Cinematography Mailing List - CML
    advanced

Wire Guard’s vs Safety Glass

Over the years we've lost the wire guards on some of our Blondes, Redheads and Lowell Dp's. I'm wondering if any of you have an opinion as to safety glass vs. wire guard option. I'm wondering if safety glass might contribute to shorter bulb life due to ventilation issues. On the other hand the wire guard is more or less a scrim.

Opinions?

John Roche, Gaffer
Liberty Lighting Ltd - NYC


>I'm wondering if safety glass might contribute to shorter bulb life due to >ventilation issues. On the other hand the wire guard is more or less a >scrim.

And I wonder how much of a real world safety issue it all is when I look at odds based on my career. I mean I can get into an accident driving a car but odds are good, if I do maintenance, and take care to drive correctly, I will not end up with a problem. I can say that I have had one single lamp explosion in all the years I have worked and would guess off the top of my head that I have worked with no less than 100,000 different fixtures including my years as a studio location lighting designer and all the field work, both incandescent and gas. I have had 1000 times as many lens cracks.

Secondly, what are the odds that even if you are using an open face such as Lowell or a Red head that you don't have some sort of diffusion material on it which helps when the fixture might have a problem? I say screens are good enough. And frankly some of the older open face fixtures simply don't really have ways of putting screens on them that is all that safe and not some sort of fire hazard.

Of course others may have been in more accidents than others but like most of these cases of statistical analysis, when you look at the group you find the problem is a small exception.

So how many of you have had explosions? Or better asked since all the explosion folks will chime up making it look disproportionately like fixtures are powder kegs, the question I should ask is how many have never had such a problem or did and it wasn't anything like the explosions that seem to me to be rare and usually caused by ignorance or sloppiness?

Now for you purists and conformists, I don't want you to think I am telling you to let children stick their fingers into a hot fixture or that you should aim a lamp strait up and have a spitting contest into the fixture. I'm simply trying to find out if some of these stories are more of a rarity than a fact which drives fear, and fear is never a safe way to work.

Walter Graff
Producer, Director, Creative Director, Cinematographer
HellGate Pictures, Inc.
BlueSky, LLC
www.film-and-video.com


Walter Graff wrote :

>And I wonder how much of a real world safety issue it all is when I look >at odds based on my career.

Since about 1992 I've had only one open face bulb explode. I remember it well as it was rigged in a ceiling and I was under it when it exploded. I was so lucky to have been looking down at something on the floor as it rained hot glass everywhere. Fortunately, I ended up with only a few minor burns on my forearm. Perhaps, once is enough.

One other note : I can't tell you how many blown 1K and 2K bulbs I've replaced that had a huge bubble in the glass. These always struck me as "almost exploded".

John Roche, gaffer


>One other note : I can't tell you how many blown 1K and 2K bulbs I've >replaced that had a huge bubble in the glass. These always struck me >as "almost exploded".

Problem is that there isn't nearly enough pressure for explosions that could really send folks flying of cover99.9% of the time. That bubble in the quartz came from someone's hand or some other contaminant that eventually caused it to bulge in a process called devitrification and the fixture loses halogen pressure or the filament starts to elongate and sag from shock and/or use (to much heat) and it gets close to the quartz causing failure. I don't call it near explosion, just failure.

Walter Graff
Producer, Director, Creative Director, Cinematographer
HellGate Pictures, Inc.
BlueSky, LLC


John Roche writes :

>One other note : I can't tell you how many blown 1K and 2K bulbs I've >replaced that had a huge bubble in the glass. These always struck me >as "almost exploded".

I'm not so sure about that.

The reason some lamps fail so dramatically is because when the filament goes through, there is an intense hot spot at the point of filament fracture which vaporises some of the tungsten. Since it only takes a little metal vapour in the lamp to turn it into a discharge lamp, an arc discharge occurs between the live filament end supports. Since there is generally no choke to limit the current, this results in a very short lived, high power discharge that causes a bright flash an often catastrophic failure of the lamp.

This is why your bathroom lamp occasionally scares the shit out of you by giving a bright blue flash, making a loud TINK! noise and tripping the lighting breaker.

If you see a filament lamp that has failed suddenly and gone "metal plated" inside, then it was a near bang experience.

The most common cause of explosive failure of discharge lamps is arc instability at the end of life, where the electrodes have burned back and also lost their electron emission ability which raises the voltage across the lamp and results in increased power dissipation. It seems to be a sudden avalanche effect which fits with the flickering then explosion effect. This is why expensive intelligent fixtures have lamp life timers built in, that shut the fixture down after a set time has elapsed since a new lamp was installed and the timer reset. Many intelligent fixtures also monitor the voltage across the lamp, and if it rises above a set level, they shut the fixture down and flash an error message.

Of course, some lesser individuals try to save money by simply resetting the lamp life timer. This sometimes results in the need for a new reflector and collimating optics!

