Help Needed: UV filter for dental curing light (setting tritium)

[archer6817j]

Hey all,

Not quite sure where to post this but I think setting trits counts as flashlight related materials...?

So I bought a UV dental curing light off eBay because I was going to simplify my life...and tritium installation. Figured the high intensity would cure Norland in a jiffy and I wouldn't have to leave my parts out in the sun for an hour to set each trit.

I got it all cleaned up and put down my first test blob of Norland 61 and...it didn't cure. After some hard googling, I found out this older model dental lamp puts out a different spectrum than I need. The light looks nice and blue but it's something like 400-500nm. I let the light run for several minutes and the Norland didn't even begin to cure. I have a (weak) 365nm LED light and I put that on the Norland just to make sure it actually worked and the Norland set up fine, so it's not the Norland, it's the light. The LED light is waaay to slow so that's why I'm not using it.

Okay, no problem. I went to Edmund optics and ordered an 85% transmission 365nm bandpass filter. It's supposed to take visible light and block out everything but 365nm (give or take a little). My dental light has a halogen source. Looking inside the little cap that came with the light I can see a filter that looks exactly the same as the one I bought. Perfect. White light in, blueish UV light out. However, when I unscrew the old filer and stick my new bandpass filter in front of the output on the dental light...I get almost no light at all and it's not remotely the purple/blue of 365nm UV...and it has no effect at all on the Norland. The filter has an arrow that shows the direction of light. I tried flipping it around. Nothing.

It seemed pretty straight forward. Remove old 450nm bandpass filter. Install new 365nm bandpass filter. Cure Norland.

Anyone have any experience and/or thoughts on why this doesn't seem to be working?

 

[MikeAusC]

Filament lights just don't produce a lot of UV light i.e. <400nm. The filter only eliminates non-UV light.

You can get 5 watt 365nm UV LEDs, but they cost $70.

 

[precisionworks]

You bring up an excellent point about the dental cure lights - the small disc filter targets the 420-480 nm wavelength & NOA61 needs light in the range of 320-380 nm with a peak sensitivity around 365 nm (info at http://www.norlandprod.com/adhesives/noa 61.html )

My light also would not cure Norland ... until the filter was totally removed. Now it cures the adhesive so quickly that the tip of the fiber optic wand has to be held at least 1" / 25mm back from the trit surface. Watch a dentist or dental assistant use this type of light & they hold the tip almost in contact with the tooth but (with the filter removed) that super heats the Norland. It took a few mess ups to figure this out.

Try your light with no filter & watch the Norland cure (I wear a welding helmet with a magnifying lens taped inside). Hold back at least 1", trigger the light, listen for ten beeps (100 seconds of light). You'll have a semi-solid precure or partial cure. Finish all the trits at 100 seconds on each one & then either sit the light aside for one week or final cure at 50° C for 12 hours to reach optimum adhesion. Most of my customers want their lights back yesterday so I final cure under an IR heat lamp - monitor the temp to make sure it stays around 50° C.

My not too fancy setup uses a light tripod that once held a halogen painter's light. The IR lamp costs about $5 & the socket/housing is another $5. As long as you remove the filter, run the 100 second per trit pre-cure & final cure at 50° C the dental curing light will work perfectly.

Since no member ever indicated much interest in the dental curing light I didn't update the thread to discuss removing the filter - my bad for not doing so.

 

[archer6817j]

No filter!? I guess that's obvious in a certain way 🙂 I'll give it a try. Thanks!

 

[TEEJ]

I'm pretty sure the NON-UV light is not KEEPING it from curing, so, a filter that only lets through what's already there, that is needed to cure it, won't make it cure faster.

 

[precisionworks]

A halogen incan bulb (without a UV glass front lens) puts out a substantial amount of light in the 365nm range. To prove this, put a drop of NOA61 in the middle of a finger nail or thumb nail & hold the probe about 1/4" above the Norland. Guaranteed you will move either the curing light or the finger in less than 10 seconds & the Norland will already have skimmed over. Screaming like a girl is not allowed as it shows bad form 🙂

Merc vapor lights (without the glass envelope) put out many times more UV at 365nmn but are a pain as the glass envelope has to be cut or broken away.

 

[smflorkey]

Or, if you have a Surefire A2, take a look at the UV LED rings offered by calipsoii. He says they emit 375nm which should cure Norland well. I haven't yet used this adhesive, but I have the UV ring for my A2 and find the workmanship very good.

Hope that helps,
Steve

 

[wquiles]

A halogen incan bulb (without a UV glass front lens) puts out a substantial amount of light in the 365nm range.

So would the output of a bright flashlight, like the SF M6 with the MN21 (600+ Lumens) be enough to cure Norland fairly quickly?

