Homebrew PCBs

"Bertho Boman" <boman01@...> writes:
> Testing the photo resist by itself might be interesting but what
> really needs to be done is to test the system.  Add your artwork and
> any hold down glass or vacuum frame and then run an exposure test
> with the step gauge over real PCB traces.

That's my next test.  So far, I've tested my system with the LEDs,
glass, transparencies, and film I'll be using.  The only thing missing
was the ink, but that's how you calibrate the system - no ink.

> In-between those two extremes you will hopefully have a good section
> if the artwork is good enough.

That doesn't really tell you what the ideal exposure time is, though.
The film is spec'd for a given exposure in mJ/cm2 to cure, you need to
expose it for that long.  If your ink isn't UV-black enough, you need
to deal with it from the ink's side.  I'm hoping that my ink will hold
at least 10 more steps (the minimum), although I'll have to overexpose
it through the step gauge to find out.

The film does give you a range of steps (7-9 for mine), so you can go
up or down a step (1.4x the time) if needed, but if it's that close,
it's too close, and you need darker ink.

Example, if it works the way I'm thinking:

I've figured my ideal exposure to be 5 minutes to step 8.

I put a printed stripe pattern on, with the step gauge et al, and
expose for 5 minutes - the stripes should hold for steps 1..8, with 9+
uncured.  Recall that ink = no cure = no copper.  Whatever happens at
step 1 *is* what will happen without the step gauge, since step 1 is
transparent.

Now, to see how far *past* "ideal" the ink will hold, I increase my
exposure 16x, or 8 steps (2^(8/2) = 16).  Now we're at 80 minutes.
The parts of the film where there's no ink should now hold to step 16
(8+8).  Whatever happens at step 9 now, is what happens at step 1 with
the "ideal" exposure.

Let's say the results were that stripes showed from step 16 down
through step 7.  Since 7 is two steps below 9, that means that I would
have to increase my exposure 2x (from the ideal, so 10 minutes) to
risk exposing through the ink.  Not a big margin.

To picture it: steps 1..7 are copper, 8..16 are stripes, 17..21 are
blank.  The real board sees what happens at step 9, so there's only
one step below it that's also striped.

If the result is stripes down to step 5, that means I would have to
increase my exposure 4x from ideal (20 min) to risk exposing through
the ink.  This is a bigger margin.

To picture it: steps 1..5 are copper, 6..16 are strips, 17..21 are
blank.  There are now three steps below 9 that are striped.

What I want is whichever inks give me the greatest margin, meaning
they're blocking the most UV.  Black, cyan, toner, whatever.