Insofar as the Y axis, you could try running the motor for the paper path
rollers though your microcontroller, then you could modulate the speed of
the motor after it has already engaged the board but before the printing has
started, until the correct alignment criteria have been met.
I can't think of much you can do about the X axis other than bulking up the
mechnical rigidness of the printer, and there are limits to that.
It'd be interesting to see a true flatbed implementation of the DLP process,
though I haven't done enough with laser printers to have a good feel for
what would be involved.
-Andrew
On Sun, May 23, 2010 at 11:29 AM, Mark Lerman <mlerman@...> wrote:
>
>
> Well, I've done some tests trying it the easy way, and while it
> SHOULD work, it doesn't, at least not consistently. The problem is
> that both the carrier and the board are (obviously) much thicker than
> paper, and when they hit the drum/transfer roller interface there is
> a slight (and variable) delay as first the carrier and then the board
> go through. There are also problems in the X axis because the carrier
> "grabs" slightly when it hits the drum and has a tendency to get a
> little out of parallel to the motion axis. In theory we could wait
> till the pcb (or a "header" of equal thickness) is under the drum
> before telling it to print. However, this may create problems in the
> charging timing since the printer does not expect there to be a board
> present at that time. The X axis problem could be handled with a
> better and longer carrier guide, but too much friction would create
> its own problems.
>
> There is no reason that DLP can't do ds boards, but I'm far from sure
> that the mechanics of your typical laser printer can be "bulked up"
> enough mechanically to produce consistent results. Probably the best
> way to do ds DLP would be to make a flatbed laser system, much as has
> been done for inkjets. Print one side, put it in an oven to fix the
> toner, flip it and print the other side.
>
> This would also be the way to go for a commercial product - anyone
> interested in collaborating?
>
> Mark
>
>
[Non-text portions of this message have been removed]