Homebrew PCBs

Don't bother with UK sunshine - unless you only want to
make boards for a few minutes on one morning a year.

Its far too unpredictable: your tests, and your production
pcbs are likely to be inconsistent.

Now, if you moved near the equator, or went for a holiday
up a mountain in the Azores or Canary Islands it might
be different - but you know they won't let you take your
chemicals on the plane :-)

Previous discussion here suggests that you don't need
far UV. (The fact that "UV" LEDs work tends to confirm.)

An ordinary fluorescent tube (or Compact Fluorescent?)
seems to have UV enough. Exposure time was fairly long
but bearable - of the order of 10min @ 10cm.

[There's more stuff on types of tube in the archive.]

When making an exposure test strip, be sure to give the
lamp a couple of minutes to warm up. Fluorescents don't
like the cold, so don't do the experiment in a very cold
place.

[For a production run, the warm-up time can be incorporated
into your standard exposure, but for your test strip it is
important that each exposure step gets the same intensity.]

Don't switch off between test steps either: if you can't quickly
uncover the next step, cover the lamp. A photographer would
use a density wedge or a sliding mask.

The exposure steps should be in a geometric progression,
not a linear one. This will not only give you equal increments
in step density over the linear exposure range but also span
usefully a greater range when the correct exposure range is
unknown  When making the exposure steps, the times
should be in the ratio: 1, 1.4, 2, 2.8, 4, 5.6, 8, 11, 16, 22, 32
which will give steps of "half a stop". Choose an initial time
so your best guess of the correct exposure time lies in the
middle of the range.

[You may have seen this range of numbers on the aperture
ring of a camera lens, where they give steps of "one stop",
halving in exposure because they measure F/diameter and
lens aperture area varies as diameter squared.)


Regards, LenW