With Dave's permission (he said he would join shortly), here is an article from his web page.
It explains the issues arising from using fluorescent tubes (rather than point light source) and not having a vacuum table.
http://www.positivityltd.co.uk/Tutorial/Losing-Detail-Exposure.html (http://www.positivityltd.co.uk/Tutorial/Losing-Detail-Exposure.html)
--------------------------------------------------------------------------------------------------------------------------------
If you are using an array of black-lights or other diffuse light source you may find that your ability to print fine detail is restricted, here's why:
This is a schematic diagram of an exposure unit.
(http://www.positivityltd.co.uk/Tutorial/images/UVTubes1.gif)
The mask is in perfect contact with the flat screen. If there was only one small tube the system would approximate to a point source and the light would cure the screen so that the image is faithfully transferred to the screen.
Compare that to the effect of the light from the tubes at the extremes, they are going to undercut the mask and result in a loss of detail and closing of halftones.
If you do not use a vacuum the chances of good contact between mask and screen are nil, and that includes a piece of plywood and some bricks.
Once contact is lost the undercutting is made worse.
(http://www.positivityltd.co.uk/Tutorial/images/UVTubes2.gif)
* So point light sources are good, but if you have a point source then most of the energy will be wasted and result in long exposure times. Even commercial exposure units will make a reasonable compromise which leads to the exposure distance being about the same as the diagonal of your screen.
* Vacuum units are essential for fine detail and halftone printing.
* The screen in the area between point source and diffuse source will receive a variable amount of light which means that some may wash out other bits won't. You will have variable detail depending on your washout.
* Over-exposure will make the problems worse because more of the semi-cured emulsion will resist washout.
* I've shown a flat screen, screens aren't flat, they follow the contours of the mesh. Even with a good vacuum contact will not be perfect.
* Light travels in straight lines - except when it hits anything, the first thing that happens is light-scattering so that even a perfect system would have some undercutting due to scatter. Tinted mesh absorbs uv light and reduces scatter.
Thank You Dave!