Clive Mitchell
http://www.bigclive.com


Walter Graff asked :

>...So how many of you have had explosions?....

I can recall only 2 in 42 years, not counting the first one I witnessed while a student: a 5K fresnel back light exploded because it had been tilted down too strongly for too long. But it was a conventional globe anyway, not quartz.

The wire guards for the Lowell Totas have to be installed and removed each time you use them and I'm not convinced they'd do all that much if the lamp did explode. I think they'd pop off. If a screen guard can be secured and not interfere with the light's function, fine. I'm for safety, even if the odds are long that it will happen. But if they are a pain to use, I go with the odds.

I haven't disengaged the airbags in my cars because the odds of them harming me and mine are so long and the possibility of their helping in an accident are good. But they are passive, in respect to the cars' operation. If the wire guards are also passive, I'll use them.

Wade K. Ramsey, DP
Dept. of Cinema & Video Production
Bob Jones University
Greenville, SC 29614


Clive Mitchell writes :

>If you see a filament lamp that has failed suddenly and gone "metal >plated" inside, then it was a near bang experience.

Now I know why blown Maglight bulbs always look this way.

Mitch Gross
NYC DP


If only we could post photos on this list, I'd supply some close-ups of several redhead safety glasses I've just received from a client who has moved to safety meshes. Yes, the glasses are about 20 years old and the client is a high-school TV studio where I'm sure the lights haven't "always been used for the intended purpose" or rigged appropriately. But my point is there are several small pits in the glasses where they have caught exploding bubbles, with some pits still containing very small blobs of what I'm assuming was molten filament that is now welded into the safety glass.

I've been on a set many moons ago where an unprotected redhead bulb exploded (fortunately whilst the actors were on break) and electrics went over and picked glass bulb fragments out of a plasterboard wall three metres from the redhead!

Now just for the record, I'm referring to 240v 800w DXX bulbs. Maybe these behave "more aggressively" than their 120 volt cousins? But I'd never have anything to do with open-faced lights that didn't have either a safety mesh or safety glass in place. Its not worth the risk, no matter how small that risk may be in some people's opinion. One only has to look at the very short rated burn time of these type of bulbs to appreciate the enormous stress they operate under.

As for the choice of glass vs wire, yes, the wires do reduce the intensity of the beam by a small amount. But I expect the glasses do too, particularly if they are "well used". I believe the wires would at least reduce the speed and limit the size of exploding glass / filament.

>One other note; I can't tell you how many blown 1K and 2K bulbs I've >replaced that had a huge bubble in the glass. These always struck me >as "almost exploded".

A local ex-gaffer here always referred to those bulbs as being 'pregnant'

Keep it safe, guys,

Clive Woodward,
video techie whose rebuilt his fair share of redheads,
Perth, Western Australia.


> "So how many of you have had explosions?" (Walter G.)

I've had enough explosions to make me fear for the safety of myself and others on the set. Here are a few of the notable ones.

2K Mighty-Mole that was just re-lamped with a defective lamp (bad gas mix). Blew up in a bank lobby, scattering hot shards of glass onto an expensive carpet. Created small fires in the 4 areas where the shards landed. I had to stomp them out (which probably looked humorous at the time). That was the last time I used an open-faced light without screen.

2K Scoop (it had a tungsten halogen lamp within a "soft" glass envelope BUT NO SCREEN) which blew out at a Big Bird Birthday party at a PBS station. Imagine a bunch of children standing around a birthday cake holding balloons. Several balloons blew up from the flying glass, but none of the children were hit. The children thought that it was a special effect and were delighted with the "magic". We whisked away the birthday cake which could no longer be eaten since it was sprinkled with bits of glass. That was the last time I used a scoop without window screen over the opening. I used to have nightmares over this one.

5K Fresnel behind a large Chimera had a lamp explode for unknown reasons.
Fixtures was angled about 40 degrees. Some of the shards of glass fell through the cooling holes in the fresnel and melted the floor beneath it. I'm still not sure what you can do about a freak blow out like this one.

1K PAR 64 on a telethon blew out. Shattered the outer globe/lens and fell to the deck. Nobody hurt. That was the last time I hung a PAR 64 that didn't have an inner retaining screen.

I believe that if I can prevent one person from being hurt by my lights that it's a fair trade for a life-time of vigilance. The statistical chance of an accident might be low, but I won't depend on good luck alone to keep people safe.

Bruce Aleksander
LD/DP/Chief Curtain Hanger
ABC/Disney
Houston, Texas


>I've had enough explosions to make me fear for the safety of myself and >others on the set. Here are a few of the notable ones.

Bruce, I don't want this to turn into "Walter said that lamps don't explode and there is no threat of safety on a set". I never said such things. What I said was that the reality of explosions is minimal. That does not mean we are not safe on a set nor should be.

