592

Optima MX-208-Y362 with the chokes flat on the back wall of the luminaire :
http://www.lighting-gallery.net/gallery/displayimage.php?album=2158&pos=9&pid=109262



Actually the one i have where the choke was burned up with Earth fault is a 5054 as well :
http://www.lighting-gallery.net/gallery/displayimage.php?album=2158&pos=5&pid=110709

This one was result of long time with stuck starters. The fault was a short from the winding (from the outside layer of the coil, but not at the very end of the winding) to Earth

But i guess there allways is some chance of a fault like this, even as sudden falt when all worked fine before - For example if some insect gets between the coil's outer layer and the core, then its the starting pulse voltage (1KV) across just the enamel wire isolation. Or if there is a winding to winding short as result of manufacture defect, and then it escalates to winding to core short when the Plastic bobbin wall melts near that place



Now if we ensure that the outer end of coil, that is likely to short to Earth is at the supply side and the inner end (that is least likely) at the lamp side, then even if the ballast shorts to Earth, there still is ~30 Ohm worth of wire resistance in series with the lamp in the worst case, even if the ballast inductance effectively drops to 0 (and does it really from 1 shorted turn ?)

Then with such "softer" short circuit, would this allready be sufficient to prevent the lamp exploding ?
Logged
Medved
Member
*****
Offline

Gender: Male
View Posts
View Gallery

Re: Connecting Fluorescent ballasts on the Phase side « Reply #5 on: September 02, 2016, 05:52:14 PM » Author: Medved
The single shorted turn means mainly high losses there, so making the enamel to degrade all around, till it runaway to the hot failure mode even when the rest of the circuit is intact (no short to ground,...). So the fixtures have to be robust against such failure anyway, or if it is not robust enough, it will pose the fire risk anyway.
Given how the winding is usually arranged (with extra paper between the wire and core), I see the short to ground way less likely than a short or leakage between the turns inside of the coil.
If that is not considered as a safety problem (I didn't do any exact tests or so, so I do not know how hot that could be before it triggers the breaker, nor exact details about how high temperatures are allowed for an "F" rated device, so I really can not judge an exact design), it really won't make any difference whether the ballast is on the Phase or Neutral side.

Even when the ballast on the Phase side is the most common here, I've seen few two lamp fixtures, where each lamp had the ballast on the other circuit end, so one at Phase, second at Neutral. It looked like made so to just save on the inner wire length (the mains connection was in the center, one ballast on one side, other on the opposite, phase runs towards one side and connects to both the ballast of first lamp and lamp pin of the second lamp, the Neutral runs to the other side, there connects to the second lamp ballast and the first lamp pin). Even with such nearly symmetrical arrangement (well, on one side was a series capacitor between the ballast and lamp; it was a lead-lag circuit) the connection terminals were marked as "L", "N" and obviously "PE"...
Logged

No more selfballasted c***

Ash
Member
*****
Offline

View Posts
View Gallery


Re: Connecting Fluorescent ballasts on the Phase side « Reply #6 on: September 02, 2016, 07:38:46 PM » Author: Ash
In the shorted turn case the lamp is still in series with the balast. so i'd expect it to fail quite fast and open the circuit. Here is a video of 36W T8 on a ballast with shorted turn :

At 13:42 : https://www.youtube.com/watch?v=mi9u7kw27ys

When the short circuit path is Phase - Ballast (full intact coil) - PE, the ballast would heat about same as with stuck starter, so no immediate danger, but it will be damaged over time

When the short circuit path is Phase - Ballast (part of coil) - PE the ballast would heat more, actually limited only by how fast it gets to the point of shorting out completely

With experiment i did with a bad ballast like that (the burned ballast from the luminaire in the link above) and 3KW space heater (to prevent short circuit when it fails), the ballast held for long time while heating and stinking, but did not short out - Current through it stayed on the order of 2A (compared to 600..800mA stuck starter on good ballast)



The extra paper crumbles after some time with stuck starter heat level. It may happen, that a luminaire styed long time with stuck starter and the ballast is on the edge of failure, but then somebody put in a new lamp and starter and it works fine. In this case the ballast will be likely to fail at some unknown time, and quite likely as fault to PE because the paper is not really there anymore

I have one luminaire just like you describe :
http://www.lighting-gallery.net/gallery/displayimage.php?album=2158&pos=84&pid=63982
Logged
Medved
Member
*****
Offline

Gender: Male
View Posts
View Gallery

Re: Connecting Fluorescent ballasts on the Phase side « Reply #7 on: September 04, 2016, 02:35:31 AM » Author: Medved
If the paper get "crispy", it keep there a void. And that void is not conductive. Well, it's breakdownb voltage would not be excellent, but still for the 330V peak or so present there during normal operation it should hold.

