Shop OBEX P1 Docs P2 Docs Learn Events
Laser direct PCB imaging — Parallax Forums

Laser direct PCB imaging

Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
edited 2012-11-03 16:03 in Robotics
I swore I'd never waste another minute making my own PCBs, but this seemed like such a good idea. Namely, spray a bare copper-clad board with paint, lacquer, or what-not, use a laser engraver to etch away all but the areas you want protected, then etch the board in the usual way. It eliminates the intermediate photo-tools and should provide micron-perfect resolution. Boy, was I in for a lesson!

First of all, not every spray-on aspiring resist etches away with a laser. When most are vaporized by the CO2 laser, the fumes redeposit on the newly-exposed copper, rendering it impervious to the etch bath. This rogue's gallery of resist pretenders includes black spray paint, aerosol lacquer, clear acrylic coating, and sprayed silicone conformal coating. This problem arose engraving top-down or bottom-up, air curtain on or off. Furniture wax had the opposite problem: the etchant attacked it immediately. The only coating I found that worked was MINWAX Fast-Drying Polyurethane (Clear Gloss) finish. It removed easily at a low-power setting and maximum speed, and its vapors left no residue.

As to etchant, I tried both ferric chloride (messy!) and HCl+H2O2 (muratic acid and hydrogen peroxide). I had better results with the latter, although it took forever to etch a board with 3% H2O2. Moreover, I had to increase my minimum trace width to 0.020" just to keep traces intact. But I was finally able to produce appears to be a workable board:

attachment.php?attachmentid=96337&d=1350533032

It's a 74HCT595+transistor drive+RGB LED illuminator for a project I'm working on. We'll see if it works. But first I have to clean, drill, paste, and solder it (and strengthen a couple over-etched traces).

Ugh! I could have had cheap, factory-produced boards in hand in the time it took to do this. I swear I'll never waste another minute making my own PCBs! Anyone who thinks it's a good idea is just nuts!

-Phil

Comments

  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2012-10-17 22:59
    I guess all's well that ends well. Here's the finished board, attached to a PropBOE for testing:

    attachment.php?attachmentid=96338&d=1350539794

    I had one snafu, in that a connected pad broke off from the 'HCT595 TSSOP pattern, and I had to bend the lead up to connect it with a wire. But the board works the way I wanted -- and, whoa! is that LED BRIGHT!

    I'm still gonna steer clear of in-house PCB fab from now on, though. It's just not worth the hassle and mess.

    -Phil
    586 x 459 - 45K
  • idbruceidbruce Posts: 6,197
    edited 2012-10-18 03:17
    Very cool Phil. However, I am very surprised that you chose that route for your next adventure in PCB creation. However, cheaper is usually better. I think if I had given up on PCB creation in the past and decided to try it again, especially with laser direct imaging, I think I would have ordered some laser direct imaging film, such as the film mentioned on this page:

    http://forums.parallax.com/showthread.php?137850-DVD-Laser-Diode-And-Optical-Block-(Future-Laser-Photoplotter)/page3

    Perhaps it would have been a more tolerable experience.
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2012-10-18 09:23
    Bruce,

    The whole idea of laser-direct imaging is to eliminate the film and the photosensitive PCB material. I've used film before, back when I could get decent transparencies from the local print shop. But print shops have advanced and go direct-to-plate now, skipping the film step altogether, and have kicked their photoplotters down the road. That's what I was trying to accomplish here.

    Although I came up with the laser resist-ablation idea independently, it's not new, and others have had varying degrees of success with it. But even those successes are tempered by the eventual mess of having to deal with -- and dispose of -- caustic liquids. It's just not worth the hassle, when I could have better proto-grade boards for cheap in a couple days from a board house.

    -Phil
  • davejamesdavejames Posts: 4,045
    edited 2012-10-18 09:30
    If the laser has enough power, why not just blast away the unwanted copper leaving just the traces?
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2012-10-18 09:39
    Copper is a nearly perfect reflector for the 10-micron IR output from a CO2 laser. It would not absorb enough energy to ablate.