I use a utility blade on a set too and I could easily say that you have to be very careful with them because lots of people cut there hands. Of course you have to be respectful of a knife. Just as you have to make sure you hang fixtures correctly. I just don't like the 'fear factor'. Fear creates unsafe working conditions.

Teach someone to ski and you find that their biggest hurdle isn't learning to sky, it's getting over the fear. When you are trained to be an electrician you are taught not to fear electricity but to respect it. Big differences. Fear is what has caused most electrocutions in the electricians industry.

As a professional, I work as safe as possible. I don't spend my day sitting in the corner hoping that nothing will explode. The point of the original post I made was to put a bit of reality to the explosion scare that I saw might have come from it all. It's just like this term we have in the states often called "going postal". It comes from a spate of homicides by postal workers in the nineties. But the reality is that these cases were isolated. It was the media that created the illusion that postal workers were more likely to kill other workers. The Postal Services¹ Commission on a Safe and Secure Workplace showed statistically that was nothing but a myth. The truth is that postal employees are only a third as likely as those in the national workplace to be victims of homicide at work. Taxi drivers and chauffeurs are 150 times more likely to be killed on the job and police rank a distant second in homicides.

But ask someone to tell you horror stories and all you hear are horror stories. You personally have an exceptional number of cases as I see it based on talking to a number of people about this. I can understand why you have an extra ordinary number of cases. Studio fixtures are rarely maintained properly from my experience as a studio lighting designer. Rather they sit on a grid until something happens. In addition the quality of the workforce in many studios I have worked in is not of the calibre of many of the field people I have worked with meaning in my mind that they don't handle things as correctly. I have seen many incorrect things done by many studio folks that led to problems with fixtures later on.

The good thing about field production is we get to touch every fixture all the time and see things happening before thy get bad for the most part, or at least more than the case of a studio fixture. That might be a reason why most folks I have talked to about this have rarely had a problem. I just want to balance the conversation by saying that explosions are rare, not that you shouldn't care. Proper handling of fixtures assures safety. On occasion a lamp can explode regardless of how safe you are. Most every time, no one ever gets hurt as you verified from you post.

But like any other accident the potential is always there. Screens or whatever you use for protection are important, but I don't want to make unsafe working conditions by having people second guess themselves because all they hear are horror stories on the CML.

Second guessing and fear is a mistake waiting to happen. You never know when a lamp might explode. Odds are good that it will never happen to any one person. If it does, and if your fixture has no protection, which sometimes can't be avoided, you get the broom and clean it up. If you have protection, all the better. Be safe and don't skimp on safety, but never fear the instrument you work with, rather respect the potential.

Point made,

Walter Graff
NYC


Mitch Gross writes

>If you see a filament lamp that has failed suddenly and gone "metal >plated" inside, then it was a near bang experience. Now I know why >blown Maglight bulbs always look this way.

It doesn't really apply to low voltage lamps, since the low voltage won't really sustain an arc. It's more likely that the dying lamp just vaporised a burst of filament in it's death throes. Can't say I've ever seen this, since my Maglite lamps tend to fail due to impact! (Another reason for LED torches)

Clive Mitchell


Bruce Alexander writes :

>I've had enough explosions to make me fear for the safety of myself and >others on the set.


I second all of what Bruce says.

I once owned a lighting rental co. and we had a rash of HMI explosions. Some of the heads came back to us looking like a small grenade had gone off inside. It was impossible for us to tell whether an exploding bulb blew off the lens or if an exploding lens blew out the bulb. Needless to say each respective manufacturer blamed the other.

All anyone on the set could say was that the light was on and was operated in normal positions, on crank ups, on level ground. Then a big bang.

Mercifully, no one was injured. However two small grass fires were started and quickly put out by the crew. And one automobile had to be repainted. the heads were another story.

There were all kinds of theories put forth as to why it happened:

1/.  The hot lens got wet. It wasn't raining.

2/.  The globe was focused too close to the lens. It happened in mid-range.

3/.  Bad Globe, old globe, damaged globe. It happened with new globes.

4/.  Bad lens. Lenses were replaced and it happened with new lens.

5/.  Bad ballast. Ballasts checked by manufacturer several times.

6/.  Bad crew It happened to several different crews. It happened on a job I was on. No one was near the light. Moderate temperature day. Head about 12' off level ground. Medium spot. On for about 15 minutes. Bang. Lens and globe shattered.

7/.  Bulb too tight in fixture.

8/.  Bulb too loose in fixture.

9/.  Lens too tight in fixture. No allowance for thermal expansion.

10/. Lens too loose in fixture. Too much allowance for thermal expansion.

11/. Moving the light while hot. It was standing still.

12/. Over voltage, voltage surge, bad generator.

13/. Under voltage.

14/. Overheating.

The final result was we changed a lot of lenses and a lot of bulbs and nothing else and the problem went away.

If memory serves there was an instance or two in which the lens shattered but the bulb remained intact. This pointed very strongly to the lens manufacturing, but no one ever admitted any problem.

