My production manager/machinist did this for me on our Tormax.
But milling serious PCBs is a lot of work. Once I get past the Dremel stage, it's better to order a real plated-through board and work on something else until it arrives.
-- John Larkin Highland Technology, Inc jlarkin at highlandtechnology dot com http://www.highlandtechnology.com Precision electronic instrumentation Picosecond-resolution Digital Delay and Pulse generators Custom laser drivers and controllers Photonics and fiberoptic TTL data links VME thermocouple, LVDT, synchro acquisition and simulation
Same problem with thermal conductivity; the grease or epoxy dominates theta.
-- John Larkin Highland Technology, Inc jlarkin at highlandtechnology dot com http://www.highlandtechnology.com Precision electronic instrumentation Picosecond-resolution Digital Delay and Pulse generators Custom laser drivers and controllers Photonics and fiberoptic TTL data links VME thermocouple, LVDT, synchro acquisition and simulation
20 years ago I built one out of a surplus 3-axis positioner. Problem I had was creating the negative image profile for the positioner. PCBs are not about what you design, it's about removing what you don't want. But it did make a nice drilling machine. 40 years ago, we had a device with a big spinning drum. You wrapped PCB material around it and the cutter "milled" off the stuff you didn't want. Worked really great until the cutter caught on a trace and ripped it off.
I keep coming back to toner transfer.
Den torsdag den 5. december 2013 20.21.27 UTC+1 skrev John Larkin:
yeh, and for anything serious drawing the schematic, doing they layout, checking footprints, etc. is what takes time
-Lasse
:
from china:
=item2a34cbf87e
than an hour
Programs that turn your Gerber into the negative image for milling is readi ly available (the DIY forums are quite active). When you draw the layout, p lace polygon pours to reduce the milling time and the wear on the drill. Ke ep close eye on the drill, when it gets dull, the copper is spun into a big mess instead of coming nicely off in spuns and the drill get hot
Cheers
Klaus
We have one of those at work, too. ...and they had one at my PPoE. Neither was used, even once, while I was there. The one here is as an anchor for copperclad, though.
That's why we don't prototype. We just design the real product, check the hell out of it, and release the documents to production. Engineering gets the first couple of units to test. We expect them, rev A, to work and be sellable. That cuts the entire prototyping cycle out of the engineering path. There is peer pressure here to get it right the first try.
I do breadboard little things like in the pics I posted. Usually, our breadboards are even smaller, just enough to characterize a part or test some unusual concept.
Some companies assume that the first couple of PCB etches will have enough problems that they can't be sold. That's self-fulfilling.
-- John Larkin Highland Technology, Inc jlarkin at highlandtechnology dot com http://www.highlandtechnology.com Precision electronic instrumentation Picosecond-resolution Digital Delay and Pulse generators Custom laser drivers and controllers Photonics and fiberoptic TTL data links VME thermocouple, LVDT, synchro acquisition and simulation
Actually, a few years ago, they came up with the technique for printing antennas on paper etc.,
Good, fast, cheap. Pick two. In this case, if you want a cheap and fast PCB, it's not going to be very good.
Take a sheet of unplated G10/FR4. Mill channels into the surface where you want conductive traces. Screen with solder paste or just slop it on, and remove the excess with a squeegee. Do not complain about the cost of solder paste. Wipe off excess solder paste. Let dry. Attach components with super glue. Heat with an IR lamp or if you're cheap, an SMT rework hair dryer, to reflow the solder paste. Instant PCB.
I did something like that about 15 years ago in order to get quick RF stripline PCB's. The only difference was that the G10/FR4 board was a single sided PCB, with the copper side down for a ground plane. I glued using epoxy a thin sheet of 0.010" G10/FR4 to the component side of the PCB. (Hint: Use a book press, not C clamps or rollers). The added 0.010" was the depth of the milling so that the dielectric thickness was the required 0.062". This produced some really nice looking milled kerfs (when done correctly) and a fairly good approximation of the final product. I wanted to do the milling with a laser, but management decided that the expense was too much for just prototypes.
-- Jeff Liebermann jeffl@cruzio.com 150 Felker St #D http://www.LearnByDestroying.com Santa Cruz CA 95060 http://802.11junk.com Skype: JeffLiebermann AE6KS 831-336-2558
Surely the surface tension of the molten solder will pull it into a series of isolated blobs?