    -Phil
  • davejamesdavejames Posts: 4,045
    edited 2012-10-18 09:43
    ...got it - understood.
  • Martin_HMartin_H Posts: 4,051
    edited 2012-10-18 10:48
    I've used Fritzing to design a few PCB's, but so far I haven't made any of them. Both toner transfer and the photo resist methods seem pretty labor intensive, then there's the etching and drilling. The PCB manufacturing services seem expensive for one offs. Hopefully someone will come up with a better way to make them.

    So I keep doing point to point wiring which is ugly and somewhat error prone. It also isn't SMT friendly either.
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2012-10-18 10:58
    I do point-to-point when I can, but the only 'HCT595s I had were TSSOPs and the LED is leadless. I didn't have a choice this time and thought I could make the PCB myself efficiently. But three days to fab a small board is not a good use of my time.

    -Phil
  • idbruceidbruce Posts: 6,197
    edited 2012-10-18 10:59
    Phil
    The whole idea of laser-direct imaging is to eliminate the film and the photosensitive PCB material.
    DuPont began formulating specialized photoresists for Laser Direct Imaging (LDI) over 20 years ago, and continues to lead the industry with its Riston® LaserSeries films. Ultra fast photospeed, high performance, and compatibility with conventional printed wiring board (PWB) processes are critical to help PWB fabricators optimize their LDI equipment investments.
  • Martin_HMartin_H Posts: 4,051
    edited 2012-10-18 11:11
    I've used Busboard system strip boards which are like using a breadboard. This works well for through hole at keeping the rats nest of wires under control. I haven't done much with SMT, but Busboard systems has some SMT friendly strip-boards too. Here's one example http://www.busboard.us/photos/BPS-IMG-SMT3U.jpg
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2012-10-18 11:11
    That's a different kind of material, Bruce. It's photosensitive and requires developing, just like film. My laser cutter works by ablation, i.e. vaporizing the already-hardened resist to expose the bare copper.

    -Phil
  • idbruceidbruce Posts: 6,197
    edited 2012-10-18 11:27
    Phil

    Okay point well taken about ablation, but by using a proven method and films, you may have increased productivity with your laser equipment and perhaps made it enjoyable. With your process, only the development stage was eliminated. Development of those films is accomplished in less than a minute.

    All I am saying is this, if there is ever a "next time", I might consider a film that is compatible with my laser system. And if not, experimentation is almost always fun.

    To be perfectly honest, your results look good to me.

    Bruce
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2012-10-18 11:32
    Hi Phil,
    I actually had a university course in etching for my Fine Arts degree. What you are discovering is that there is rather a vast array of materials that can be used as an acid resist for copper etching. After all, even a Magic Marker will work or a transfer of the black from a laser printer.

    You might just try cleaning the copper board and giving it a thin layer of automotive car wax. Just rub on and wipe off. The whole point is the acid resist doesn't have to thick and doesn't have to be hard. There are art formulas that are purposely soft and thin. Even wax will work. Good old tar diluted with turpentine is often rubbed on.

    Regarding etch, the muratic acid undercuts the copper as it is aggressive; whereas the ferric chloride tends to gently etch straight down without undercutting. The undercutting will actually loosen foil from the board in some cases. There is quite a bit of information about how to 'enhance' etching formulas in the art world to get exactly the right amount of detail and to have the plate hold ink well.

    You might try researching intaglio printing formulas for etching copper plate printing and acid resists.

    One's first DIY PCBs tend to be rather knarly. But use pre-heated ferric chloride and don't rush the etch if you want crisp results from the etchant.

    And while obvious everyone tries to skip the rubber gloves and a plastic apron. Ferric Chloride will stain any thing, Muratic acid will show up as holes in your clothes after the next laundry cycle.
  • idbruceidbruce Posts: 6,197
    edited 2012-10-18 11:38
    @Loopy

    I think Phil has been allured to the voluptuous fab houses (promises of less mess and quick turn around). I don't think he will be repeating his experiments anytime soon, although I hope does.
  • Martin_HMartin_H Posts: 4,051
    edited 2012-10-18 11:48
    Loopy. I've heard people say that muratic acid mixed with hydrogen peroxide forms cupric chloride after it is used as an etchant. Cupric chloride itself is supposed to be a good enchant as well, so it becomes reusable.
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2012-10-18 12:06
    Is muratic acid merely another name for hydrochloric acid, or a less pure, industrial form of acid?