IMHO, putting up a light where there are people around without some kind of barrier in front of it, is like hanging a light on a grid without a safety cable or chain. It's very inexpensive insurance.

Brian Heller
IA 600 DP


>I once owned a lighting rental co. and we had a rash of HMI explosions. >Some of the heads came back to us looking like a small grenade had >gone off inside. Needless to say each respective manufacturer blamed >the other.

And how long until you were replacing lamps in these fixtures? Give me some
HMI wattages and the expected average life you guys were getting out of them.

Walter Graff
Producer, Director, Creative Director, Cinematographer
HellGate Pictures, Inc.
BlueSky, LLC


Walter Graff writes :

>And how long until you were replacing lamps in these fixtures?

We normally replace lamps based on color temp not burn time.

We used to track burn time very carefully, but found it's not worth the trouble. Some fixtures don't have timers and some timers don't work with square wave ballasts.

At the time of the incidents some bulbs had well under 100 hours. As I stated, we strongly suspected it was the lenses and not the bulbs.

Give me some HMI wattages.

6K and 12 K fresnels. We have had other incidents (Bursting globes) with nearly every wattage of incandescent, from FEV to DPY. I don't recall any DTY bursts however. I think I'd remember.

>and the expected average life you guys were getting out of them.

6Ks are rated for 400 hours, but we seem to be relacing them at about 250 hrs. 12Ks are rated for a slightly longer life and longer replacement interval. I can't recall the exact numbers.

Brian Heller
IA 600 DP


Brian Heller writes :

>The final result was we changed a lot of lenses and a lot of bulbs and >nothing else and the problem went away.

Occasionally there seems to be a run of bad lamps from a particular manufacturer. Sometimes it pays to switch manufacturer if you have some bad lamps.

Another issue with some modern colour changing floods has been the excessive heat build up on the lamp by the optical system focussing light back onto the lamp itself. I can think of one particular brand that uses sliding dichroic cylinders around the lamp, that has had a big problem with it's higher power fixtures. Fortunately the expensive dichroic cylinders contain the blast in a controlled manner. NOT!!!

Clive Mitchell


> We normally replace lamps based on color temp not burn time.

Ding Ding! Huge bells go off! I'll tell you what I was taught over the years. First lets talk about life expectancy for HMIs. Read this carefully, there is NO such thing as a life expectancy with an HMI. I know you thought you saw it listed once, or the union told you this. No, you saw what is called average life expectancy. BIG DIFFERENCE. There is not a single manufacturer who will give a life expectancy for HMIs and for good reason, HMIs bulbs are completely unreliable light sources and no manufacturer wants to do anything to get themselves in trouble.

What do I mean when I say "unreliable"? Many white papers on HMI light sources exist and all say the same thing. HMI are not reliable light sources unless under very controlled conditions. As a result, you get averages from manufacturers and in return they don't get lawsuits. So an average of 500 hours for a bulb could easily mean 250 or 750 hours of usable life. I just used a very important term. That term is "usable". Some HMI bulbs will burn for 1500 hours, even 2000. In fact for most people, it's burn them till they explode or become so green as to be usable. But just because an HMI hasn't exploded doesn't mean it should be in service. I find that seems to be a big problem, using HMIs well past when they should be. So for a average life listed as 400 hours that means past 100 you are taking chances. (I'll tell you why I say that in a moment).

And you answered two questions for me. First, you were part of a rental company. When I think of many rental companies I often think of the slogan, if it works it's rentable. Great for incandescent but not so for HMIs. Not accusing you of this attitude but it has been predominant from where I stand.

Second, you used lots of HMI's. Lets face it, most folks don't replace bulbs when they should. There is only one real reason why HMIs explode. It's called devitrification. Devitrification is a complex reaction of the elements in quartz glass that causes the glass to turn to crystal. Only four things cause that to happen, three of those things in at least two combinations. First and most prevalent is heat and time, better known as too much use. Second is heat and alkali contaminants such as you hand or elements in the air such as hydrocarbons from automobiles. The fourth factor is over voltage. I can guarantee that most all explosions of HMIs where caused not by anything other than simply using the bulb far longer than it should have been, followed by contamination, followed by over voltage. An Osram technician told me that you should never use any HMI bulb for more than 25% of the bulbs rated average life expectancy. And every manufacturer will tell you never to judge bulb life by color rendering.

>We used to track burn time very carefully, but found it's not worth the >trouble.

And probably a big reason why you had so much failure. HMI bulbs are a pain
in the ass but when properly logged you can get incredible reliability. I have a friend who owns a large inventory of HMIs. He is meticulous at maintenance and has never had a single problem. That in my book is often the problem with rental fixtures. Rental houses say its too much trouble to keep track of time so just watch the bulb and see how the color changes, that will tell you about life expectancy. That is about the most incorrect statement that can be made. The complex changes and molecule exchange that goes on across the glass barrier during burn time is a factor in why that color might change faster in one place or at some time than another.

