John's homebrew pages
Making "optical" printed circuit boards
For a long time I had intended to make printed circuit boards
"optically", that is from a printed mask rather than a mask painted on
to the copper (the earliest technique I used) or drawn on with a
permanent marker pen (a later technique) or with cut out vinyl tape.
All these techniques had stood me in good stead, even for my first
boards for surface mount components (SMD), but for better quality
boards, especially for SMD devices with lots of closely spaced legs, I
realised that I would have to move to using an optical technique.
The simplest optical technique is to produce a mask at the same
scale as the final board, and to make an optical "contact print" onto
the photoresist. This was not really a new technique for me, since I
have made many a photographic contact print in my time - I started
those when I had a Kodak Box Brownie!
Especially as I had started to use the PCB package for drawing out
my boards (even for painted or vinyl boards), I realised that making
the print for the mask would not be difficult - I could print it direct
onto laser printer transparency material. Then at the Magnum Rally in
2010 I came across a UV lamp - it only has a little 4W "black light"
tube (150mm / 6 inches), but I thought it would probably be OK for exposing photoresist.
As I am retired I tend to try to do things as cheaply as possible,
so had decided to try using "spray on" photoresist as it's a lot
cheaper than the ready sensitised boards. However the photoresist can
relatively short shelf life (a year or so), so I don't yet know if this
is a false
economy! I looked at many web pages looking for best practice, but
couldn't really find consistent advice, so decided to "suck it and
see". I'm currently using Electrolube PRP spray-on positive
photoresist, and Electrolube PDN photoresist developer.
Before I made the first board, it was essestial to establish the
best exposure time under the UV lamp. To do this, I drew a simple PCB
pattern with several fine SMD device pad layouts on it, and copied it
several times. This was then printed out onto laser transparency
material ready for test exposures.
The first board (top board in photo below) was a bit of a disaster.
I sprayed on a thin layer of photoresist and left it to dry. Then I
used initial exposures from a minute to about ten minutes, mixed up the
developer at the right temperature, put the board in and developed
it for the recommended time (use the manufacturer's instructions that
come with the developer). Great! - it had produced what I expected, an
underexposed board at
one end and overexposed at the other - so I could work out the best
exposure. I hadn't taken a photo of it at this stage. I then put it in
the etch solution, again at the right temperature; I reckoned that as
that was wet as well there should be no problems. Wrong! The nicely
visible pads and lines on the board started to dissolve away, and I
pulled it out of the etchant before it was really finished. The result
was the odd board at the top in the photo - some of it looked OK, but
most wasn't - and I've no idea what the rounded patch is (thumbprint?).
I realised that on all the web pages I had read, instructions included
drying the board after development. So, for the second try, that's what
I did, using a hot air gun (though no hotter than I could cope with
holding the board between my finger and thumb). This came out much
better - see the middle board in the photo. I realised though that one
end of the board was not being exposed properly, as it was a bit
further from the UV tube.
The final test (bottom board in the photo) had the illumination
fixed, with exposures up to 20 minutes, and it came out really well. I had also realised by this time that
making a spray-on board needs just a single thin coat of the photoresist
(only about 2 seconds spraying - 4 seconds is too long) - it's best not to get it too thick.
So I was now ready for a real board. This is a design that includes
a small SMD chip as well as larger SMD chips and even a couple of DIL
devices. This photo of the board after development but before etching
shows how sharp the result is; it's best, of course, to print the board
layour as a mirror image so that the laser toner side of the
transparency is in contact wth the sensitised layer during the UV
exposure. However, you can see that I have made the coating a bit thick
towards the left hand edge of the board.
This is the same board after etching, and after a first bit of
tidying up the stripline on the left side of the board, which the thick
resist had prevented from being etched properly. You can see a few
other little blemishes (click to enlarge) that need tidying with a burr
in the rotary tool, but it's generally pretty good.
Finally, here's the board touched up and drilled ready for the
components. As a first optical board, I was pretty pleased with this.
With the basic technique sorted out, I have also tried using
pre-sensitised board, which is even easier to do of course, though more
expensive. Below is a photo of my first developed and etched board
using this stuff - I was a bit disappointed as it wasn't as clean as
I'd hoped, needing a bit of work with the burr. However the bits with
most detail were good enough so I didn't re-do it.
After a bit of work with the burr, and the holes drilled, it's fine
for use. There's a photo of this board with its components in place
near the top of this page.
So I'm now pretty happy with this process, and can't see me using
anything else for the more complex boards I need; though I will still
use vinyl tape where the board is very simple. I won't be going back to
enamel paint though, I don't think!