    Ferric chloride actually gets better with use. I suspect the cuperic chloride evoves. The fresh red liquid is not as good as the muddy brown. And when it gets greenish black, it is no longer useful.

    Artist use hydrochloric acid, never muratic - I am not sure why.

    Yes, the fab house are great if you plan to make multiple boards. But there is always the desire to make just one board for a special project. I do this sometimes and have my own methods that I religiously follow.

    I just wish I could find an ink jet printer that has a bed for printing CDs and could adapt it to directly printing the ink onto a clean copper circuit board. So far, I haven't found one that I care to fool with, so I print transparencies on an HP Ink Jet and use a photo exposure process with ferric chloride.
  • Martin_HMartin_H Posts: 4,051
    edited 2012-10-18 12:12
    Muratic acid is a common name for hydrochloric acid. If you walk into a hardware store it is what you ask for.
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2012-10-18 12:15
    idbruce wrote:
    ... by using a proven method and films, you may have increased productivity with your laser equipment ...
    My laser equipment would blast the resist away before there was even a chance to develop it. It's not the same kind of laser direct imaging equipment used by PCB houses.
    ... even a Magic Marker will work ...
    I tried that. Oddly enough, the laser didn't touch the Magic Marker ink.
    You might just try cleaning the copper board and giving it a thin layer of automotive car wax.
    I tried wax. The etchant attacked it.
    Regarding etch, the muratic acid undercuts the copper as it is aggressive; whereas the ferric chloride tends to gently etch straight down without undercutting.
    Both etchants caused excessive thinning of my traces, due I'm assuming, to undercutting. Ferric chloride was the messier of the two to use, since it stains everything it comes in contact with.

    -Phil
  • idbruceidbruce Posts: 6,197
    edited 2012-10-19 04:42
    Phil
    My laser equipment would blast the resist away before there was even a chance to develop it. It's not the same kind of laser direct imaging equipment used by PCB houses.

    I knew that was coming before you even said it. :)

    Okay, since I have no idea what kind of equipment you are using, I am only offering suggestions, but I really don't have any idea how far fetched they may be.

    I would assume that your equipment is some type of XYZ stage assembly, with you laser being mounted on the Z axis. If this is the case, then you should already have most, if not all of the necessary components. So here is couple suggestions:
    • Modify the modulation or power output of the laser, and perhaps the focus.
    • Make an adaptable and attachable laser head with a bypass, specifically for imaging PCBs.
    In reality, developing and etching PCBs is a snap, once you have a decent work area and system. If and when you get the laser imaging part corrected, I would imagine that you could produce a very nice looking PCBs in 15 - 30 minutes or less.

    In my experience, with the exception of opaque positives (not applicable in your case) the biggest problems associated with making PCBs is having a dedicated work space, a good plan of procedure, and a good plan for dealing with the chemicals. Considering that you are not too fond of the idea, well this eliminates half the battle, because you would only use the system for quick prototypes. And also considering that you would not be using a photoimaging method, well that eliminates a lot hassle of dealing with an exposure unit and the space required.

    If you got a laser head to expose film properly, all you would need is a nice little dedicated developing and etching work area, perhaps 3'H X 3'W X 2'L. Proper chemical temperatures make developing and etching a breeze. Please refer to the following thread for imformation pertaining to a dedicated work area and temperature control of the chemicals. It should be noted that even though the photos of this link do not show lids, all of the chemical containers have lids.
    Phil, I hope you do not give up on it, because it is not as horrible as it sounds. Just get your procedure organized, and I am sure, it won't be a bad thing. Besides, I would like to learn from your quest. If not, then at least I have made an attempt to try and persuade you.

    PCBs in less than 30 minutes. You can do it and enjoy it :)

    Bruce
  • jdoleckijdolecki Posts: 726
    edited 2012-10-31 18:38
    Just use the laser to etch shallow canals on the board then fill the canals with some

    Conductive substance.

