Exploding PSMH Ballasts

[edwardchcg]

This isn't a new topic, but by the time I searched it out today, the prior threads were closed. That being said, I had an indoor, 150 watt, pulse start metal halide, wall pack fixture wired for 277 that had its ballast explode about 6 to 8 months ago, destroying the fixture. I never figured out the cause, chalked it up to something that couldn't be determined after the fact, and just replaced the whole fixture. Last night the new fixture, in the same location, had its ballast explode even more violently than the first one.
There was plenty of speculation in the previous threads as to the cause of pulse start metal halide ballasts exploding, but aside from that, a few of the posters claimed to be dealing with similar occurences and were waiting for responses to inquiries they put in to manufacturers. I wonder if anyone has heard anything definitive explaining what's happening here. I don't want to simply replace the fixture again, as this is in a production area. But before moving on to a new lighting technology for this location, I'd be happy to hear any answers. Thanks for your time!

 

[Dennis Alwon]

I would certainly contact the manufacturer. I assume the proper terminals were used with the appropriate voltage.

 

[kwired]

I can see a lamp exploding or even a capacitor exploding (though they are supposed to have a fuse inside that prevents this) but can't really see a ballast which is pretty much just a transformer explode. I can see an insulation breakdown and a flash over event occuring, and if violent enough maybe rupture of whatever it is enclosed in. A lot of this would be dependent on available fault current at the place of the incident. Individual fusing of each luminaire could help if this seems to be a frequent problem.

 

[mgookin]

Is this fixture one of one? Or one of many?
Does the production have heavy loads? Steel mill or ladies cutting fabric with scissors?

 

[edwardchcg]

I installed the fixture myself, but I would be the first to say that doesn't guarantee it was done correctly. I'm human. The problem is that there was no way to verify the installation afterward due to the damage caused. It did operate correctly for 6 to 8 months as best as I can tell, though.
This fixture is one of many that we have in the facility. We get them through a contractor that we use all of the time, and the contractor/owner is doing the leg work of contacting the manufacturer. When I get a response from the manufacturer and/or contractor, I'll be happy to share it with the forum. Until then, I'm really just looking for someone who has gotten anything in the way of information that can be considered a concrete explanation.
Thanks for the insight everyone.

 

[GoldDigger]

I installed the fixture myself, but I would be the first to say that doesn't guarantee it was done correctly. I'm human. The problem is that there was no way to verify the installation afterward due to the damage caused. It did operate correctly for 6 to 8 months as best as I can tell, though.
This fixture is one of many that we have in the facility. We get them through a contractor that we use all of the time, and the contractor/owner is doing the leg work of contacting the manufacturer. When I get a response from the manufacturer and/or contractor, I'll be happy to share it with the forum. Until then, I'm really just looking for someone who has gotten anything in the way of information that can be considered a concrete explanation.
Thanks for the insight everyone.

Since the pulse-start ballast has to contain a way to produce very high pulse voltages from normal AC input, I would be concerned about the effects of over voltage or transients on the input AC producing too high a voltage which could then break down insulation leading to an eventual explosive failure.
But that is just a guess. I would do what I could to verify the power quality to the fixture though. That would be independent of your wiring.

 

[jusme123]

was it connected to a multiwire branch circuit?

 

[kwired]

Since the pulse-start ballast has to contain a way to produce very high pulse voltages from normal AC input, I would be concerned about the effects of over voltage or transients on the input AC producing too high a voltage which could then break down insulation leading to an eventual explosive failure.
But that is just a guess. I would do what I could to verify the power quality to the fixture though. That would be independent of your wiring.

I see that as being more likely to rupture the ignitor or capacitor than the transformer though. The transformer is probably stout enough to absorb transients and if effected by them would cause additional heating and eventual insulation breakdown. You would think if bad enough there would be a lot of other equipment failure in the facility also.

 

[edwardchcg]

was it connected to a multiwire branch circuit?

Nope. It came straight from the panel with just a hot and a neutral.

 

[TNBaer]

I had the same issue several years ago which I believe I posted about here. Twice in the same installation a PSMH ballast exploded on a recessed step-light (70 watts, I think). Both explosions were so violent is sheared off the heads of 4 screws holding the luminaire's face on and flung the faceplate across 3 or 4 lanes of traffic. I never did learn what caused it or what could have caused it. We were lucky no one was hurt.