Homebrew PCBs

Chris,


Laser-printer scaling problem:

Edit the PDF file.  The PDF is just a big text-file, and the
scaling
parameters are right there at the beginning.  You can adjust each
axis
to suit.  There's tons of info on the web about PDF language and
how to go about doing things like this.


Exposure and Density:

If it's taking 15 minutes to expose your board, then your
light-source
is MUCH too weak.  The longer the exposure, the fuzzier the
image;
because the UV also moves -sideways- through the resist; not just

straight into it.  Not only do your traces get fuzzy, but a
pinhole that
might make an easily etched micro-dot turns into a 'spot' that
doesn't
want to etch away.

Also, the molecules being exposed/crosslinked by the UV always
affect
a few molecules nearby....so the longer the exposure, the more
nearby
molecules that get randomly added to the crosslink as well.  This
too
spreads out your lines.


Bottom-line:  the faster your exposure, the junkier the artwork
that
you can get away with using (i.e. poor density etc.).


As far as printer-density goes, laser-printouts have always been
marginal,
in my experience.  I use them; for non-critical, undemanding
work; but
if it's anything important, or fine-line, I will send Gerber
files to the
local film-house.

Photoplot films are usually only 5 or 10 bucks a layer; for 8x10"
size.
15 bucks is worth it for an important project....and they're
ready the
next day.



Dry-Film Resist:

I ran Riston dry-film for years.  Even with a shop-made exposure
unit, I never burned longer than 2 minutes.

This unit was a simple wood enclosure, with the bottom lined with

aluminum foil, then eight 20w germicidal-UV lamps (these are the
clear glass ones); and a piece of 1/4" plate glass on top; about
an
inch above the bulbs.  Not at all a collimated source; but with
10
bulbs mounted right next to each other all the way up the box; it

faked 'collimated' quite well...lol.

Note:  it is WELL worth the trouble to find and buy the
germicidal
lamps.  They burn an image far faster than 'black light' bulbs.


I used a piece of 1" thick aluminum tooling-plate about 12x16" as

a "pressure-shoe" on top of the board.  This flattened out warp
in
the board and kept the film up tight to the board over the whole
board area.

That is important, by the way.  Many times, a poor result is
caused
more by poor contact between artwork and board; than by poor
artwork.

This unit was built to expose 8x10" images evenly.  Average
exposure
for good density in Riston 4700 was around 90 seconds, as I
recall.

Note:  most common types of glass eat UV...i.e absorb it,
attenuate it.
So don't use any thicker glass than what you feel you need for
strength.


Laminating Dry-Film Resist:

You don't need a "dry film laminator" to use Riston.  I used an
antique
GBC graphics-art laminator that I pulled from the trash
somewhere.

In a pinch, I've also 'laminated' it onto 2" square items using
nothing
but a plain old clothes-iron.

Another "laminating" method that works fine on small boards is to
heat
the board in the oven (to roughly 200-250F), lay a pre-cut piece
of
dry-film on the board, and then simply use a medium-durometer
rubber
roller to 'laminate' it.


Riston comes up on ePay all the time...usually pretty cheap too.

Although it is claimed to only have 6-12 months shelf-life; that
is
wildly conservative.   It lasts FAR longer than that.

Good storage conditions help a lot (i.e. cool and dry and dark,
and
keep it sealed in thick-poly bags).

I am still using the last of a case of 2 rolls which I bought 15
years
ago; and they were old/surplus -then- !  ...lol.


Liquid Resists:

If you're not the type that knee-jerks a freak-out about
chemicals,
then there are many excellent liquid resists you could use.  I've
never
had any problem ordering a quart of whatever I needed.

Some of these are aqueous-alkaline-developing like most
dry-films;
although many of them strip via a solvent.  Some use a solvent
for
both strip and develop; as KPR did.


Expect to pay $40-90/quart.  However, note that your processing
(i.e., your etch-bath etc) is likely far more benign than what
the
resist is capable of handling at 'normal' coating thickness; so
you
may not need 'normal' thickness.

Most of these resists were designed to withstand Piranha
etch...and
Oxygen Plasma etch...etc..

So in a bath that's relatively mild by comparison; a much thinner

layer might work just fine for you.  If so, you may be able to
'cut'
the resist 2:1 or 3:1 with solvent; and get far more boards out
of a
quart.  But even when run 'straight', a quart does a LOT of
boards.

I used a lot of KPR in the past, and quite a bit of Shipley stuff
as
well; and I never bothered with spin-coating or spray coating.

I found that simple dip-coating gave excellent results; so why
make
things more complex than they need to be?

For dipping, I made up a very thin tank by TIG-welding some .062"

316 stainless from the scrap-bin.  This tank was only about 1/2"
thick, so that it only needed a bare-minimum of resist to fill
it.

I also fabbed a stainless lid for it; into which I fitted a thin
rubber strip
as a gasket.  I cut a hole in the counter-top and mounted the
tank in the
counter.  With the lid on, the resist never lost any solvent.

I didn't bother with a motorized puller either....I simply pulled
the board
out slowly by hand; and hung it right above the tank to drain and
set.

The trick for getting good results via simple dipping like this,
is to
prevent the coating from drying too -fast-.  You want it to stay
non-viscous long enough to fully drain off.  That will give you a
nice
even-thickness coating over the whole board.

The way to make it dry slowly (especially with solvent-based
resists
like KPR), is to make the air around the board stagnant.  I did
this
by putting a cardboard box upside-down on the counter, over the
board where it was hanging above the tank  This box was no larger

than necessary; i.e. holding the minimum amount of air.

This produced a solvent-rich environment in the box; which caused

the coating to dry at a slower rate.

The whole process was dirt-simple and worked like a champ.  We
routinely held razor-sharp 2-mil lines and spaces in chem-milling

nickel foil; and likely could have done much finer.

Note:  for success with resists, you must CLEAN the board first.

After the preliminary scrub with dish-soap and a 3M sponge-back;
I use an alkaline-cleaner soak for 5 mins; then rinse; then into
a
microetch bath of dilute sulfuric/peroxide (about 5% of normal
strength).  30-60 seconds in that, and I'm looking at brand new
copper.  Then rinse, blow off liquid with filtered shop-air; and
laminate soon....don't let it sit around for days and oxidize, or

collect new dirt.

Note:  I don't like the HCL/peroxide.  It fumes more, and it's
harder
to regenerate.  The sulfuric-peroxide etch is very fast, produces

fewer fumes; and can be regenerated simply by chilling it.  The
copper falls out as copper-sulfate crystals; which can be easily
dip-netted out.


But back to resists...  In general, liquid resists are great....I
really
like 'em.  Dry film only has the advantage in high-volume
production
work, imo.

Best of luck with your efforts,

Richard

--
============================
Please do NOT attach pictures without contacting me FIRST.
DSL isn't available here.  Do not add address to any lists.
============================