A Light That Sometimes Trips Breakers?

[tommygdawg]

Hey all,

Quick question. I have a ballasted 1200 watt discharge lamp. When plugged into a room with nothing else on the circuit, it will sometimes trip my 15 amp house breaker when I strike the lamp. Other times, it strikes just fine. My question is, could this be a result of faulty electronics in the ballast? It seems that if it doesn't trip breaker on one ignition, it shouldn't on any others.

As a side note, I have another similar 1200 watt discharge light with a different type of ballast that never trips my breaker when I strike it.

Any thoughts would be appreciated! Thank you!

[alsetalokin4017]

Slightly scary...

First thing I'd do is put a 20-amp breaker in place of the 15-amp you are using now. Then if the light ever even once trips the breaker again--- get rid of the ballast and replace it with a new one. Destroy the defective ballast so nobody else might try to use it.

[tszaboo]

Which country? 110 or proper voltage?

[EPTech]

Hi There,

Modern circuit breakers work twofold. One function is to prevent overloading. It is a slow function that involves an element in the breaker heating up a bi-metal that trips the breaker in a well defined time at a certain overload current. The second function is short circuit prevention. In this case the circuit breaker is actuated magnetically and that will trip it almost instantaneous.

I think that last function is bothering you. Upon striking the stike coil of your lamp will draw a very large amount of current in a short time. The breaker "sees" this as a short circuit and it will trip. You can choose to place a slower breaker (different class) but still with the same overload current (15A). You need to check local regulations because it may not be allowed.

Also, your breaker may also be worn. If you get a lot of those peaks on a circuit, the magnetic actuator may actually wear out the trip mechanism of the fuse.

[HackedFridgeMagnet]

First thing I'd do is put a 20-amp breaker in place of the 15-amp you are using now.

You so cannot do that.
The breakers are rated to protect the wire not the load.

get rid of the ballast and replace it with a new one.

Yeah it seems the ballast is not limiting the current.

[alsetalokin4017]

First thing I'd do is put a 20-amp breaker in place of the 15-amp you are using now.

You so cannot do that.
The breakers are rated to protect the wire not the load.

 
You might not be able to do that, but I can.     

get rid of the ballast and replace it with a new one.

Yeah it seems the ballast is not limiting the current.

[tommygdawg]

Thanks for the thoughts, everyone! To clarify, I am on 120 volts here. Usually my voltage at any of my outlets varies between 118-122. I do have 20 amp circuits in the house (bathrooms, kitchen, etc.). Perhaps I'll try running it off of those a few times and see what happens.

Hi There,

Modern circuit breakers work twofold. One function is to prevent overloading. It is a slow function that involves an element in the breaker heating up a bi-metal that trips the breaker in a well defined time at a certain overload current. The second function is short circuit prevention. In this case the circuit breaker is actuated magnetically and that will trip it almost instantaneous.

I think that last function is bothering you. Upon striking the stike coil of your lamp will draw a very large amount of current in a short time. The breaker "sees" this as a short circuit and it will trip. You can choose to place a slower breaker (different class) but still with the same overload current (15A). You need to check local regulations because it may not be allowed.

Also, your breaker may also be worn. If you get a lot of those peaks on a circuit, the magnetic actuator may actually wear out the trip mechanism of the fuse.

Thanks for the detailed thoughts! Would your second explanation account for why sometimes the breaker trips and other times it doesn't? I will say that it's happened on a couple different circuits in the house.

[bobcat]

If it has a magnetic ballast, the problem may be residual magnetic field present and power phase of opposite polarity (luck of timing). This causes a high current draw while collapsing the residual magnetic field.
You can try a soft start for the power or maybe even use a solid state relay with zero crossing turn on.

[tommygdawg]

If it has a magnetic ballast, the problem may be residual magnetic field present and power phase of opposite polarity (luck of timing). This causes a high current draw while collapsing the residual magnetic field.
You can try a soft start for the power or maybe even use a solid state relay with zero crossing turn on.