> At the time of the incidents some bulbs had well under 100 hours.

As I said not a single manufacture will give you life expectancy, only average life expectancy. But if you are having multiple problems under 100 hours I suspect alkalis could the cause, not the lens. Just because you put a glove on does not mean your bulb is being safely handled. Bulbs should only be handled with fresh clean alkali free gloves anytime they are handled, not grip gloves that you've been using for three years. Pain in the ass yes, but think about this, when a manufacturing plant such as the one in NH receive the white sand that is already very clean and even has a tracking label so they can track impurities, it is further processed until the impurities are no more than 20 ppms. If you don't know, that is an incredible amount of work and says something about how incredibly important it is that that glass does not get contaminated by the gloves someone used to unload a truck the day before while eating pizza, pick up a dead rat in a studio, or pick his butt with.

>6Ks are rated for 400 hours, but we seem to be relacing them at about >250 hrs.

That statement counters what you said before where you said you couldn't afford to worry about time, but rather relied on color, or are you talking about then and now?

Walter Graff
NYC


> If only we could post photos on this list,

You can, you send them to me or David Walpole and we put them on a web-page.

We're also interested in any comparison tests of any kind that you do.

Cheers

Geoff Boyle FBKS
Director of Photography
EU Based
www.cinematography.net


>You can, you send them to me or David Walpole and we put them on a >web-page.

>We're also interested in any comparison tests of any kind that you do.


Geoff (and list),

Thanks for that.

I just tried to photograph the offending items and discovered two problems :

1/. On closer inspection of the glasses I have here, I only have 'chipped' ones handy. But I think I know a client who has some with embedded metal filament blobs so will track these down...

2/. Trying to convince my auto-focus digital camera that I wanted the blemishes in the glass, not the objects behind in focus! A bit of 'RTFM' might help "

So yes, I'll get some evidence and send it to David Walpole for posting... and while I'm at it, I guess I should pull out a light meter and take readings to compare light losses: open / mesh / glass.

Cheers,

Clive Woodward,
probably a forensic scientist in a former life,
Perth, Western Australia.


>Occasionally there seems to be a run of bad lamps from a particular >manufacturer.

I wonder if that statement has any truth to it. Tell that to a company like Osram. They have such a method for tracking the sand they get, to the batches they then process to remove impurities, to each and every piece they cast and where it goes that they would tell you that is not true, or if it was, they would know exactly where the problem is. The creation of fused silica glass is a very scientific process and the companies that make it have incredibly tight manufacturing standards.

A bad batch might be bad from a bad storage situation but I would doubt that with the money spent investing in making this stuff that companies are simply turning out "light bulbs" that have occasional bad batches. I have been to a plant that makes the glass and I can tell you just from the looks of it, quartz glass manufacturing has the same manufacturing standards as the silicon electronics industry since it's used in both.

Walter Graff
NYC



>Occasionally there seems to be a run of bad lamps from a particular >manufacturer.


I wonder if that statement has any truth to it. Tell that to a company like Osram. They have such a method for tracking the sand they get, to the batches they then process to remove impurities, to each and every piece they cast and where it goes that they would tell you that is not true, or if it was, they would know exactly where the problem is. The creation of fused silica glass is a very scientific process and the companies that make it have incredibly tight manufacturing standards. A bad batch might be bad from a bad storage situation but I would doubt that with the money spent investing in making this stuff that companies are simply turning out "light bulbs" that have occasional bad batches. I have been to a plant that makes the glass and I can tell you just from the looks of it, quartz glass manufacturing has the same manufacturing standards as the silicon electronics industry since it's used in both.

Walter Graff
NYC


* I really don't think there is any way for you to know what I thought I saw, or did not see.

You know I had the honour of having dinner with Ira Tiffen last week after a SMPTE conference. Ira and I had communicated on the phone and via internet for some years but never met in person. At dinner Ira noted that I am nothing like the curmudgeon I portray on the CML. And I am not. I have this problem of thinking more words than I actually write, so instead of me saying things in general, I often use the term 'you' as my pronoun of choice, but Brian I was in no way directing it at you personally. I should have said it in another wordy fashion but my laziness made me pare it down to a simple sentence that made it seem like I was implicating you. I was not.

* but most of the people who deal with them on a day to day basis find them reliable enough :

I do not disagree, just saying a fact which is that as a fixture they are not reliable and hence why folks seem to have trouble with them that they often can not understand.

* Thirdly, as far as bulb life is concerned, the only bulb life that is of concern to the HMI user is bulb service life.

Absolutely but from a life expectancy of the actual bulb, color is not an indication of anything other than what gels you need to correct it of whether it is 'unusable' anymore. I just wanted to make the point that most of the explosions out there are from folks using lamps much longer than they should and/or mishandling bulbs.

