Close

esot.eric's random over-thinking on zeroing

A project log for Mini PCB printer

Melting toner (as etch resist) directly onto copper clad board using a laser moved by a DVD drive XY system.

esot.ericesot.eric 06/01/2016 at 02:490 Comments

Contemplating my first PCB to be made on this device, which will most-likely be the actual controller *for* my mini-pcb-printer. (Whereas for the first-runs, it'll be done on breadboard). The thought occurred to me, I'll most-likely need double-sided in order to fit two motor-driver-chips and my uC. But how to align those sides...? And some other overthinking on alignment.

Heads-up, this is overthinking, sometimes that's a problem... And, for non-dense and/or one-sided designs this is *most definitely* unnecessary overthinking.


So the quick-thought is this: Drill two holes in the PCB near opposite corners. That's the "easy" part, right? Next, install on the "bed" two pins aligned for just that. That alone might be enough, but another thought is:

a) skew (maybe my two axes aren't perfectly parallel, but good 'nough that say a TQFP could be soldered-up... That'd work fine for one side, but what about alignment on the other, wouldn't the trapezoids align such that their sharper corners are askew to each other?

b) zeroing the axis repeatably (on either side) might be a bit difficult

Now, if we can get *rounded*-head pins (or maybe even flat-headed?), surely they exist, with a slight taper such that they center themselves on the PCB-holes...

Now, I'm already reusing the DVD-drive's optics, which not only include a laser, but *also* photo-diodes... And our PCB material is coated in black toner... so the pins themselves would be a very bright reflection of the laser-beam.

So, darned-near perfect alignment could plausibly be achieved by manually moving the laser-head/bed *near* the pin-head, and a short "search" algorithm for that reflection. Do this twice, once at each pin, and we can also determine (and automatically-subtract) [small] potential-skew between the two axes.

Again, it's probably overkill.

A few other-such-overthoughts:

Design a "part" in the PCB-layout software to auto-place these holes (and avoid placing other parts in that area, or even plan to use them as vias).

These holes needn't (maybe *shouldn't*) be at the far-extremes of the axes' motion.

These alignment-guides could auto-calibrate the system for inches-to-stepper-steps. (On my system, the two axes are clearly different thread-counts-per-inch)...

A jig could easily be made for the actual drilling of these holes; drill one first, place that hole on a pin on the jig, Bam.

Another(?) jig could easily be made for alignment for drilling the other vias/holes.

Inserting the pins into the "base"/"bed": Currently my "bed" is nothing but a sheet of acryllic... Drill holes in some (single-sided) PCB-material (maybe multiple layers simultaneously, for stability and verticality). Rest pins and PCB-jig(s) atop the base. Heat Pins Hot. (Soldering-iron, heat-gun? PCB-jig(s) acting as a heat-shield?) Press down into base.

(Better to use thin "pins" e.g. sewing-needles, or wider ones such as docking-station alignment-guides? Latter is more likely to have a large enough surface for reflection...)

Sharper-pins = easier "post-it-note focussing"?

(brightness ==> focus?)

(will two pins actually be enough to determine skew when axes' steps-per-inch are unknown?! Hmmm)

Again, it's probably overkill.

Discussions