Exploding lamps

[Sayen]

I was reading through old posts, and found this: http://www.controlbooth.com/forums/lighting/4866-stage-lights-bulbs-2.html

Ship said, "This all as opposed to real blow outs where the filament possibly due to a voltage spike while already operating at times will shoot thru the outside wall of the glass in the glass having like a bullet effect worth of hole frozen in it, and the opposing side of the lamp having a puckered like appearance due to the gasses sucked out of the lamp."

I have a number of HPL and EHG lamps that failed this last year, and when I pulled them out they all showed outward explosions through small holes, some taking bits of the filament through the envelope. I'm familiar with finger print breaks, but and these didn't look like anyone touched them. I assumed I had purchased a batch of bad lamps, possibly flawed glass, until I saw your post. We have a new building that is bug ridden beyond description - how likely is a voltage spike on a regular basis? Is there a way I can test for this? If I can't prove the problem exists on my own, I'm out of luck.

I lost two expensive reflectors to exploding lamps. Real fun.

I appreciate the help.

 

[JD]

Many times, when a filament ruptures, the lamp briefly turns into an arc lamp with no current limiting, bypassing the resistive aspect of the filament. This can cause a sudden super-spike in the internal lamp pressure. This "explosion" can blow the filament through the quartz shell of the lamp. This is not the norm as in most cases the filament stays inside and partially vaporizes leaving a flash mark and some beads, or beaded ends of melted tungsten, or, the lamp fully shatters.

Still, "The filament has left the building" does occur. The best place to capture all the players in this attempted escape is in a VNSP par bulb that has undergone filament ejection. I am a big fan of VNSP lamps so I have seen a lot of blown ones over the years. I can only remember one or two that did the trick. Pretty neat looking, seeing a quarts tube with a small hole and a big (usually stretched out) spring rolling around inside the outer glass jar!

 

[ship]

Fairly rare the filament leaves the building as said. Seen a few both of halogen and incandescent which says if one is soft lime and the other quartz glass it's not persay about fingerprintes but than again could.

Bugs bad - had one really ugly one in my garage tonight. Could be from a vaporized bug on a lamp but really hard for them to get in. Mostly not from fingering the lamps one would assume for the most part though similar. Possibly a reflection on the glass causing heat to build up in an area opposing that reflection - the opposing side of the glass, and within the globe given gasses/heat circulate from filament - hottest part to cooler parts, that if something is intensifying heat there is a certain wind blowing a filament towards an opposing side of the lamp - this also would bubble the glass. That stretched bubble - see them in all types of lamps called a divitrification is often in a moving light said to be caused by a small area of dirt on the lamp and has been at times attributed to dirty lamps. I more call it elephantitus of the globe to the same effect.

Now have a few things that could cause a bubble to form into the glass, even blow a filament towards the glass including the filament even melting its way thru the glass while still working - got one of them also.

Could be a scratch in the globe but that should shatter in forming a weakness that wouldn't stretch. More something on the globe either causing a reflection that the other side gets reflected light on & as it gets heated it gets blown away as if a baloon, thus also cooler and grows from there in a wind effect. Or something on the bubble side of the lamp absorbing heat which causes a cool spot and a bubble.

Bubble eventually blows and takes all the heated gasses out of the globe towards the lower pressure of the atmosphere. This wind - filament having been blown towards the divitrification or not sometimes on a more warn filament that has less holding it in place, or a filament with less filament supports could get sucked out instead of vaporized as normally happens when a lamp suddendly gets an inrush of cooler air.

Hard to say, could also be a bench focus problem with the fixtures in causing a hot spot by way of proximity, dirty reflectors etc. Lots that could cause it and if you are seeing a lot of it, it should be further investigated. Same fixtures or at random for instance? Same hang positions? Same lot number of lamps?