*Your term "usable life" seems to invite use until failure.

That was my point. A 575 could end up with a usable life of only 200 hours, but color temperature looks great. Then again it could last 1000 and look great. Simply noting that we have come to accept that HMIs work and that they must be reliable, but when you take a carefully manufacturer device such as this into the harsh cruel world, things don't always act as you think they should.

* As I thought I made clear, real world rental experience has proven it is wildly impractical if not impossible to reliably track HMI bulb hours.

Absolutely. It is nearly impossible but once again, expect explosions because of it.

* Mishandling accounts for the greatest number of bulb explosions by far.

You didn't read what I said or I didn't write it clearly enough. The only reason a bulb explodes is devitrification. Mishandling a bulb causes devitrification as does pollutants (dust/handling).

* Furthermore, not all devitrification failures result in explosions. Sometimes devitrified bulbs just go quietly.

Maybe when the fixture is off. Rarely when it's hot. When the wall of fused glass devitrifies while hot, the pressure of the hot gas can only explode.

* Are you sure he wasn't an Osram salesman pretending to be a tech. I'm sure they'd like their bulbs replaced even more often

It's a valid question, but no, this guy was on the tech side of Osram. I met him up at their US offices where we were producing a video one year. He was rather candid. I don't know if the company would have told me as much as he did.

* Every manufacture has told us the exact opposite.

I am talking about explosions and you are talking about color temperature variables over usage. Not the same. There is a formula for life expectancy. It is a complex equation that assumes lamps are in pristine controlled atmospheres. But the tests done on these bulbs are not done in fixture cavities and hence the resultant explosion potential, to the lifetime can only be assumed and not calculated. In fact the general rule as I was told it by the gentleman from Osram is that 10%-30% derating is necessary for HMI lamps depending on the cavity they are placed in. That is how he came up with the 25% percent rule to avoid the potential for explosions.

Of course who in the real world is going to adhere to that except performance stage lighting directors and tour events. In a laboratory under testing conditions e know the rules, but my point was that in the real world many factors can and do cause unpredictable variations with HMIs. That was the point of my post, not that HMI color temperatures changes, simply that they explode and here are some reasons why outside of someone spitting on one. (I almost use the word "you" in that last sentence and that would have changed the whole way you perceived it. Caught myself). But lifetime as in beginning to end, not whether it looks good on camera has nothing to do with color rendition as far as I got from my discussion.

* Of course many people don't replace bulbs when they should. Many people don't change the batteries in their smoke detectors when they should. Or change the oil in their cars when they should. And if this post was about smoke detectors I could probably explain why folks don't understand why there houses are burning down even when they had a smoke detector, or if it was about car problems I could explain why folks cars seize even though the car originally came with oil in it.

* Lighting companies that survive in the market place do so by meeting their customers demands.

After a while but I have seen plenty of lighting companies that offered crap and stayed in business for quite some time because they were cheaper than anyone else. No that doesn't apply to every genre of production, but you'd be surprised at some of the stuff I've seen which was advertised as being in much better condition than it was. But once again the cheap rule applies.

Thanks for the conversation. I'm glad you stayed on subject other than the misunderstanding in the beginning. I always enjoy your thoughts and ideas.

Walter Graff
Producer, Director, Creative Director, Cinematographer
HellGate Pictures, Inc.
BlueSky, LLC


Walter wrote:

>.... but I would doubt that with the money spent investing in making this >stuff that companies are simply turning out "light bulbs" that have >occasional bad batches

It may be that the glass itself is made consistently to high standards, but apparently bulbs are not. The principal weakness seems to be in the "seal," the pinched glass section through which a foil passes. This is a tough engineering problem, since the foil and glass have different coefficients of expansion and over many on-off cycles-of differential expansion and contraction- the seal may fail.

I buy lots of 575 single ended HMI bulbs and am having a failure rate of about 10%...bulbs which are replaced by the manufacturer. All bulbs failed during first ignition; either the outer shell leaked and a yellow-fringed, white cloud appeared, or the lamp flickers repeatedly and the lead to the outboard electrode melts. When one tries to track down the causes, it seems to be that these bulbs are so difficult to make that they cannot do much better, despite their automated facilities. The company, however, admits a failure rate of only about 2%, and are pretty cagey when discussing the matter.

Anybody else have these problems?

Jerry Cotts
DP/LA


Walter Graff writes :

>I have been to a plant that makes the glass and I can tell you just from >the looks of it, quartz glass manufacturing has the same manufacturing >standards as the silicon electronics industry since it's used in both.

It's not just down to the quality of the quartz used in the envelope, but also factors like stresses induced in the seal of the lamp during manufacturing. These could be caused by a gradual failure of the pinch sealing equipment, or during adjustments to the sealing machines operation.

I once had a full batch of Phillips XOP-25's that gave major problems with both the end pinch cracking, even during careful installation of the lamp, and complete detachment of the electrode inside the tube.

