UV light and damage to a CCD? (Ping Phil Hobbs)

I believe certain types of CCDs can get damaged, I'm sure google can tell you but, - won't the pictures have to take through a welding mask or something to get any useful pictures?

-Lasse

The main damage mechanism for CCDs is thermal, and silicon is a really great heat conductor. You're unlikely to bleach or photodecompose anything in a brief exposure like that anyway--CCDs are fine with direct sunlight focused down on them.

However, arcs can be several times brighter than the Sun, and of course any light bright enough to damage the CCD will completely blow out your retina.

Also, really bright sources will cause the CCD to bloom and spoil the image. A CMOS imager is quite a bit better because the excess charge can't easily spill out into neighbouring pixels along a row, so although those pixels will be saturated, you won't have ugly vertical white lines in the frame.

So the net is to use the same sort of sunglasses for the camera that you do for yourself (if it's a CCD), or something a couple of shades lighter (if it's CMOS).

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal
ElectroOptical Innovations
55 Orchard Rd
Briarcliff Manor NY 10510
845-480-2058
hobbs at electrooptical dot net
http://electrooptical.net

Indeed although they tend not to like it so much with a larger lens or telescope. Eventually the thermal load gets to them.

Which is probably the thing to remember in a welding environment. Human eyes and strong uv sources really do not get on and you won't feel a thing until much later.

Excessive UV may bleach the least stable of the colour filters on the Bayer mask which in the regions where the bright spots are focussed and could alter the colour balance permanently. The CCD silicon is undamaged but the organic dyes get bleached. I doubt you will harm a monochrome device. Bleeding into the columns would occur though as Phil has described.

An ordinary UV/Skylight filter will go a long way to protecting the sensor from any ill effects. The lenses being glass don't let much short wave UV through to begin with.

Preferably something a bit closer to optically flat than sunglasses or so thin that it doesn't matter. ND8 filter or aluminised melinex sold as photographic solar filters by Baader observatory would be ideal.

Regards, Martin Brown

Well, the idea that your eye will be blown out pretty much suggests a neutral density filter and an eyepatch (har de har har) for the eye not using the viewfinder is in order.

But UV light is also a contributory factor to cataracts, so I think a good UV filter (Schneider 415nm) and ND2 would do the trick.

t h s

d.

u r

Ah, I was thinking of the silicon, not the filters in the camera. Now this could be a problem worth mitigating.

She won't get anything - the arc is brighter than the sun. At best, the frame will be washed out (overexposed by 1000x), and at worst the IR (from the arc) will burn a hole in it. Shoot it through the window from a welding hood.

Hope This Helps! Rich

Finally !! An interesting (non-spam) thread on SED.!!

I'm what I would consider a "semi-pro" when it comes to photography. An MIT-Grad friend of mine (Electromagnetics), and himself a multiple blue-ribbon prize winner is mentoring me on how to take better photos. So, I'm always interested in weird little photo projects. Shameless plug: See some of my photos at:

Now, I don't plan to go photographing TIG welders anytime soon, but I do have a couple comments / questions?

Haven't we seen welding on TV before? (Like in all those horrible junior high school career counseling/planning videos they used to play?) And then there are the movies: U-571 & The Sum of All Fears come to mind - where the submarines are getting welded in the yard. Hey, what can I say.... I like submarine movies.

Next - wouldn't a UV filter help? It might fool the camera's metering(?), but you can always switch to Manual mode.

Next - or how about one of those cheap disposable cameras? If you fry that... who cares?!

As for exposure, it might be best to take two images and overlay them. One to expose the surroundings, and other to capture the arc. The two can then be combined either in-camera (like the Nikon D-90), or using HDR (High Dynamic Range) post-processing, or maybe even PhotoShop. But if two (or more) separate images can't be taken and later recombined, I guess this option won't work.

-mpm

Oops... I meant Hunt for Red October (not Sum of All Fears). Got my Clancy novels mixed up there....

Eproms, with quartz windows, aren't damaged by extensive UV exposure.

John

Silicon can stand an enormous amount of abuse by strong light and even hard UV. It will erase eproms but that is all. Organic dyes and human eyes are by comparison rather fragile in that UV light environment.

It is worth having a sacrificial glass filter over the front lens element to avoid problems with sparks and slag from the welding hitting the optics. You can always scrap a filter - lenses cost a lot more.

A DSLR probably would suffer no real ill effects because the focal plane shutter keeps the sensor in the dark except for a few short exposures. A video camera or P&S with continuous imaging onto the CCD and an LCD viewfinder might suffer some problems if you were unlucky.

Regards, Martin Brown

m

I think Rich is right.

