I am now working on producing ds boards. My first test was to see if
I can print multiple times on one side of a board with good
registration. If I can do that, I'm certain I can do the rest. I
added a reflective sensor to locate the actual position in the motion
(Y) axis. I could have used one of the leftover E260 interrupter type
sensors, but it was easier for me to use black tape an white labels
than to drill holes. In the final version, in the spirit of
recycling, I might use a leftover. There are some pictures of the
sensor and new mcu board on my pbase site <
http://www.pbase.com/mark10970>.
As I mentioned previously, when the carrier moves through the
printer, the leading edge first hits the moving transfer roller,
which pulls it up into the drum/transfer roller interface. This
causes a variable delay, which messes up the registration. The "fix"
is to wait till a couple of inches of carrier is completely through
the interface before starting the actual timing (timing starts when
pis is pulled low). I still have some work to do, but I'm pretty
sure I can get it to work I printed a test with 5 mil traces - it was
perfect in the Y axis, but about 5 mils off in the x axis. There is a
picture on my pbase site.
One problem I foresee is how to tell whether the sides are registered
before etching. You can't see through the board before it is etched,
and you don't want to etch it without knowing it is correct.
Mark