Phillips denied that there was any problem, but strangely we haven't had a batch with that problem since! Because we were made to feel that our equipment was causing the problem, we didn't chase Phillips for compensation. It's a horrible feeling when yet another $120 lamp cracks as you install it.

A good example of an acknowledged lamp problem is the adoption of a diffused area over the pinch section of a lamp. The diffusion of the frosted quartz is to counteract some fixtures that have a reflector system that causes hot spots on the seal.

Clive Mitchell


Jerry Cotts writes:

>Anybody else have these problems?

Yes, oddly enough about 10% of the time

Brian Heller


Walter writes:

>You didn't read what I said or I didn't write it clearly enough. The only >reason a bulb explodes is devitrification.

I would agree with you completely if you said most, or nearly all, HMI bulb explosions are the result of devitrification. In some cases, despite every precaution in handling etc., bulbs with very low time do explode. This appears to be the result of some kind of manufacturing difficulty.

I'm guessing it's uneven wall thickness that allows for uneven heating.

Unfortunately, it is very difficult to get crews to save all the evidence for failure analysis by the manufacturer. Understandably, productions only want to get going again ASAP. Also manufacturers don't seem to be as interested in this as they once were. As Jerry Cotts says, they can be less than forthcoming with explanations.

Brian Heller :

>Furthermore, not all devitrification failures result in explosions. >Sometimes devitrified bulbs just go quietly.

Walter Graff :

>Maybe when the fixture is off. Rarely when it's hot. When the wall of >fused glass devitrifies while hot, the pressure of the hot gas can only >explode.

With the higher wattages yes. With lower wattages, more often than not, they seem to go quietly. Since these bulbs are still more or less intact, it's far easier for the Mfg. to determine the cause.

We generally get one of two answers : problem with the seal, or devitrification. We're not in any position to challenge the mfgs. on their determinations. Maybe they don't want us to know the real reasons, and just tell us that to end the discussion or maybe they don't know or don't really want to bother.

They give us a new bulb, and we move on.

Walter Graff :

>but my point was that in the real world many factors can and do cause >unpredictable variations with HMIs.

Absolutely. The latest cause of failure in the larger wattages is bugs,(flying insects) getting inside. They are attracted by the light, fly over the fixture, get cooked and fall into the vents.

Even a very small insect or part of one landing on the bulb can cause fairly
rapid failure.

Brian "bugged again" Heller
IA 600 DP


While we are on the subject, it is my experience that 575 SE HMI bulbs will not reliably hot-restrike after 1-200 hours despite manufacturers advertised bulb life of 700-1000 hours. I know there is a lot of frustration and myth regarding hot-restrike, but does my experience comport with yours?

Even with new control gear (ballast+igniter), a bulb that will not hot-restrike with an older igniter will just barely do so and not for very long.

Jerry Cotts
DP/LA


>The principal weakness seems to be in the "seal," the pinched glass >section through which a foil passes.

I never discussed that. That is a long post. My discussion concerns explosions, seals cause failure and not explosions like crystallized glass

Walter Graff
NYC.


Jerry Cotts writes:

>While we are on the subject, it is my experience that 575 SE HMI bulbs >will not reliably hot-restrike after 1-200 hours despite manufacturers >advertised bulb life of 700-1000 hours.

I'll try to get some hard info on our shop experience dealing with perennial issue so that we're not just chasing "frustrations and myth."

Brian Heller
I 600 DP


The avg life that Walter talked about with HMI's is the same as for tungsten and incandescent - it is only an average...if you have a lamp with an avg life of 50 hours, for instance, that means that when the manufacturer put 100 of them in a room and burned them at spec voltage, after50 hours, 50 of them will still be burning It is that simple...and that misleading - Half of the lamps will NOT meet their avg life, and the other half will exceed it.

Mark H. Weingartner


Walter writes :

>Not all HMI lamps are designed for hot restrike. Are you talking about >the single ended 575 w/se? If so it is not designed for hot restrike.

>As for average life, the manufacturer rates the AVERAGE at a different >number than you post. What exact model number are you referring to?


Walter...

The Osram 575 w/se is rated at 750 average life, while for some reason the double ended lamp lasts for 1000 hrs. These ratings also specify a precise number of on/off cycles, a number that can be found somewhere on their web site. I seem to recall 8 hrs on at a time, although it may be one hour cycles....or 750/8 (750/1?)=# of on/off cycles. So your mileage will definitely vary.

