Homebrew PCBs

At  9:08 pm ((PDT)) Sun May 27, 2007, bhleavi wrote:
>[attribution lost]
> > >btw - for collimation, my box is all white inside instead of
> > >aluminum foil. I think this works well to even things out.
> >
> > Evenness of illumination is not the same as collimation and
> > yours is diffuse illumination, more or less its opposite.
>
>Collimation is light just coming from the direction of the source.

Only if the light source is sufficiently far away that the rays may be
considered parallel within the accuracy required by the application.

[ http://www.google.co.uk/search?q=dictionary%3A+collimation

http://www.thefreedictionary.com/collimation
1. To make parallel; line up.
2. To adjust the line of sight of (an optical device). ]

For some purposes the Sun is such but if you need something
more portable, less energy-consuming and 24/7/365 it is usual
to use a compact light source and a collimating device.

This might be as simple as a long tube but that is inefficient
so it is common to use a focusing device such as a lens or
curved mirror.

A collimating lens gives a highly parallel beam but the edge
illumination falls off and target area is limited by the aperture.

A parabolic mirror is easier to make for large apertures but the
edge illumination falls off even further and direct illumination
from the source may need attention.

On the other hand, the illumination inside a hollow black-body
radiator is extremely uniform but totally un-collimated.

A lamp in a white box is somewhere between these extremes.

A commercial PCB exposure device I saw on the web had an
row of parallel discharge tubes covering the same area as the
stage but spaced well above it and behind a baffle grid so that
there was no squint illumination of the stage. Thus the
exposure was effectively an array of overlapping beams from
discrete light sources each far enough distant to approximate
a parallel beam. A rear reflector could have been used, I don't
know, but lighting efficiency isn't a big issue here.