 

[precisionworks]

The Norland site says that the recommended energy required for full cure is 3 Joules/sq cm. Based on non-scientific observation the dental curing light has that energy level at around 1" distance & a much higher energy level when the probe nearly touches the Norland. Not sure about the MN21 incan - it surely contains some UV component but my guess is that the UV energy is pretty low. Some members say that bright sunlight works well but it contains a high UV component.

 

[CKOD]

You guys are overcomplicating this 🙂 A $5-10 black light bulb (CFL or incan) works fine. It sets the NOA in 10-15 seconds, and just for good measure I let it cure under it after I'm done for an hour or so.

Also, the dental curing light has a light pipe/guide of some sort I presume? What is it? If its plain glass, 3-6" of transmisson may knock out all the 365 nm light even with the source filter removed.

 

[precisionworks]

the dental curing light has a light pipe/guide of some sort I presume? What is it?

It's a fiber optic bundle, optically polished on each end, and rotates 360° making it easy to get a comfortable grip on the gun. These curing lights use a 75 watt halogen UV specialty bulb & a reflector that focuses the light at the start of the fiber optic probe. There's enough 365nm energy at the end of the probe to boil Norland if the tip is held close to the trit. The microscopic bubbles grow with heat & start rising to the surface, only to be trapped as the surface hardens over. Try it 🙂

sets the NOA in 10-15 seconds, and just for good measure I let it cure under it after I'm done for an hour or so.

Norland calls the initial setup a pre-cure, meaning that the adhesive has not reached full adhesion to glass. They state that NOA61 sets up fully in 7 days at room temp or in 12 hours at 50° C.

The Norland at the very top of the slot sets up first as it receives the highest dose of UV. The adhesive below the top may range from semi-solid in the middle to liquid at the bottom of the slot, especially on the deeper slots like 2.0mm or 2.5mm. The 1.5mm installations cure many times faster but I don't feel comfortable touching or handling the light until the 12 hour heat cure is over.

 

[archer6817j]

Well here's a funny story. I took the filter off my dental light, put on my welding hood, hit the button for 30 seconds....annnddddd....nothing. Norland is still totally wet. I hit it for another 30 seconds and no improvement.

I've had the box of the curing lamp apart and it appears to have a normal halogen bulb...or not. The only thing I can think is someone replaced the bulb with the the wrong kind, but I can't see why that would fit or even be available. I'm going back to the shop now to check out the bulb and get some part numbers off it.

Interestingly, when using the filter the light came with, it produces a bright blue light. When I put on the 365nm filter it produces no light at all. Combined with the above, could that seem to suggest there is no 365nm light coming out? Seems strange since it IS a uv curing light.

I know I could do a bunch of other stuff but I thought I'd have the "perfect" solution for just $100 bucks...so I'd like to make this thing work rather than scrap it. Space is also a bit of a premium on my bench and it's a lot nicer to have a little wand I can wave around rather than set up some giant contraption with black lights etc.

eBay Item number: 110797673039

 

[PhotonWrangler]

I think you've got the wrong bulb. If it's a standard halogen socket then chances are the previous owner swapped the bulb for a less expensive one with a different envelope and gas fill. I think you're better off retrofitting a CFL blacklight bulb in there. Less heat, lower power consumption and lots more UV output.

FYI here's a note from a GE Lighting FAQ -

8. Does an MR16 halogen lamp provide much UV?
Tungsten filament lamps, such as halogen and incandescent, provide minimal UV. GE's ConstantColor® MR16 lamps are made using special quartz, which has properties that enable it to filter out nearly all of the UV portion of the spectrum.

So the chances are your EFR lamp is designed for visible light output with minimul "real" UV <400nm.

Here's an MR-16 drop-in that will get you into the 395-400nm range. It should be compatible with your existing power supply.

 

[Al Combs]

Interestingly, when using the filter the light came with, it produces a bright blue light. When I put on the 365nm filter it produces no light at all. Combined with the above, could that seem to suggest there is no 365nm light coming out? Seems strange since it IS a uv curing light.

I looked at the link in your first post for the interference filter you bought from Edmund. The full width half max wavelength is a 10nm wide window. Half above and half below means if peak transmission is 85%, there is only 42% transmission at 370 nm and drops off rapidly after that. Since the average person's eye can't see anything below 400 nm, my guess would be the filter is working.🙂

Seriously though as TEEJ pointed out, the filter doesn't amplify 365 nm, it screens out everything else. Are you outside your 30 return window? Those little suckers are expensive.😱

 

[Al Combs]

Is the dental light you are using the one in the eBay link? If you'll forgive the Wiki reference on their Dental Curing Light page, "The dental LED curing lights use LED’s that produce a narrow spectrum of blue light in the 400- to 500-nm range (with a peak wavelength of about 460nm), which is the useful energy range for activating the CPQ molecule most commonly used to initiate the photopolymerization of dental monomers."