The 575 w/se is in fact advertised as a hot re-strike lamp and there is literature-again on Osram’ s web site-dealing with various aspects of this feature. For instance, a cold bulb requires 5,000 volts to "ignite" while a hot bulb....and Osram waits a full 15 seconds before the hot re-strike...requires 25,000 volts. The problem is that most lamp owners and operators expect the lamp to hot re-strike reliably and continuously which they wont. So you have the DP calling for a lamp to be turned off...and then on...and then perhaps off and on again while s/he makes up his/her mind. (You know what Haskell Wexler (and others) says; something like "light the set as you think best and then go around turning lights off!") More experienced HMI operators try to pan the light away and back to avoid turning it off...or they flag it. Less experienced operators incur the annoyance of the DP when the lamp fails to re-strike on demand...and the whole family of bulbs/electronics gets a bad reputation (also derived from ballasts which have been kicked around and head cables which go bad).

Jerry Cotts
DP/LA

HMI bulbs seem to take a terrible beating, in engineering terms. There is the seal problem, an 800 degree Centigrade temperature of the glass enclosing the plasma, the complex component gases and compounds within the plasma chamber, and so on.


>The 575 w/se is in fact advertised as a hot re-strike lamp and there is >literature-again on Osram’ s web site-

Interesting. I have an older cut sheet that does not list the SE as a hot strike but only the W/GS as a hot restrike 575 bulb.. The website carries the same 5/20kv strike info for the SE and SEL as the W/GS but it only says cold and not hot for the SE although shows cold and hot voltage requirements.

An important thing to keep in mind is that just because a bulb is rated as a hot restrike does not mean the fixture is capable of hot restrike. Some wonder why HMIs don't always restrike and very simply, if the igniter isn't capable of it, you have to wait. 575w fixtures can be tough to restrike due to the short gap and smaller igniter setups.

Walter Graff
NYC


Speaking of HMI re-strikes and bulb life... A DP told me that if you do not let an HMI come "fully-up" ie., you shut if off before its up to "temp" you shorten bulb life. Any truth in this? He said it does not matter if it's from a cold start or a re-strike. I've heard this before, but have no hard facts. Myth or reality?

John Roche, gaffer


John Roche, gaffer, wrote :

>Speaking of HMI re-strikes and bulb life... A DP told me that if you do >not let an HMI come "fully-up" ie., you shut if off before its up to "temp" >you shorten bulb life.

Reality...Warn your operators to let the bulb run 4-5 minutes before turning off. However, if for some reason you have a blackened bulb which will strike, in most cases simply burning it for 5 minutes or so will restore it.

Jerry Cotts
DP/LA


Jerry Cotts writes:

>Reality...Warn your operators to let the bulb run 4-5 minutes before >turning off.


This has been our experience as well.

Not allowing bulbs to warm up before shutting them off is not good for the bulbs.

We have gotten rid of all the Double ended 575s over the re-strike issue and now only have SE bulbs.

I'm told since DE 575s are no longer manufactured, the only place to get DE 575s is on eBay.

Even the newest 575s need some "off" period before re-strike. It seems to vary quite a bit depending on mfg of fixture. and of bulb and the burn time on the bulb.

Brian Heller
IA 600 DP


I'll chime in on this one since I don't believe that anyone has yet mentioned this point.

I think that the best indicator as to when to change an HMI lamp is apparent on visual inspection. Look for devitrification of the inner globe wall. If there's much more than a few specks of "schmutz" in the glass it's time to change lamps. These devitrified areas are weak spots that will absorb more heat (since they block the light). Uneven heating of this envelope due to these opaque areas contributes to (if not initiates) the failure.

I suggest doing a visual inspection of the lamp before you plug the system together. It's a great opportunity to look for other problems like bugs on the reflector or loose connections. My experience suggests that if the envelope is cloudy or heavily pock-marked, you will soon have an explosive failure. These lamps don't even start off looking pristine, but the more "schmutz" on the inner envelope the greater the likelihood of an explosive failure. Anyone else check the lamp condition before firing up the light?

Bruce Aleksander
LD/DP
ABC/Disney
Houston


Brian Heller writes :

>Even the newest 575s need some "off" period before re-strike. It seems >to vary quite a bit depending on mfg of fixture. and of bulb and the burn >time on the bulb.

I'm not so sure about this. Osram waits 15 seconds before re-striking on their tests and I believe this is the most "difficult" time for the bulb to re-strike. For some reason an immediate re-strike is easier...I think. As I wrote, my experience is that bulbs will not hot re-strike after 1-200 hours. With age apparently the electrodes become shorter and the gap therefore longer. Also various types of crud build up on the electrodes.

Because of these phenomena, perhaps more than the available 25KV is required when hot.

Jerry Cotts
DP/LA


My experience suggests that if the

> envelope is cloudy or heavily pock-marked, you will soon have an >explosive failure.

I don't disagree with Bruce on this, but bear in mind that the witches' brew of stuff in an HMI globe tends to condense somewhat on the inside of the globe as it cools and evaporates again when it heats up, so some cloudiness may not be a sign of trouble - It takes a trained eye or a lot of experience to see the difference between devitification and condensation.

I no longer claim either
Mark Weingartner
LA based but not there now



Copyright © CML. All rights reserved.