It's an electronic ballast so that doesn't apply, unfortunately. My kill-a-watt reports a running amperage of 14amps (give or take) at full intensity, so I'm not surprised that it would sometimes trip a breaker. I just wish I knew why it doesn't *always* trip the breaker. Thanks for the thoughts, though!

[Gyro]

If it has a magnetic ballast, the problem may be residual magnetic field present and power phase of opposite polarity (luck of timing). This causes a high current draw while collapsing the residual magnetic field.
You can try a soft start for the power or maybe even use a solid state relay with zero crossing turn on.

Actually it's the other way around. Absence of residual flux from the previous half cycle can cause the ballast (or other cored inductor / transformer) to saturate on the first half cycle if turned on at the zero crossing, causing a high peak current. Zero cross switching will cause this to occur at every switch-on. Inductive circuits are best turned on near mains peak voltage.

[tommygdawg]

If it has a magnetic ballast, the problem may be residual magnetic field present and power phase of opposite polarity (luck of timing). This causes a high current draw while collapsing the residual magnetic field.
You can try a soft start for the power or maybe even use a solid state relay with zero crossing turn on.

Actually it's the other way around. Absence of residual flux from the previous half cycle can cause the ballast (or other cored inductor / transformer) to saturate on the first half cycle if turned on at the zero crossing, causing a high peak current. Zero cross switching will cause this to occur at every switch-on. Inductive circuits are best turned on near mains peak voltage.

Could this be the case even after several hours or more of inactivity?

[Gyro]

It's not a case of residual magenisation or the core in the 'permanent magnet' sense. Ballasts / choke cores are sized for steady state operation, when the current voltage rises on every half cycle it relies on the collapsing flux from the previous half cycle to avoid saturation. If switched on from a zero crossing, there is no collapsing residual flux to oppose and the core hits saturation. The reason the breaker is tripping sometimes is due to the luck of the draw in the point of the cycle when it is switched.

The simplest solution would be to add a low value resistive ballast (an ohm or two) to limit this saturation peak to a point where the breaker doesn't trip.

EDIT: It is possible for the laminated core material to degrade with time and temperature to the point that saturation occurs more easily, it's a known problem in transformers salvaged from old tube gear. In that case a new ballast is the solution.

[tommygdawg]

It's not a case of residual magenisation or the core in the 'permanent magnet' sense. Ballasts / choke cores are sized for steady state operation, when the current voltage rises on every half cycle it relies on the collapsing flux from the previous half cycle to avoid saturation. If switched on from a zero crossing, there is no collapsing residual flux to oppose and the core hits saturation. The reason the breaker is tripping sometimes is due to the luck of the draw in the point of the cycle when it is switched.

The simplest solution would be to add a low value resistive ballast (an ohm or two) to limit this saturation peak to a point where the breaker doesn't trip.

EDIT: It is possible for the laminated core material to degrade with time and temperature to the point that saturation occurs more easily, it's a known problem in transformers salvaged from old tube gear. In that case a new ballast is the solution.

Well, it is a new ballast. Albeit a new Chinese ballast. I'm not sure I completely understand your explanation when it comes to things like zero crossing and draw at certain cycles. Would you be willing to elaborate more?

Also regarding adding a low value resistive ballast, is that an external device you're referring to? Or literally inserting a resistor into the internals of my Chinese ballast?

[Gyro]

Sorry, I probably didn't explain very well, and it's not an easy one to explain....

With an inductor, current always lags behind voltage, ideally by about 90 degrees. That means, in steady operation, as the voltage rises after a zero crossing the current and magnetic flux are still falling from the previous half cycle of the opposite polarity. If you switch on from a zero crossing, the flux is already at zero, allowing it to rise further in the first half cycle.. to the point of saturation in a cheap, tired or undersized core. At the point of saturation there is no longer any reactance to limit the current.

