Desktop PCB production

CrossRoads:
Sintering, yeah, just showing my lack of knowledge in that area.

How about low vacuum pressure vaporization? Use laser to pretreat the areas to be coated, then put the board in a little oven for the treated areas to attract the vaporized metal (like Aluminum)? Might take a few passes for the metal to build up enough thickness to be durable for reflow soldering to attach parts.

I was just looking at UV curable inks to use as a carrier for some type of metal powder. Seems like it would create a lot of waste. What ideas do you have for the "pretreat" part? If the material is nonconductive, it may be possible to use static just like a laser printer and some type of metalized toner.

Maybe even something along the lines of an etch-a-sketch which would use a magnet beneath the material to hold the conductive powder. Again, the difficulty is finding something with a melting temperature above soldering and finding something capable of providing that amount of heat in a localized manner.

I really appreciate the ideas and input!

CrossRoads:
Maybe use the salt, hydrogen peroxide, and (vinegar? alchohol? this mix was just posted in the last couple of days) as etchant.

I seen that the other day. Definitely something I would like to experiment with.

If you wanted to go the electroplating route you could first use conductive paint to mark out your traces, then use electroplating to make those traces thicker.

Pretreat, I don't know. Not sure what you could do to a piece of fiberglass to make something else want to stick to it.
Static electricity, conductive paint, seems like everything turns into a multi-pass process.
Maybe just burning off the clearences around traces with a laser is the way to go.

Here is a nice step by step guide, with videos of how one company does PCB manufacturing. They use photo-resist, and electroplating. Might give you some ideas.
http://www.eurocircuits.com/index.php/making-a-pcb-eductional-movies/35-front-end-tool-data-preparation-

CrossRoads:
Pretreat, I don't know. Not sure what you could do to a piece of fiberglass to make something else want to stick to it.
Static electricity, conductive paint, seems like everything turns into a multi-pass process.
Maybe just burning off the clearences around traces with a laser is the way to go.

WARNING: 100% noob question ahead!

Won't the laser be reflected by the copper surface of the board? Not only that, but you'll need an extremely powerful laser to do that. I have no idea how to calculate the required power of such laser, common sense tells me it'd be over 15W. How much does such a beast cost?

Focusing this laser to burn 4mil-10mil wide clearances between tracks and pads would also require some effort creating the focusing lenses (reuse a set of Nikkor or Canon lenses??), not to mention that there must be an excellent cooling system for the dissipated heat.

I for one would love a system where blank copper board goes into an end and a drilled, ready to solder PCB comes in the other.

I still have a working, ancient Epson Stylus Color II. they don't even make cartridges for it anymore, I guess. I'll try to work something out of it.

Indeed, it would reflect. I don't suspect that would be an issue so much, though. Enough of it would be absorped. Cutting metal with a laser is nothing new. I don't think the laser would need to be that powerful since we are cutting a very thin layer of copper. Removing the cut pieces, I am not sure about. I don't expect that the material would simply vaporize, but maybe that is how it works.

I would also suspect that focusing wouldn't be extremely difficult given that a laser printer is capable of such a fine pitch.

And you are right, a very large heatsink would probably be required to keep the overall copper cool while cutting it.

Come to think of it, one of the stupid things I tried when I was a teen was to play with the flyback transformer from an old 5 inch B&W TV I took apart. I remember being able to cut through soda cans like butter (and shocking myself several times while experimenting!) I wonder if something similar would work for this purpose? However, the main reason that worked is that it would burn through the material to create a complete circuit. Obviously that isn't what we want to do here (cut through the substrate.) But I wonder if placing both pieces on the copper just far enough that it can't arc on its own, so that the copper would complete the circuit and "blow" a line following the path of the eletrode and cathode. Obviously we know what happens when we blow a trace on a board, the copper trace appears to evaporate.

Sure, but just burn it at an angle, let the reflection be absorbed at the transmitter side. Or, pulse the laser with just enough power to remove the copper and not destroy the fiberglass underneath.
Could also be very thin copper, with things like liquid tin applied after burning to make what's left thicker.
I don't know how much power would be needed.
Also need to address vias and plated thru holes for component legs.

Board to drilled PCB - I think only CNC machines can do that right now, with the via/thruhole limitations.
You can get boards done overnight, just cost a lot more than sending away and waiting a few weeks.

CrossRoads:
Pretreat, I don't know. Not sure what you could do to a piece of fiberglass to make something else want to stick to it.
Static electricity, conductive paint, seems like everything turns into a multi-pass process.
Maybe just burning off the clearences around traces with a laser is the way to go.

heat/uv curing adhesives perhaps?

CrossRoads:
Board to drilled PCB - I think only CNC machines can do that right now, with the via/thruhole limitations.
You can get boards done overnight, just cost a lot more than sending away and waiting a few weeks.

If it were possible to get a laser to work, it would be capable of also doing the drilling/milling/slots, etcs.. by allowing it to cut all the way through some areas. This would need some type of adjustable columnating lens for varying the thinkness of the beam (well, or just multiple passes to cut away material.)

I am no laser expert, so if someone reading this is, I would definitely appreciate some insight or even a "No! That isn't gonna work!" feedback to stop thinking in that direction.

For vias and throughholes, I had seen a few years ago a system that deposited a bead of conductive adhesive and used a vacuum to pull it through the hole.

This guy made a double-sided board at home including vias and through-holes. Lots of steps, though.

