On 08 Jul 2008 11:29:40 -0400, DJ Delorie <dj@...> wrote:
>
(snip)
>
> With the inkjet and photomask, it looks like 6 will be reliable as
> long as I'm careful about mask defects (dust, bubbles) and the etching
> process (i.e. 6 mil traces are easy, 6 mil space might be an issue,
> 5/7 might be more reliable than 6/6).
I had a few questions I've been meaning to ask ever since I saw your
incredible inkjet results. Do you have a detailed description of your
process, end-to-end?
Specifically:
How do you make your transparencies? You mentioned you use Silkjet
but what type of printer do you use at what DPI? Do you use the
manufacturer's ink or refilled/third-party ink? I've used Apollo
laser transparencies with reasonable success (Some problems with
pinholes and patterns) on a Lexmark office laser (Optra R+) under
Linux. I've also tried a Epson Stylus 860 which is amazingly good at
solid, perfectly black printouts with the density in Gutenprint set to
4.0 like you detail on your web page. Without that density there are
very noticeable lines between each horizontal line since it doesn't
put out quite enough ink to fill it in. This is with Apollo inkjet
transparencies: CG7039-20, using the "premium glossy photo paper"
setting. I'm using Epson cartridges for the moment although it'd be
nice to try something cheaper.
How do you laminate your boards? That is, what type of laminator,
temperature used, and the method you apply the film to the board. I
usually spray a mist of water on my boards and then lay the film
across it slowly, letting the water push bubbles and excess water out
at the other end. I then run it through a pouchless laminator
(Modified with larger input/output slots and a switch to turn the heat
on and off). The first run through is cold, the second as it's
warming up, the final after the "ready" light is on. This ensures
there are no bubbles trapped and also pushes bits of scraped-off film
out (the first time through) instead of sticking them to the board.
How do you expose your boards? I use two U-shaped 25W (?) bug-zapper
florescent bulbs in a side-by-side configuration mounted on fiber
board, about 6 inches away from the board. My secret for ensuring
perfect contact between my transparency and board is using a
vacuum-sealer pouch with the transparency on top of the oversized
board, held in place by a bit of tape. After the air is removed
everything is perfectly solid and I don't have to worry about the
negative shifting. It takes about 6-8 minutes to expose this way.
The vacuum bags are inexpensive enough and you can get a sealer for
around $36 at Walmart. You can reuse the bags several times,
depending on how careful you are not to poke holes and how big your
boards are. It helps to have a large piece of thick PCB behind it all
to make sure everything is rigid once the vacuum is applied.
I understand you use the HCl+H2O2+Cu method to etch your boardds. I
use the same although I've been meaning to try and find some way to
circulate the etchant across the board to speed etching and reduce
undercut. Someone on the net came up with the idea to use a 1/2-inch
PVC pipe with both ends capped. You cut angular slots in the part
below the etchant level and then drill holes above where the etchant
is forced out by centrifugal force; you use a motor mounted above the
solution to drive the pipe around although I think you'd need some
type of simple bearing at the other end to stabilize it. At one point
I was using a small "pond pump" which was really nothing more than a
sealed assembly with a tubular magnet that was mounted to drive a
plastic piece with vanes on it. It worked very well until I left it
in the solution for a couple weeks and the glue holding the magnet to
the vane disintegrated (the magnet got kind of crumbly too).
Thanks-
Larry