Barry mentioned that Norland's optimal cure wavelength is 365nm. If the dental light is optimized for a 460nm wavelength, perhaps it's the fiber optic wand and not the bulb causing the problem. Can the fiber optic be removed so a sample of Norland 61 can be put in the beam as a means of testing the fiber optic?

I used to work in the inspection department of a company that sold among other thing the interference cubes used in fluorescence microscopes. They typically used short arc 100 watt Mercury bulbs or the 75 watt Xenon depending on what exciter frequency they needed for the dye used with the sample. But all of the cubes they sold also worked with an ordinary 100 watt halogen lamp housing. They were just much dimmer. Quartz has excellent UV transmission characteristics. I doubt very much if the bulb is the problem. Bulbs with a UV block coating must be more expensive than a standard bulb that would work just fine.

 

[archer6817j]

Yeah, the filter I got costs as much as the light 🙂 The light came with a filter that screws onto the tip of the wand. I realize a filter doesn't amplify (because it's a filter 🙂 , but if I only need 365nm...then it seemed like a good idea to filter everything else out.

I appear to have this bulb.

Just got back from the shop. I put a drop of norland on some paper and exposed it directly to the halogen bulb for 60 seconds and it seemed about 75% cured. Then I used the fiber optic wand (no filter) and no curing at all.
Then I put the bandpass filter directly in front of the bulb, in the opening that normally holds the wand, and put a piece of paper in front of that. After 60 seconds it was a tiny bit cured, but the light is very diffuse. So I'd say the halogen bulb works, and the filter works...but somehow the actual wand is blocking the UV? There is a boat load of visible light coming out. I don't quite understand that since the whole thing is built for the specific purpose of delivering UV light.

 

[Al Combs]

but if I only need 365nm...then it seemed like a good idea to filter everything else out.

Except that depending on the lamp housing, visible light might serve as a reminder it's still turned on.

I don't quite understand that since the whole thing is built for the specific purpose of delivering UV light.

Considering that 460nm is ordinary violet and not ultra violet light, it's probably a familiar if not technically correct industry misnomer.:shrug:

The money from the returned Edmund filter will go a long way towards the purchase of an Eprom Eraser of the right size. Imagine an ordinary fluorescent tube with no phosphor. Then imagine a quartz envelope to allow the UV the phosphor would normally absorb to pass through... We could always tell we left it on by the light 'bleeding' though the edge of the drawer.

 

[precisionworks]

Welding structural steel today please excuse the short answer - you need the UV curing bulb in your light :nana:

The UV bulb puts out lots of UV across the spectrum & the small disc filter (that you removed) filters the light so that only 400-500 nm waves pass through.

Sounds like the ends of the fiber optic wand may not be perfect - they should look like mirrors. Polish with Semichrome if they don't.

 

[PhotonWrangler]

The money from the returned Edmund filter will go a long way towards the purchase of an Eprom Eraser of the right size. Imagine an ordinary fluorescent tube with no phosphor. Then imagine a quartz envelope to allow the UV the phosphor would normally absorb to pass through... We could always tell we left it on by the light 'bleeding' though the edge of the drawer.

Eprom erasers use germicidal UV lamps with a much shorter wavelength, somewhere in the 245nm range. If you need this short of a wavelength, you can go to Target and pick up a GTL3 replacement germicidal bulb for their air purifier. Find an intermediate-base socket for it, wire it in series with a 40 watt bulb for a ballast (or a 15w magnetic fluorescent ballast, same results). Keep in mind that shortwave UV is REALLY bad for your eyes and skin, so you'll need to make an appropriate shield for it.

 

[Al Combs]

It's true that germicidal bulbs are used for the 253.7nm wavelength needed to erase the eproms. But the Mercury vapor in the germicidal bulbs also has a spectral emission line at 365 nm that would make it ideal for curing Norland. Two downsides to an eprom eraser are the cost and the possibility the drawer not being large enough for the flashlight to fit. You could make your own setup from say a regular 15 watt fluorescent fixture and a germicidal bulb. The germicidal bulbs are G15T instead of F15T. But as PhotonWrangler mentioned, there is the danger of Ultraviolet-C you have to take into account. You'd need to make an enclosure of some type that would completely shield the light but be large enough that it doesn't overheat. Perhaps a large cardboard box might be a workable solution for intermittent use. Also I'm not sure if germicidal bulbs use the same ballast as a regular fluorescent bulbs. Black lights would eliminate the danger of exposure to Ultraviolet-C. But there are so many different types it would be difficult to tell if you had the right one.