I just looked at the exposure programming for the Nikon D-90. Even at f/22, the shutter speed is going to be very high. -17EV, or

1/2000th shutter speed. (Assuming 50mm, ISO=3D200 here.) This will practically ensure only the arc gets exposed. The rest of the frame may be pitch black. (?) Of course, these values will change with UV and neutral density filters added - probably to the point where the camera metering won't function anyway. Which would render my above observation moot.

I think you'll have to photograph this in Manual mode. UV Filters are cheap compared to some of the more expensive camera bodies.

Mental note: I wonder how much UV protection the lens provides?

Many people in the European laser show industry, where audience scanning is allowed, (if you watch your MPE levels), can tell you that a lot of CCD and DLP devices can be damaged by focused visible laser powers as low as 50 mW. I've personally nuked vidicons with 5 mw

633 nm light, when well focused. CCDS do a lot better then PbS, but..... Color CCDs have a lower damage threshold.

If you want a accurate level for the damaged thresholds, call up Spiricon, who make CCD based laser beam parameter measuring sets. Spiricon is unique in that you can use any CCD with their product. They will know exactly when to stop and add ND filters.

Besides, long before you get to the damage threshold, you will saturate some non imaging part of the CCD, and you will see whole columns or rows go saturated.

Because of the differences in focusing coherent light and a arc, assume you can be perhaps a order of magnitude higher with the arc then the laser.

Steve Roberts

Hmm, my Canon S1 IS comes with the warning: "Do not aim the camera directly into the sun or at other intense light sources which could damage your eyesight." Just covering their asses?

Excessive contrast certainly screws it up though.

Tim

I can't comment on optical effects. However, she should mind the EMI from the arc. Unlikely to damage the camera but if it's one of those consumer-grade plastic things it could upset the electronics in there.

When I did defibrillator discharge tests on my designs I managed to upset LANs and one time a PBX system quit, needed a re-boot.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

Colour CCDs are used extensively outdoors. The integrated UV dose from that is likely to be quite a bit larger than that from a relatively brief session photographing welders. On the other hand, the dose is probably a bit more uniform across the CCD on average.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal
ElectroOptical Innovations
55 Orchard Rd
Briarcliff Manor NY 10510
845-480-2058
hobbs at electrooptical dot net
http://electrooptical.net

A 5 mW beam focused down really small is much brighter than any nearby thermal source. At NA=1, a 633-nm beam will have a spot diameter of something like 400 nm, which when focused on some strong absorber embedded in silicon would raise the temperature by something like (5 mW/400 nm)/(150 W/m/K) or 80 C. (3-D heat conduction is pretty good.)

Did you do a postmortem?

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal
ElectroOptical Innovations
55 Orchard Rd
Briarcliff Manor NY 10510
845-480-2058
hobbs at electrooptical dot net
http://electrooptical.net

Hopefully no one was expecting to get decent arc welding photos or video without a suitable filter (making the issue of CCD damage moot). By far the best result is obtained with monochromatic illumination and a narrow bandpass filter, see US patent 4,225,771. I have seen video of MIG welding taken with this method where the arc is minimally visible and the droplets of molten metal in the arc and the puddle surface can be seen clearly, and the same method has been used for remote TV monitoring of automatic orbital TIG welding in radioactive areas. Probably too expensive for hobby use, so I suggest trying a standard welding filter of about the same density as the operator is using or a couple shades lighter (perhaps a 10) and lots of background light; direct sunlight is not bad, or a pair of 500 watt QI floodlights a few feet or less away will provide decent background illumination through enough filter to make the arc observable without complete washout. This will not let you see through the arc to the puddle surface very well but will allow the arc, torch, workpiece and filler rod to be seen fairly well. (I use a pair of 500 watt QI floods on my welding workbench when I want to see small parts clearly through the welding filter with or without arc, or when it is cold.)

I should point out that most "UV" filters for cameras don't filter much UV. I have a 385nm UV flashlight. Besides hunting for scorpions in the desert, you can light up some rocks, fungus, etc. [Allegedly s**en stains if you want CSI.] Two items laying around the house/lab are uranium glass and old style analog scopes with green phosphor. [Everyone has uranium glass on hand, right?] Both glow green with this flashlight. You put a TIffen Sky-UV between the flashlight and the uranium glass or phosphor, and nothing happens. Now I have an Andover

420nm low pass filter, and it just kills the UV. I'm guessing the Scheider glass (B+W) isn't in the liars club and actually filters UV.

Making a trip to the high quality tool store, Harbor Freight (cough cough), the welding glass said #10 on it. I don't know if that is like a ND10, but the glass is damn dark.

I showed it to a student, who waited till I left and burned his initals into the whole field of view. With a indium seal on the end of the CCD, baking it for recovery is out of the question. So it got trashed. RCA early gen security grade vidicon.

Steve