As for voltage spikes, easy enough to meter, an electrician would just hook up a meter with diagnostic tools to your system over a period of time and be able to analize for instance when the amusement park next door shut down for the night you might due to a sudden lack of current draw get a spike in voltage now that a huge user is no longer drawing as much. Could at times also be because of proximity to a power station or transformer for an area and other things with switching between. Mostly if that's the case of a voltage spike the lamps will just blow without blowing up in the above way.

 

[Van]

Testing for a voltage spike isn't too awful difficult if you have a good meter. A lot of meters have a peak hold or record function incorporated. Simply plug the leads into a socket in the area you want to check. place a sign near by that say "do not touch" and walk away. The meter will hold a record of the highest and lowest voltages present. I was able determine a daily voltage flucuation in an old shop I used to work in by using this method.
As to the lamps, Ship is Da Man when it comes to lamps, I think he'll agree that bad batches happen, but they are rare. A lot of times you can burn through a lot of lamps in the first couple seasons at a new facility. I find this is more often due to peoples unfamiliarity with new systems. Popping on all the stage lights without warming them first, leaving stage lights on over night or using janitors using them for work lights is not uncommon either. Now if lamps keep blowing in this same manner after you work through your "initial" batch of lamps there may be a deeper issue. Are they 110 lamps and you dims are pushing out 115 - 120? this does occur and can significantly shorten lamp life.

 

[Sayen]

I love these posts, thank you all. I thought it was rare, and therefore weird when I suddenly had a stack of lamps all doing the same thing. I'll take all your suggestions - we're in the middle of cleaning and benching all of our instruments anyhow, and I did tag the ones that blew last year. New batch of lamps should work for a test, just hope this doesn't get expensive.

Thank you too Van - I'll have to hunt down a meter that's slightly fancier than mine.

 

[quarterfront]

Slightly different situation but I went through a period a couple years back where I had FEL's popping right and left for a while. The kind of lamp failure where the lamp explodes with a pop that sounds like teeing off a golf ball with a steel driver and you're left with nothing but the base and a little stub had been really rare, then gradually it got to where I was seeing this happen twice or more a week. Doing lamp inventory I noticed that I'd been ordering more FEL's more often over the past 8 or so months. Ordered a different brand of lamp (+~$2ea.) and over the next two months the problem decreased as the old stock worked it's way out of inventory.

Don't rule out quality control as a factor - "inexpensive" and "cheap" are two different things....

 

[JD]

One way to catch voltage problems is to use an APC UPS supply. (sold for computers) Many of them come with "power doctor" software and a USB cable that allows you to track the power conditions on the computer. At $40, it's a real cheap way to monitor suspect power.

 

[LightStud]

Hmm. I wonder if JD's solution is more or less expensive than this one: Dranetz-BMI Products

 

[visigoth]

I had a batch of EHG's blow out the side, about 6 or 7of them. Returned them to Barbizon and they replaced them, so dont throw them out get a replacement, they aint cheap!

 

[cdub260]

A couple of years back, when we first started using dimmer doublers at the Pageant, I had an exploding lamp incident while bench testing a Source 4. My 120 volt fixtures use standard stage pin connectors, while my 77 volt fixtures use L5-15 twist locks. That first year using the dimmer doublers, I built adapters from stage pin to L5-15, figuring that in a pinch, I could simply change out the lamp and slap an adapter on the fixture. BIG MISTAKE! When I was maintaining this fixture, I didn't notice that I had a 77 volt lamp and a stage pin connector. I had the light cleaned and inspected, though apparently not thoroughly enough, and ready for testing. I plugged it in and the lamp exploded, damaging the reflector and scaring the crap out of me.

Needless to say, I no longer use adapters to turn a 120 volt fixture into a 77 volt fixture. I used this incident to justify buying spare lamp bases, so that I could quickly go back and forth between 120 volt and 77 volt. Another safety measure that just occurred to me, which I will be implementing as soon as I show up for the show tonight, is I need to separate my 77 volt lamps from the rest of my lamps in their storage cabinet.

 

[LightStud]

Ouch!, and a little bit frigthening. Yet another disadvantage of the DimmerDoubler. I'd always have a Variac on the workbench.