PCB exposure box - bulb burnt. Options for replacements

[DGM]

From the picture, I'd venture to guess it's a F12T8BL lamp. Not uncommon for institutional insect killers. Yes, ordering fluorescents online is usually costly. They either want you to buy carton quantities or they have to deal with a ton of packing and high risk of breaking in transit.

One way you could plausibly do an A/B test of cure rate between the BL lamp and an off-the-shelf CW lamp would be to slap a couple cardboard dividers between the lamps. Done that way, a board that straddles the divider would be exclusively illuminated by one lamp on each side.

Not that they would fit in your enclosure, but anything with the BL suffix would be a good start. Everything else is down to getting sufficient and uniform illuminance.

Everyone is poo pooing on clear germicidal tubes, so I'll share my own build. My curing box uses the naked quartz arc tube out of a 175W mercury vapor lamp. When you're putting out enough shortwave UV that you need a ducted exhaust system, you need to be careful.

 

[mik3ca]

Ok so the standard white bulb this time around wasn't worth it I don't think. I literally had to add another tablespoon of MG chemicals "developer" and run the outsides of the container through hot water just to make the board from 10% satisfactory to 75% ok. Nevertheless I did scrape the extra unwanted tracks with a utility knife.

I found the following listing on ebay advertising "insect killer" lamps which a guy online by the name of Mike recommends using on his PCB page.

Here's the listing:
https://www.ebay.ca/itm/6x-8W-T5-Ul...nsect-Fly-Killer-Mosquito-Zapper/322199166926

And the code for these lights contain BL.
Would they work?

 

[DGM]

Do the original lamps truly have no markings? Is the F8T5 size correct?

I would think that a common BL lamp would work. BL/BLB tubes peak around 350-370nm. While there are other phosphor blends (that won't be marked BL/BLB) that have different peak wavelengths, I kind of doubt they'd use anything special for this application. F8T5 BL and BLB lamps are super common. 1000bulbs.com sells them for about 2.70USD. I've bought them at the local store as well.

I don't know that a BLB lamp would work well though. I have a feeling that the visible light fraction from a BL lamp would be a non-negligible contributor to the curing, and filtering it out would only reduce the effectiveness. That said, BLB lamps in the F8T5 size are going to be more common in retail.

If it is something special, there should be some extra info on the lamp. Either a suffix to the lamp designation or a code or a peak wavelength spec, depending on manufacturer.

 

[mik3ca]

So the full part number ebay puts out for this light is F8T5368BL and in the description it is "T5 368BL" Does the 368 mean it outputs light only at 368nm?

 

[mik3ca]

Oh and I checked stores for such lights and all they can provide are curled lights or double bar lights but not the ones I need for my exposure box.

 

[DGM]

F8 = 8W
T5 = 5/8" diameter
368 is the peak emission wavelength of interest -- dictated by phosphor type
BL = emits UV + visible light (tube looks white)
BLB = uses doped glass to block visible emission (tube looks purple)

 

[mik3ca]

F8 = 8W
T5 = 5/8" diameter
368 is the peak emission wavelength of interest -- dictated by phosphor type
BL = emits UV + visible light (tube looks white)
BLB = uses doped glass to block visible emission (tube looks purple)

Thats the most helpful piece of info I found on the Internet so far. 368 must be perfect for photoresist because I think the optimal wavelength for UV exposure is between 350 and 450 is it? I couldn't remember the exact range but I think its in that ballpark

 

[DGM]

UVA and near-UV lamps can be found from the low 300nm, all the way up to a royal blue 450nm, but the BL/BLB lamps are either around 350 or 370. If i'm not mistaken, the ones you'll tend to find currently are around 370. My understanding is that the common 350nm phosphors are all lead-doped, and probably just run afoul of RoHS restrictions.

I don't use any dry film resist or coated boards, so I don't really know what they'd work well with. At a glance, it seems most of the references for PCB resists are sensitive in the UVA range.

To differing degrees, a particular product may be sensitive to a relatively wide range of wavelengths. While you might hope to find a perfect match for the old ones, chances are they're significantly worn anyway. If you replace the whole set, you'll probably have to adjust your exposure times anyway.

 

[throbscottle]

I would replace the tubes with LEDs. I made a UV light box some time ago using these, a pack of 100 cost £3.58 on eBay in 2016. Mounted on 2 pieces of 10cm x 22cm perf board, connected in 3 series chains (I think I used 96 LEDs in total, so 32 per chain), each with it's own CC supply, all done very cheaply in a wooden box. I later had a conversation on here (with Mike I think but I can't find it now) that series resistors alone would probably have been adequate. Anyway I like that it's regulated since it takes out variance in mains voltage using a clipper circuit. Very efficient.
As mentioned above, these are the same tubes used in insect-electrocution devices, definitely not the same as "black light" tubes.

 

[DGM]

I don't really think a LED conversion is warranted, unless you're building your own or cannot obtain the correct lamps. Even assuming the wavelength sensitivity isn't an issue, you'd still want a comparable total radiant power and a uniform distribution. I'm assuming if we're talking about 100x LEDs purchased on ebay, we're talking about 5mm 15-30 deg LEDs.

Using 5mm or other lensed LED's in a shallow box is going to present difficulty in getting uniform illumination. Adding a diffuser or using diffused LEDs reduces total output.

Again, disregarding the SPD differences, a cursory survey of the specs for legit 5mm LEDs (395-405nm) gives a radiant power of ~25-50mW. A single F8T5BL puts out about 1.4W UV. You'd need about 200-400 such LEDs for a comparable nominal power, and that's assuming that the ebay LEDs are comparable to the legit parts in output power and won't degrade quick enough to give the longevity advantage back to the fluorescent option.

 

[throbscottle]

All I can say is I need a 24 second exposure for my boards, so power isn't an issue. Quality might be but I've had no problems yet. I admit I do use quite a deep box so the illumination is very even.

 

[DGM]

I suppose the power really only matters if one expects to maintain an equivalent exposure time. I don't use photoresist for etching, so I don't really know what to expect of it in terms of speed. I use my UV rig for photopolymer curing, and exposure time is 10-20 minutes depending on the geometry. I wouldn't want to have an exposure time like that doubled or tripled.

 

[throbscottle]

Ah, different kind of goat then. Good luck with tracking down those tubes!