-- John Devereux
Blobs will appear if I overheat the solder paste for too long a time. The solder paste flux sorta "hardens" after it's exposed to the air for a while, in effect gluing itself to the rough milled grooves in the fiberglass. The glue is temporary, but holds long enough to keep the solder in place. Isolated blobs do not ocurr on a regular PCB because the powdered solder first wets to the underlying copper traces. I've found that if I work quickly, and not overheat the solder paste, I don't get solder blobs. One does see this effect in the form of tiny solder balls if not all the solder paste is properly stuck to the fiberglass. I had that problem when using old solder paste, that was pre-hardened in the jar, and had lousy adhesion. Smearing on some more solder paste, letting the flux harden, and re-heating with an air gun, did a tolerable patch job on the wider traces. However, for fairly narrow traces, the solder mostly stayed in the milled channel and did not form balls. Where it didn't, it was easily repaired with an ordinary soldering iron.
What makes it work is careful temperature control and following a reasonable thermal profile. Here's the basics: I can do it with a hot air SMT desoldering station if I watch the PCB quite carefully. The solder changes surface "texture" when it melts. I didn't know any of this 15 years ago, so I just played with the hot air gun until it worked.
It can also be done with a toaster oven and controller. I haven't tried an oven yet, but have friends that have built PCB's using one. I've been told that getting the temp profile accurate takes some trial and error. I'm not sure what it will do with my all solder paste PCB.
On wide traces, I also had problems with the center of the strip line trace being thicker than at the edges. I just scraped the trace flat with a razor blade. I also had some narrowing of the trace with wide traces, which I solved with some solder paste and heat gun touchup.
I haven't done this in perhaps 15 years, so things may have changed. For example, I was using non-RoHS solder paste and mostly non-SMT parts.
Jim wanted fast and cheap. He didn't say anything about good.
What killed the idea was the emergence of PCB vendors, that could turn around a complex PCB overnight for reasonable prices. What may resurrect it are 3D printers. Silver nano-particle ink on paper. No clue how they connect the component leads to the PCB or achievable bulk resistivity.
-- Jeff Liebermann jeffl@cruzio.com 150 Felker St #D http://www.LearnByDestroying.com Santa Cruz CA 95060 http://802.11junk.com Skype: JeffLiebermann AE6KS 831-336-2558
Yes, but that is why I could not see it working when there is nothing to wet to.
Cool, thanks!
-- John Devereux
Great! are those your final products?
Jamie
Has any one tried the photo gloss Laser (not inject) paper?
I've been told that it releases like the original inkjet paper used to before you they made it not run anymore ?
Jamie
My milled PCB is very different from a normal PCB. It's a collection of shallow gouges filled with solder, where there would normally be elevated copper traces. I could probably do a tolerable job just filling the gouges with roll solder by hand. The hard part is getting it to stick to the PCB material, which is handled by the residual solder paste flux.
In theory, you could take the finished milled PCB, bang it upside down on the table, and all the traces would fall out. Some conformal coating (acrylic or urethane) should prevent that from happening.
Please note that I'm not suggesting that this is a replacement for conventional PCB's, the next big thing, or even an improvement on existing methods. It's a kluge that works well for making a quick prototype while waiting for the real PCB to arrive in the mail.
-- Jeff Liebermann jeffl@cruzio.com 150 Felker St #D http://www.LearnByDestroying.com Santa Cruz CA 95060 http://802.11junk.com Skype: JeffLiebermann AE6KS 831-336-2558
OOohhh! Dental burrs! Super! Thanks!!
List? All of the CDrom labeling systems i have seen use paper of some sort' pre-cut donuts so to speak.
Back in the day, there were a bunch that printed directly on the CD. I took a quick look at conversion and determined that they weren't really conducive to printing in the area where a CD has a hole. Nothing theoretically wrong, but practicality induced aversion. There's still the issue of ink that resists and won't clog.
I found a flat-bed pen plotter more satisfactory, until the .005" resolution became an issue for small geometries. Somewhere around here, I've got a VB program to read gerber files and plot to a Tek 4662.
DROMs
The CD is put in a special tray and they print directly on it look here for example:
-Lasse
ebay, about $1 each. They disappear copper beautifully. I bought a dozen and, so far, the first one still cuts fine. It takes a little practise to get good at it. One side of a cut tends to scallop, so I usually slice a path then rotate the board 180 degrees and trim again.
The other trick is to rub the board with Scotchbrite and soapy water after you dremel. That removes any burrs and preps the copper for soldering.
-- John Larkin Highland Technology, Inc jlarkin at highlandtechnology dot com http://www.highlandtechnology.com Precision electronic instrumentation Picosecond-resolution Digital Delay and Pulse generators Custom laser drivers and controllers Photonics and fiberoptic TTL data links VME thermocouple, LVDT, synchro acquisition and simulation
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.