Context: I have never used a CNC, but I’m experienced with 3D printing, electronics, and general assembly. I’ve done a fair bit of research so I feel like I’m at least to the point I can ask meaningful questions.
My current thought is to build a LR4 with a 24”x48” working area so I can work with 1/4 sheet goods (I’m in the US). A Lowrider V 1/4 if you will. I’d like to get a picture of what I can reasonably expect the machine to do and if there are specific parts of modifications I should consider.
I know I can expect to work with foam, wood, various plastics, and eventually aluminum once I develop the skills. I would also like to work on very small things.
Is the LR4 precise enough to make PCBs? What track width and separation is reasonable?
Making 1 or 2 layer PCBs would be neat. Precision achievable will depend on quality of prints, materials, assembly, time invested precisely calibrating dimensions/squareness, flatness of spoilboard, feedrate, router bit sharpness, etc… That said, have seen people also mill sub-millimeter accuracy. The smaller the gantry the easier it is to get precise cuts.
Along these lines, for very small precise tasks, I considered adding temporary clamps to X and Y belts to effectively shorten the belt length, but I never got around to testing this out. Mentioning incase it helps…
Am hoping someone straps a fiber laser to their LR4 (or LR3) to make relatively fine PCB traces compared to what’s possible with router bits.
Guessing you’ll hit router bit tearout as the first precision limiting factor, before hitting precision limit of LR4 XYZ motion (for a very nicely assembled, and well tuned build). Curious what others think? Am wanting to mill some large simple fat trace PCBs at some point, yet another lighting project…
Ok, sounds like it’s feasible. I wouldn’t expect ultra precise boards with tiny components, but as long as I can get through hole PCBs to work I will be happy.
Is there a concern about maintaining alignment between switching tools?
Should I plan to drill and cut traces with the same tool, or could I switch between drill and cut with a V bit?
I watched a video about the Carvera Air and it has some sort of “Auto Z” feature. Is it possible to put something like that on a LR4?
Regarding quality of prints and tuning, I have a Prusa MK4 that should be able to deliver good quality. I also have a 350mm Voron Trident which does a very good job, after plenty of tuning so I’m used to that.
I’ve had my eye on making PCBs with my Premo. I’ve consumed a number of articles and videos, and, as outlined by @MakerJim, the hard problem is the flatness of the material. You have to dial in the flatness of the fixture, plus solve for any curvature in the board, and deal with any inconsistencies in board thickness. Even slight variations not accunted for will trash the PCB.
Using a laser on a MPCNC is a more forgiving approach, but involves a more traditional acid bath. Steps:
Clean the board
Paint with lacquer
Laser off the lacquer to expose the copper where needed
We have a high performance PCB mill at work and it’s quite different to a regular mill. It uses a floating Z stage with a pressure foot so it glides over the surface of the material. It sets the cutting depth by adjusting the position of the bit in the chuck relative to the foot using an automatic tool-changer type system and a sloped surface to drop the bit onto…
It’s not just the position of the PCB, it’s also it warping as you cut, especially if you’re rubbing a lot of copper out on the top surface.
Honestly, though, unless you’ve got a very specific reason to want to make PCBs yourself, we just get them from JLC or PCBWay despite having a ~$100k PCB making workflow that’s pretty easy. Home-made PCBs are just kinda crap, quality wise, and lead to a lot of other problems…
Edit: And a 60K RPM spindle so you can actually get some performance out of those tiny cutters… Not to mention sub 0.01mm runout…
If you’re just after through-hole boards for the hell of it then that’s probably fine, 0.1" spacing should be pretty easy to hit with a square 0.5mm endmill and cutting ~0.5mm into the PCB. It’s using a V-bit and trying to get ~0.2 or better traces/spaces where things get nasty.
I bought some blanks thinking I would mess around with trying to mill some simple PCBs my son and I designed. As we finished the design, I suggested we send it to JLCPCB just to see how much it would cost for 5 boards. It turned out to be about $32 USD for all 5 (4 layer boards) with the new user discount. The boards took 10 days or so, but the democratizing effect of having services like that is crazy.
Yeah. In my experience the cost and capability of PCB fabs has gotten amazing while the hassles of dealing with home-made PCBs remain the same or even get worse as parts shrink, etc.
It’s truly insane what you can do with a $5 PCB, $5 stencil, some solder paste, a credit card as a squeegee and then reflowing it all in an old skillet. I can’t think of many other industries where there’s so little gap between what can be accomplished by big companies with massive production lines and a single person mucking about in a couple of hours after work.
I love jlcpcb and pcbway for most boards anymore…but I do still etch my own from time to time.
Can’t say I’d consider milling them though. I’ve seen quite a few people mill them and I’ve never been that impressed with the results.
Now I would consider using my CNC to drill the holes and/or cut the outline of boards. My old MPCNC I didn’t trust enough to try - but that was because I build it well beyond the intended specs and gave up rigidity for size since it fit my primary need. My new LR4 I’m working on…I would definitely consider drilling a board or cutting an outline.
I’ve found etching to be way easier than it was when I first tried it many years ago. Photo resist film is cheap (I paid $12 for a roll of Amazon that’s likely going to be a lifetime supply) easy to work with (subdued room light is sufficient - no need for a darkroom or anything fancy.) I use a cheap $15 laminating machine to apply it and washing soda to develop it. I also built a frame with some 3D printed jigs to hold strips of UV led’s at a set distance to get very repeatable and even exposure.
The hardest part for me was getting negatives to print - but that was mostly due to trying to do it on a laser printer - switched to a cheap inkjet with some film designed for screen printing and started getting great results.
The other big difference is I gave up on ferric choloride for an etchant and now use a mix of muratic acid I buy at the grocery store in the pool supplies section with standard drugstore hydrogen peroxide. Works so much better and I don’t even bother to heat it usually (but I also work outside in the desert so it’s pretty warm to begin with!)
That said … if you really do want to mill them…there are solutions. But it would take a good bit of trial and error. Using marlin with bed leveling and a vacuum table could probably make it workable. And ESTLcam had a demo back in v8 of surface scanning and engraving on an uneven surface: https://www.youtube.com/watch?v=ApBfHW7AhsM
