Hi Vicent,
Some of the eBay ads which can be found by searching for:
36W Nail UV Lamp Acrylic Gel Curing Light
show that they contain four long U-shaped fluorescent lamps. This ad:
http://www.ebay.com/itm/310556445104shows one of these types of lamp - and they have "365nm" written on
them, which is most helpful! The Riston General Processing Guide at:
http://www2.dupont.com/Imaging_Materials/en_US/tech_info/datasheets/index.htmlindicates peak sensitivity is in the 350-380 nm near-UV range, with some
sensitivity from 300nm (UV) to 450nm (blue).
I am not sure what wavelengths other photoresists are most sensitive to,
such as positive photoresists.
I think it is much better to have a relatively compact source of light
than something broad like these.
In some previous messages to this list, such as message 30440, I
described using a 500 watt linear halogen incandescent lamp with
reflector - a cheap hardware-store "floodlamp". Such lamps put out
enough near-UV to give me a sun-tan when I extensively used an old
process camera which had four of these lamps. Glass would reduce the
tanning/burning shorter wavelength UV rays to some extent but I think
would hardly affect the transmission of the 350-380nm light of interest.
The 500 watt lamp radiates most of its energy in the visible and
infra-red (with some being lost as hot air), and only a fraction of it
is the light we are interested in. I couldn't find a suitable source of
dark-violet-looking "Woods glass" to make a filter which would reject
most of the visible and IR I don't want. This visible and IR energy is
substantial and will warm the board, which might cause differential
expansion of the phototool and board. However, this can be largely
solved by blowing a fan on the board, with its phototool and glass or
other surface which clamps the phototool to the board.
This way, most of the light comes from an area about 10cm wide and 45cm
away from the board. These rays of light are relatively parallel, so
the shadow from the phototool remains relatively sharp even if the
phototool is not completely flat against the photoresist.
With a broader source of light, as would inevitably be the case with
these less intense, larger, fluorescent lamps, the shadow would be much
less sharp.
It is difficult to get the phototool sitting flat against the PCB. The
PCB is reasonably rigid and is not necessarily flat. If the top cover
is glass or acrylic, then any dust or mismatch between the flatness of
the glass and the PCB will cause a gap and loss of sharpness.
It would also be possible to use a vacuum frame with a soft, flexible,
top surface, such as clear PVC sheet (retailed at fabric and other shops
as a table covering), which would press the phototool close to the PCB
even if the PCB was not completely flat. This would also largely avoid
problems caused by dust, since it would tend to stretch the phototool
over the dust particle and so press it close even one or two mm away
from the particle, whereas with the glass or acrylic approach, the dust
particle would lift the phototool and glass off the board for a larger
surrounding area.
Another likely problem with fluorescent lamps of all types is that their
output would vary considerably with how long they have been on for,
which would be affected by their starting temperature and therefore how
long they had been on for in the last 30 minutes or so. This would make
it difficult to run them from a timer. This might be OK if you ran them
all the time and had a way of simply placing them over the PCB with
phototool etc. However, I prefer to place the phototool and my existing
glass sheet arrangement together by a safelight (a bunch of red LEDs)
and place it under the lamp, before turning the lamp on for a fixed
time. The 500W incandescent lamp would have its near-UV output affected
somewhat by mains voltage fluctuations, but I doubt if this is a
significant problem.
- Robin
http://www.firstpr.com.au/pcb-diy/