Laser Etched PCB Production

Like most of us, [reonarudo] isn’t satisfied with the current methods of homebrew PCBs, so he put a laser on a reprap and started burning some boards.

The basic procedure is to cover a copper clad board with matte black spray paint. A laser was installed on the X carriage of the reprap. [reonarudo] used cad.py to convert the board files into g-code and fired up his laser. The quality of the boards is highly dependent on the accuracy of the laser so after working through some mechanical problems [reonarudo] managed to make some passable boards.

We’ve seen just about every manufacturing method imaginable applied to PCB production. Etch resist and toner transfer do the job and a reprap milling machine is pretty neat, but lasers are so much cooler. While it may not be perfect (yet), printing PCBs with a laser shows a lot of potential. Check out a video of [reonarudo]’s bot burning some copper after the break.

[youtube=http://www.youtube.com/watch?v=Chq6TcTeDJc&w=470]

42 thoughts on “Laser Etched PCB Production

      1. For very simple designs, I agree. But when you start dealing with SMT components and small package sizes it ends up being a huge time saver, rather than having to make multiple home prototypes and wasting material.

  1. I still dont understand the OCD of “I gotta have this board now!!!!!!” that people have.

    I’ll gladly wait 2 weeks for a board house to make me one that does not look like it was home made.

    1. I normally have time to do projects on weekends, if i use toner transfer, i can make the layout on friday and saturday, etch on saturday evening and drill and populate on sunday.

      I also like to use fab houses, but not for the first or second prototype, especially if the board is only 3cm x 3cm. In the two weeks it takes to get a board from the fab house i can sometimes go throught the whole development cylce to the finished product. If i than need more than one board or need to switch to small SMD components, i will go to a fab house.

    2. 2 weeks? Make that 3 to 4. And it’s quite expensive. If you don’t need tiny SMD components, and don’t need a lot of boards, then home etching is worth it.

      I did need some good quality boards a while back, and http://www.jackaltac.com/ was the cheapest one that could deliver within a few months (makepcb is cheaper, but unreliable in how long it takes)

    3. Because when you’re in the development/fucking around stage you don’t want to sit on your ass for weeks on end waiting for a PCB to show up so you can keep working. