Homebrew PCBs

"Bertho Boman" <boman01@...> writes:
> I am a little surprised at the long discussions about selecting
> exposure for just the photo resist.  To me that is artificial and
> not real life.  To make PCBs, there needs to be an artwork and a
> hold down vacuum frame or glass.  Both will absorb some UV so that
> will affect the result and the "opaque" sections of the artwork are
> not really perfect, they will let through unintentional UV that also
> will affect the outcome.

If you had been following the discussion, you would have known that I
*included* those items in the test exposures, so I *am* compensating
for loss through the glass, artwork, etc.  That's the whole point of
calibrating with a step gauge - it tells you exactly what exposure you
need for YOUR SETUP.

> Further, there is undercutting by light creeping around the edges and
> development and etching issues.

That doesn't affect the exposure, that's something to account for in
the artwork.  I have a test pattern I use for doing that, and PCB has
a setting that lets you compensate for edge bloat/shrink.

> In an ideal situation, the variables should independently be
> measured but most hobbyists do not have all the equipment to do
> accurate measurements of the individual variables.

Um, yes we do.  The step gauge measures the process exposure, and a
test print measures the undercut and other edge conditions.

> Also affecting the result is how the light source is colliminated.

The LEDs have a 10 degree range.  We've talked about this before.

> Make a parallel test pattern of the skinniest lines and spacing
> required and long enough to match the step gauge.

I did say that would be my *next* test.  It's not the *first* test
because you have to calibrate the exposure first, which is done
without a pattern to interfere with your inspection of the "last held
step" (which is itself partially etched).

> Run an exposure test as the real boards will be exposed, develop
> just the resist, eyeball the result,

I've done all that.  Please re-read my posts.

> Also for the tests above I leave a little piece of PCB not covered
> by the artwork (punch a hole) and I also place a small piece of
> aluminum foil near the test pattern.  That will give you two
> references: A perfect opaque pattern and a perfect transparent
> pattern.

The step gauge does all that for you, and it's easier and calibrated.

> Since we are on this subject, I see little discussed about the artwork

Did you miss my post about how to determine the UV opacity of your
selected inks?