The ballast degradation takes time even with shorted starter, so the usually required monthly inspections (looking for nonworking lamps and mainly for stuck starters) should be sufficient.

But anyway, is there is stuck starter and it is so long it really degrades so much, the the fire may start just from the ballast overheat itself, regardless of which side is connected to Neutral directly and which via the stuck starter.

Logged

No more selfballasted c***

Ash
Member
*****
Offline

View Posts
View Gallery


Re: Connecting Fluorescent ballasts on the Phase side « Reply #8 on: September 04, 2016, 04:04:59 AM » Author: Ash
When the stuck starter is still present, the voltage across the paper is very low (voltage across 2 cathode filaments + few turns of the choke). When a filament of the lamp burns out its suddenly 230V, or if lamp is replaced its up to 1KV starting pulses. In either case the isolation failure may happen not immediately, but at unknown time later

Monthly inspections are not allways done. It is possible that a ballast was standing in that condition for long time before the lamp was replaced, and possibly so it would take only few days more till a fault to PE if the bad lamp would remain there. Most electricians dont open the cover to visually see the condition of the ballast when replacing a lamp with stuck starter

As long as there is lamp in series with the failing ballast, there is no fire risk - When the ballast fails (whether by short between turns or short to PE) this will overload and blow the lamp cathodes, well before it heats to extreme temperatures. If the ballast is on Phase and shorts to PE, then there is no lamp cathode (fuse) in the circuit, so nothing to limit its heating - Its current may be way higher than stuck starter current and not much there to limit the heating. In my experiments (ballast with PE fault in series with 3KW heater) the current there was on the order of 2A, while normal stuck starter is no more than 800mA. The ballast ofcourse got very hot
Logged
hannahs lights
Member
*****
Offline

Gender: Female
View Posts
View Gallery

Re: Connecting Fluorescent ballasts on the Phase side « Reply #9 on: September 04, 2016, 04:43:58 PM » Author: hannahs lights
We had 2 light fittings fail with burnt choke. in work I'm not sure if the tubes failed and it killed the choke or the other way round but the fact of it is is that they really stank badly I think if I could smell it in a stinky factory then anyone would soon notice a problem in a more normal environment. The circuit for those lamps where protected by a 7.5 amp trip switch so around 1500 watts could of been passing through the shorted choke in theory anyway
Logged
Medved
Member
*****
Offline

Gender: Male
View Posts
View Gallery

Re: Connecting Fluorescent ballasts on the Phase side « Reply #10 on: September 04, 2016, 06:20:16 PM » Author: Medved
For sure there are some bits favoring one connection, some others favoring the other one.
In many cases the ballasts tend to go out with smoke on their own (or as a result of a longer time overvoltage or so), failing into smoke mostly without causing ground fault at all. At the same time there are many fixtures with stuck starters glowing for years and yet the ballasts apparently survive that.

I don't think one connection would be any more nor less safer than the other, only when it happens to fail in some nasty way, people (and mainly the ones like insurance agents or lawyers) start to ask, if the unusual connection didn't played a role in that particular case. If that happens and the connection you used does not correspond to what most others are using (that is the ballast on the phase side), you will have hard time explaining why you considered your connection safer - they will judge only about that particular case...
If the connection corresponds to what is commonly used, no one will ask whether there were anything better or so, just because you have followed "the industry standard", so you will be fine from legal consequences...
Logged

No more selfballasted c***

Ash
Member
*****
Offline

View Posts
View Gallery


Re: Connecting Fluorescent ballasts on the Phase side « Reply #11 on: September 05, 2016, 12:12:07 AM » Author: Ash
If the luminaire was tested by a standards lab and found eligible to be marked with the standards approval sign, can the lawyers have any claims to the design ?
Logged
Medved
Member
*****
Offline

Gender: Male
View Posts
View Gallery

Re: Connecting Fluorescent ballasts on the Phase side « Reply #12 on: September 05, 2016, 02:48:43 AM » Author: Medved
If the luminaire was tested by a standards lab and found eligible to be marked with the standards approval sign, can the lawyers have any claims to the design ?

