UV light bulb for photo-etching PCB

[milen]

Do I need A, B or C spectrum of UV light?
If I read right both white and black fluo UV bulbs are in A spectrum, where UV C fluos have clear glass - which one is better?

Can I use scanner glass schield, or I need special UV emitting glass?

PS Can I use 2x x W balast and single starter and connect by this diagram?

 

[Resqueline]

Photoresist seems to be sensitive in the 340-420nm UVA wavelength range, so ordinary blacklight tubes can be used.
Scanner glass is relatively thin and not very green so it won't affect the wavelengths of interest very much.
You can't connect fluorescent tubes in parallel, they have to be in series, and you have to use two series-starters - one for each tube.

 

[milen]

on this site I found graphs for A, B and C spectrum of UV light. There are two fluos in A spectrum - one with white coating and one with black. Which one should I use - or isn't significant?
I also found some special device for PCB uv radiating that uses PHILIPS TL8W/05 lamp, and I think this also is in A spectrum.

Beside I would like to know how dangerous is this light if you accidently turn on the switch when hatch of the scanner isn't closed. Do you instantly feel any pain in the eyes (like watching the sun) when watching so you know that you shouldn't watch it, or pain is formated after a while when you allready have damaged eyes?

PS will this be the correct wiring? Ballast has two paralleled inputs and two paralleled outputs so basicly I understand that wiring is the same as having two ballasts, one for each lamp. And I found a wiring diagram here

 

[Resqueline]

That site says "BLB tube Peak: 352nm or 368nm" (dotted line) which means they don't know what they're talking about/ selling..
Just like the white looking tube the black looking tube has a phosphor with a peak emission at 352nm (and a range of 295-425nm).
Then for the "black" tube it says "Deep blue filter glass absorbs visible light". That's either Wood's glass or just an enamel coating, having a peak transmission around 365nm. Thus the BLB tube gets its peak output moved to 368nm (and a range of 328-408nm). It'll be "safer" to look at but will diminish the wavelengths of interest also.

PHILIPS TL8W/05 tube has its peak at 365-370nm (but ranges 300-460nm)
Datasheet/spectrum: www.polychrome.nl/file_download/18/Philips-05.pdf

All these lamps will work but you want a(n unfiltered) peak as close to 380nm as possible for maximum efficiency, so the TL/05 seems better suited.
A 100W incandescent bulb will do the job too, it just takes a little longer.

Eye damage is like for a sunburn. You don't feel anything while it's going on. It takes a while to build a damaging dose and so any pain will come when it's already too late.
I believe we're talking minute exposure here rather than second.

I don't think that circuit will work either, it would require two separate coils & cores.
Connect the tubes like at the end of page 4 here: http://www.capakor.com/.../(37)General_technical_information_of_fluorescent_lamp_starters.pdf
Use St 151 series starters.