Homebrew PCBs

Please see my previous message for a discussion of the desirable
properties of phototools, and about printing with a 1200 DPI laser
printer.  In the future, I will place an updated version of this
material at:

.  http://www.firstpr.com.au/pcb-diy/

Firstly, I will consider a direct laser-printed phototool on plastic
film which has diffused optical properties, like drafting film.  I guess
this material is polyester, but I am not sure.  The sheets I have here
are from:

.  http://www.megauk.com/artwork_films.php

Laserstar Film A4 sheets, packet of 10, UKP6.15 plus postage.

My HL-5250DN prints like a charm on this.  It has a matte finish on both
sides.

In order to reduce the diffusion of this phototool, I can do two things.
 Firstly, place the toner side against the object to be exposed, such as
the Riston layer on some circuit-board material.  (Perhaps even removing
the protective mylar sheet from the Riston, to reduce reflections -
however, maybe the Riston would bond to the toner and plastic film.)

Secondly, soak the whole arrangement in oil, such as cooking oil.  In my
preliminary trials, this does not upset the Riston.

Then, the light enters a piece of glass which is used to press the
phototool onto the PCB, but instead of exiting the glass into air, it
exits into oil, which has a refractive index much closer to that of
glass than air.  This means less light is reflected at the bottom
surface of the glass.

From the oil, it enters the top of the plastic sheet of the phototool.
Since this top surface is matte finish, this normally involves quite a
lot of sideways splaying of the light, which we are trying to avoid.
This occurs to a reduced degree since the refractive index of the oil is
closer to that of the plastic sheet, compared to that of air.

This is due to the angle of light being bent at angle surface between
two materials depends on the difference between their refractive
indices.  This doesn't change the shape of the irregular matte surface
of the plastic.  However, with oil and plastic, there is less bending
than with air and plastic for the same irregular surface.

When the light exits the bottom of the plastic sheet, it travels
straight into oil.  This again reduces sideways splaying of light
(diffusion), compared to the amount of splaying which occurs with the
light exiting the matte surface plastic into air.

Then the oil takes the light directly to the Riston, or to the mylar
film on top of the Riston (between this film and the Riston there is no
air-gap).  So there is a better matching between the refractive index of
the oil and that of the Riston, compared to having an air gap.

The result is more light hits the Riston (assuming the near UV
absorption of the oil is not a problem - better to use a clear rather
than a yellow oil, I guess) and that this light is in general travelling
more in a straight line from the light source than would be the case
without the oil.

The result should be reduced exposure times and a sharper image -
provided the light source is a narrow one.

Looking at the path of light from a green laser pointer, I find that
without the oil, there is a broad diffusion of light.  With oil, there
is a clear, straight, beam of light, with reduced diffusion.  I haven't
quantified how much light is in the straight beam, but I am sure this
arrangement will give a sharper image when used with a light source
which produces relatively parallel beams of light.

There's no point in trying this if the light-source is broad, such as a
tubular fluorescent or nearby compact fluorescent lamp.  Those broad
light sources will create a broad fuzzy shadow no matter how much we
reduce the diffusion of the phototool.

I find that for Riston, which requires near-visible "Near UV" light
around 350 to 380nm, that there's no point in using a specifically "UV"
light source.   I used to use a 400W mercury vapour street light with
its phosphor-covered glass envelope removed.  This is an extraordinarily
powerful light source and is very dangerous to the eyes.  It needed to
be cooled with a fan.  The biggest two problems were that its light
output varied with how long it had been on and with its temperature -
and that it could not be turned off and on rapidly.

Here is a much better approach.  For $16 I obtained a 500 watt
incandescent floodlamp, with a ~12cm long quartz halogen lamp, with an
aluminium reflector, like this:

.
http://www.jastimber.co.uk/products/electrical/lighting-floodlights/floodlight-portable-500w-fppsl500p-product.html

These lamps put out plenty of UV.  I used to have a process camera with
four such 500 watt lamps, with no protective glass.  My legs developed a
suntan!  That would be caused by much shorter wavelengths than the 350
to 380nm light to which Riston is most responsive.

I have one of these floodlamps (Arlec HL110 in Australia) with the glass
plate removed.  It exposes Riston rapidly at distances such as 60cm.  At
such distances, the light source is relatively narrow and I would expect
the benefits of reducing the diffusion of the phototool would be apparent.


An even simpler technique is to print onto plain paper and do as above:
have the toner face the Riston and soak the whole thing in oil.

My preliminary tests indicate that this works well too.  The oil reduces
the diffusion and reflection of the paper.  For repeated use, it would
probably be best to make a fresh paper phototool, whereas the plastic
sheet material referred to above could be re-used without fuss.

Also, the plastic sheet material would be more stable than the paper,
which might be important for larger boards.

I have only done preliminary tests and haven't looked at double-sided
arrangements.  For double-sided PCBs, the plastic sheet would be far
preferable, since the two sheets, oiled or not, could be aligned and the
PCB material slipped between them.  This would probably stretch or tear
the paper sheets if they were soaked in oil.

 - Robin