The resistor I was suggesting would be a wirewound one for surge capability. It needs to have a wattage rating sufficient to withstand the voltage drop at normal lamp current (with a safety margin). Actually an NTC thermistor designed for surge limiting (as used in switch mode power supplies would probably a better and more efficient solution. I'm not sure about mounting, I don't know your mechanical setup - obviously it would need to be electrically safe and somewhere that would withstand its heat dissipation. Sorry, that sounds a bit vague.

[timb]

Might be a bad breaker. They do occasionally fail and randomly tripping is a symptom. Can you try it on another circuit or is the ballast permanently wired?

[_Andrew_]

Inrush current for discharge lighting can often throw MCB's as the current at switch on will be in orders of magnitude over the nominal operating current for a few milliseconds.

I use 400w metal halide lighting with magnetic ballasts that operate at 240v an have a 1.6A normal operating current and an order of magnitude of about 13 x nominal on switch on.

So for a few milliseconds with 400w lights I have theres a potential for a 21A spike during switch on.

[tommygdawg]

Might be a bad breaker. They do occasionally fail and randomly tripping is a symptom. Can you try it on another circuit or is the ballast permanently wired?

I did have the issue on another circuit in the house, though to be fair, on that circuit I was trying to strike the lamp at full power rather than dimmed down to 50%. When it tripped the breaker at 50% on my other circuit, I assumed it would do it on any circuit. And I do have another 1200 watt discharge lamp that I run on the same circuit (not at the same time) and I've never once tripped a breaker with it.

Inrush current for discharge lighting can often throw MCB's as the current at switch on will be in orders of magnitude over the nominal operating current for a few milliseconds.

I use 400w metal halide lighting with magnetic ballasts that operate at 240v an have a 1.6A normal operating current and an order of magnitude of about 13 x nominal on switch on.

So for a few milliseconds with 400w lights I have theres a potential for a 21A spike during switch on.

Wow! Have you ever measured your spiking voltage when you turn it on or are you just basing that off the maths? I've read from a few sources that discharge lamps (mine is a metal halide as well) with electronic ballasts (versus magnetic like yours) don't have that huge inrush current, but rather slowly build up their current draw. This is represented on my Kill A Watt, but I don't know if that thing has a resolution fast enough to display a huge inrush current that lasts only a matter of milliseconds.

-EDIT-
Just tried it on a 20amp circuit with no issues, but will have to try it a few more times before I know for sure that it will strike each time.

[bills]

Inrush current for discharge lighting can often throw MCB's as the current at switch on will be in orders of magnitude over the nominal operating current for a few milliseconds.

I use 400w metal halide lighting with magnetic ballasts that operate at 240v an have a 1.6A normal operating current and an order of magnitude of about 13 x nominal on switch on.

So for a few milliseconds with 400w lights I have theres a potential for a 21A spike during switch on.

Andrew
I did some testing on 400 w highbay lights with magnetic ballasts (277v) and found the current ramps up very slowly
and doesn't reach peak until the lights are fully lit.
maybe you had a bad ballast or the wrong/bad cap.
HID lamps don't conduct much until the mercury vaporizes.

[_Andrew_]

The inrush current can last just a few milliseconds so MCB's rated on the low side for the inrush current can periodically trip.

I'm in the UK so this is all at 230~240V
A from experience I have found with the lights I have that a 16A type C MCB will handle 4 x 400w units with out nuisance tripping.

Found this document on discharge lamps that suggests a inrush current of up to 25 x the start current for the first 3 ms then 7 x the start current for the next 2 ms after that.
http://www.veelite.com/wp-content/uploads/Protection-of-Lamp-Circuits3.pdf   

[Xplode]

Actual average current draw will ramp up slowly, but I was under the impression there is still the brief spike while the ballast is energized, or possibly when the bulb actually strikes the arc.  Cheap electronic ones could be seeing a large spike as they charge up capacitors internally as well I would think.