    What about using a 3 d printer to lay down some traces

    Just some random ideas
  • eldonb46eldonb46 Posts: 70
    edited 2012-10-31 20:26
    I have been building PCB using the Toner Transfer Method for several years, with good results. My typical board contains 8/8 mil traces. See the following blog search link for examples (Note: this search link contains some non-toner-transfer PCB projects).

    http://wa0uwh.blogspot.com/search/label/PCB

    Once the Tower Image is printed, the remaining process takes about 20 minutes, which includes the clean up. I recommend it for the Hobbyist.

    I use Fab Shops for the more complex double sided boards.

    Eldon
  • idbruceidbruce Posts: 6,197
    edited 2012-11-03 16:03
    Thanks for the link Eldon.

    Talk about a scary challenge for the homebrew guy: EDIT: I would be more inclined to try some of the challenges, but that is a big hunking chunk of PCB material for testing.
  • Here's a video directly related to this thread: https://youtube.com/watch?v=OTGZcY7WyYI
  • Wow, how did I never see this thread???? Been wanting to try this with my laser for a while, just never got around to it. Thanks for the insight into the challenges Phil and thanks for the more recent video of what seems to be a success, Bruce.

    Funny thing, I have a few thousand of those Osram 1.5 watt RGB LEDs along with a matching number of Silicon Touch Technology DD313 constant current driver chips that have been waiting for me to design a PCB for them. Imagine an RGB LED module similar to a WS2812, but with brightness enough to be literally dangerous!
  • Heater.Heater. Posts: 21,230
    WBA Consulting,
    Imagine an RGB LED module similar to a WS2812, but with brightness enough to be literally dangerous!
    You mean like this 3 watt RGB chainable LED module:

    http://hackaday.com/2015/11/06/three-watt-individually-addressable-rgb-leds/
    https://www.adafruit.com/products/2741

  • Andrew

    You are welcome.

    I shared the video as a possible solution to making PCBs at home, but also as a source of humor. I think those two guys were pretty comical. I got a pretty good chuckle out of them.

    I tried the muriatic acid and hydrogen peroxide combination once, but not as strong a mixture as they used, and it works very well, but the fumes are devastating and you just know form the smell and the reaction that it is very dangerous, so I stopped using that process.

    Yep, I got a good chuckle from that one!

  • TorTor Posts: 2,010
    edited 2017-03-31 11:43
    What they did was nearly insane. No precaution against fumes, droplets, potential accidents, what have you. There's no way the fumes from that stuff could be healthy. What I did with ferrium chloride back in the day was nothing in comparision.
  • Tor

    Yea, ferric chloride doesn't even come close.

    One other time, I was using a muriatic acid and water mixture outdoors, to strip the zinc plating off of some bolts. The wind changed directions and I got a big whif of those caustic fumes, and I must say that it literally took my breath away for a few seconds.

    When using muriatic acid (hydrochloric acid), a person should be very careful, and use all the precautions necessary to maintain safety. Those guys talk about using water to stop the reaction, but water only dilutes. They should have at least had baking soda, water, and a fume hood at the table.
  • WBA ConsultingWBA Consulting Posts: 2,933
    edited 2017-04-01 04:33
    Heater. wrote: »
    WBA Consulting,
    Imagine an RGB LED module similar to a WS2812, but with brightness enough to be literally dangerous!
    You mean like this 3 watt RGB chainable LED module:

    http://hackaday.com/2015/11/06/three-watt-individually-addressable-rgb-leds/
    https://www.adafruit.com/products/2741

    Yep, remember seeing that one a while back. That was one reason I kept the parts I had. I figured there was interest. With the constant current driver IC I have, I can make small boards that would allow direct connection to 3 IO pins, yet also allow full brightness since Vled feeds directly from a different power source. The project these were used on had the DM412 daisy-chain, 3 channel driver chip feeding the DM313 constant high current chip which drove the LED elements. Unfortunately, I don't have any DM412 chips left, just the DM313. However, with a simple board design, I could easily throw these together to sell.

Sign In or Register to comment.