Yes, of course. It is your responsibility to make the design safe. The standard lab testing does not lift the responsibility.
If you ultimately succeed in making the design safe, nothing can happen (but that is just a hypothetical case).
For that you have to gather all knowledge and eliminate all risks you can reasonably identify and address.

Very significant part of the "knowledge" is, how similar things are normally done - it is usually the result of decades experience in the field, which was as well build on many disasters that have happened before. If you think some are not relevant for you anymore (because you think your approach have eliminated the risks in another way), fine, but you have to have strong backing in hands (safety analysis, FMEA's,...; I may tell you most dreaded things among engineers in safety critical businesses like automotive, aero,..., because their formalism tend to be the target for auditors, who do not know a s**t about what you are doing).
So if you do not want to spent your whole career on improving an electrical circuit for a desk lamp, better just stick to how the things are usually done, you could be sure the crucial mistakes were already filtered out by the test of time...
And on top of a sound design practice, you have to perform validation measurements and tests, which should prove you haven't overlooked anything.

Now for the design practice, it could be checked by anyone any time. But the validation tests are supposed to cover all things that are not that visible, mainly the overall performance in some specified critical situations.
There you rely on those being carried out properly. And even when the test procedures are prone to man made errors and misunderstandings, the codes require these to be performed by certified labs, where part of the certification process is staff training, usage of proper equipment and its calibration. You may have such lab as part of your company, but I may ensure you, is a heck of expensive business. So only some large companies use that way, all smaller just subcontract these tests to some certified lab.
Of course, these test labs do a bit more service than just plain measurement (just to save your money - they inspect the thing as well and if found something weird, they will notify you and ask if you really insist on spending money on testing that design, when they think it is likely to fail even before they start). And of course, if you insist, they mention the findings in the report, together with the test results - and it is up to you, what you do with that...
Logged

No more selfballasted c***

Ash
Member
*****
Offline

View Posts
View Gallery


Re: Connecting Fluorescent ballasts on the Phase side « Reply #13 on: September 05, 2016, 04:26:18 AM » Author: Ash
In this specific case there is a country where it is the normal practice to connect the ballast on the Neutral on purpose. That is New Zealand. They have been doing that for decades, and apparently nothing bad happens. Additionally, what about a Class I (with PE) Fluorescent luminaire with plug ? There is no way to ensure which way the plug goes in in 2 of the most common socket types in Europe. Yet those luminaires (magnifier desk lamp with circline, ...) are supposed to remain safe anyway

And what would happen in the "safety theater" if a FL luminaire made the classic way, choke on Phase, fails in nasty way (no lamp in the circuit, 2A through partially shorted coil), and for example, heats the luminaire enclosure to the extent that its outside temperature get so hot that it badly melts some nearby drop ceiling material, and claims are made that it could have started a fire

How is risk management done generally when developing new stuff ? (othen than by trying not to overlook allready well known things from previous experience that apply)
Logged
Medved
Member
*****
Offline

Gender: Male
View Posts
View Gallery

Re: Connecting Fluorescent ballasts on the Phase side « Reply #14 on: September 05, 2016, 07:45:37 AM » Author: Medved
If it is an isolated event, it will fall among the "remaining issues that can not be prevented and that is understood as rare enough". In fact it is nothing else than that - there will always be some failure modes associated with each approach, I don't think either of the connections is fundamentally worse nor better.
If that becomes a more frequent issue, detailed analysis will start among the industry and only then the result will make it into a "recommended practice". Or not, because the outcome may well be all the alternatives are the same or worse...

And there is another aspect in doing it how it is common in that particular region: Once something fails there and it needs to be fixed, once the electrician has to deal with something unusual, it is way more likely he will make some mistake (because he may misunderstand some reasoning and do the thing incorrectly for the used arrangement). And it will be just because of this kind of mistakes, which may make the unusual wiring less safe (= more frequent dangerous failures).
Logged

No more selfballasted c***

Print 
© 2005-2022 Lighting-Gallery.net | SMF 2.0.18 | SMF © 2021, Simple Machines | Terms and Policies