Just wanted to show a quick mock-up using the shapelock to make some stepper motor mounts for the X-Y table. It only took me a few minutes to make these and I will be redoing them. The steppers I have here are miniature stepping motors. They have two tiny holes in the face of them for mounting but I don't the matching screws. So what I did was to push the shapelock onto the face and make sure I pressed it into the screwholes. Otherwise, the body is just round, so I also needed to envelope it in the shapelock and create flat plates underneath them to drill for mounting. The plate here is a repurposed breadboard complete with rubber feet.

The shaft on these steppers is splined, so there is something for the shapelock to bite into when I make a coupler for them. I don't have the proper threaded rod for these right now so I couldn't go any further at the moment.

After I reshape the stepper mounts to be more accurate, I will fashion some end brackets with an embedded bearing out of the shapelock. Threaded rods will be attached to the motor shafts and smooth rods along the non-driven edges. The smooth rods will not rotate, so they will be fixed with no bearings. I will just create a shapelock bracket around the rod ends, drill through the shapelock and rod and insert a pin.

If I can't get the precision I want from these, I will look at doing it with DC motors and linear encoders. This is just a proof-of-concept right now.

BTW, you can re-use shapelock. So all the bits that I cut off these brackets get thrown right back in the tub of pellets. So there is an advantage to cutting off excess. And of course, I can get those motors out of there easily by simply heating up the shapelock and peeling it off.

When using shapelock, don't overheat it as it will end up very sticky and runny. So here for example, it might actually penetrate all the way into the motor mechanism blocking movement. It becomes like hotglue when you overheat it.

Found some of my shapelock parts I had made. You have a set of eyelids, two jaw pieces and a servo bracket. All except the servo bracket are rejects though. The jaw parts are way too think and I created a thinner/lighter pair that went off into an installation I don't have access to anymore. They eyelids distorted because water got trapped between them and the measuring spoon because I hadn't dried them off well enough before pressing them in. Also, the edges were cut too close so I didn't have enough material around them to glue them. These will be repurposed (thrown back in the tub.)

The servo bracket is on a micro-servo. It is press fitted, so no screws are even necessary. This was done pushing the shapelock into the screw hole and slightly around the edges of the mount. These lock in place very tightly so it takes some effort to snap the servo out of it, but is easy enough to do.

In the second picture is an XY coupler I just made. Those axles slide smoothly through the shapelock due to its self-lubricating properties while remaining tight enough that there is no noticeable play in them. Top is pushed down a bit so that a screw can be tapped into it for mounting a tool.

Since this is off-topic (but probably interesting to some) I should move the shape-lock stuff to a separate thread. Where would be the appropriate place?

I'd leave it in Project Guidance.

Some more tips when working with shapelock:

  1. It will only stick to itself when it is hot. It does not stick to cold pieces. So it is possible create things like a ball-joint by first creating a ball and letting it cool (in fact, throw it in the freezer!) Then soften another batch and wrap it around the ball and let it cool. The ball will now be locked inside the socket, but the ball and socket do not stick to each other. The ball can move around smoothly. I have also used this method for making universal joints. I'll look into making one to show.

  2. Because of number 1, you need to get your part how you want it first. The only way you can add pieces together is by heating them both up first.

  3. And following on 2, what I do is first get the general form the way I want it then let it cool. If you need to work it, quickly dunk it back in the 150 degree water for a few seconds to soften it back up. The consistency will vary over the cooling period. For some shapes and for things like bends, it is easiest to do these when it is about the consistency of a gummy bear. If you dip your fingers in cold water you can help cool it quicker.

  4. Use a heat gun to smooth, polish, and sharpen edges. Remove material you don't need so you can re-use it.

  5. To get more pretty parts, squeeze the material between glass, and also wait until it has started to harden a bit, then press edges against a flat surface to make them more square.

  6. Look for things around the house/workshop to mold around for various shapes. I use various cylinders when I need to make rings. And I mentioned using measuring spoons for stuff like half-spheres.

  7. Consider using nails while the material is soft so you don't need to drill holes later.

And using the fact that hot shapelock does not stick to cold shapelock, another trick is to make a form out of shapelock and let it cool. Then use that to form your final piece.

Cool - I am picking up a lot from this discussion!

Glad it is useful. I believe you mentioned being into animatronics a bit, so this stuff is very useful for that. You can make just about everything including custom servo horns, levers, rod ends, and I have even heard of gears being made with the stuff (want to try that myself!)

This stuff has held me over while trying (unsucessfuly so far) to save up for a 3D printer.

Due to time constraints with the real job and a marriage with step-kids, I had to give up my gigs building interactive exhibits and most of my tools etc.. got boxed up and put in storage. That was my extra play money to buy tools and stuff like a 3D printer. Trying to get back into it all again.

Oh you mentioned puppets with mouth movements. The shapelock is springy when it is thin, so if you create an upper and lower mouth pallete and make the fold line very very thin, you can operate it easily as a mouth. I have a mechanism drawn up here in one of my notebooks to actuate that, but I'm sure you can figure it out. Basically, you just push a wedge shaped piece on that edge and it squeezes the hinge and the mouth closes. Pull it back and the mouth springs open.

I forgot to mention that there is another useful thing you can do with this stuff...

Just did a quick mold of my face. I didn't get it an even thickness and moved it around a little too much while it was setting so it isn't perfect. Much easier when you are doing it on someone else. lol

The trick here is to not heat it up anymore than necessary. It is comfortably warm at the right temperature. Spread it out very thin and let it cool some. There is a perfect moment where it will mold well and pick up fine detail while not squishing too much from the pressure of pushing it into areas. If you get it just right, you can make a perfect copy of your face.