As for the original poster and his tripping breaker, I know here in Canada we're only allowed to load our breakers to 80% of their rating (I'm an electrician, so I'm talking about electrical code, not everyday people just plugging things in) so you don't have this sort of problem.    But if you are saying that your 1200 Watt lamp is drawing 14amps at 120v, then you've got some large inefficiencies there.   Most of the time I see things like 400W HID fixtures rated at something like 465W (on the labels).   Scaling that to a 1200Watt fixture, you should still only be drawing around 11.6amps... not 14.  Usually my fluke clamp on meter doesn't even show the rated numbers, its lower...  Something doesn't seem right with your numbers.


Also, how absolutely positive are you that there isn't something else on that circuit?  Houses are wired with multiple outlets/light fixtures on a circuit (up to 12 usually). And then out of further curiosity, what brand of circuit breakers are they?

[bills]

PDF of trip times for circuit breakers.http://static.schneider-electric.us/docs/Circuit%20Protection/Molded%20Case%20Circuit%20Breakers/0100-400%20A%20Frame%20FA-LA/FA-FC-FH/0600DB0105.pdf

I stand corrected but even with 25x load for 3ms I don't think it would trip a residential breaker.
 We are also comparing solid state ballasts and magnetic ballasts, what are the specs. on the solid state ballasts?

[tommygdawg]

Thanks, everyone, for the additional thoughts Some additional info:

I'm positive there's nothing else on the circuit. It's a bedroom that no one is using at the moment so there's definitely nothing else plugged in.

Regarding my ballast, I don't know exactly what the specs are. Since it's a Chinese knockoff it's difficult to find exact specs on it. I can tell you the specs of the Arri ballast it's based on and I can also tell you it doesn't match up For instance, it's supposed to have a power factor of .98 but the Kill A Watt measures a power factor of .72. Even with a power factor of .72, a 1200 watt light shouldn't be drawing 14 amps. That said though, the ballast does have additional features such as a flicker free (converts the electricity into a pure square wave I think) and 300hz mode, so maybe that additional draw is to power some internals that make those modes work? It's also labeled as a universal 575/1200/1800 ballast, so it can power lights of any of those power ratings (though I've only tried 1200). Maybe that affects it somehow?

I don't know what brand the breaker is in my house, it's unlabeled.

[johnkenyon]

Why don't you use a breaker with the same current rating (e.g. 15A), but with a different trip profile.

Or to use old technology, don't use a fast-blow fuse, use an anti-surge/slow-blow fuse.

Note that a slower tripping breaker may mean that you have to have cabling with a higher current rating (because fault which results in a high but not short circuit level current will take longer to trip the breaker).

(Note that this post comes from the 240v side of the Atlantic...)

[tommygdawg]

Why don't you use a breaker with the same current rating (e.g. 15A), but with a different trip profile.

Or to use old technology, don't use a fast-blow fuse, use an anti-surge/slow-blow fuse.

Note that a slower tripping breaker may mean that you have to have cabling with a higher current rating (because fault which results in a high but not short circuit level current will take longer to trip the breaker).

(Note that this post comes from the 240v side of the Atlantic...)

Well, the issue is I need portability. These are film lights and as a filmmaker I generally need to be able to plug them in wherever I'm shooting, so what I've been trying to figure out is why this ballast occasionally trips breakers.

[_Andrew_]

I have a fare amount of dealings with production power and even though when I'm powering kit at 240v single phase, when it comes to a discharge lamp at 1.2kw I would drop a clean 16A supply to it on a 16A type C RCBO.

I would still be inclined think that the problem you are experiencing is lightly to be due to the inrush current and on the odd time the trip pops is just down to it being unlucky to have been turned on at just the moment of the mains AC cycle where the highest inrush current is lightly to occure.

Found another article on inrush current. This person has gone in to a lot of detail on the subject.
http://sound.westhost.com/articles/inrush.htm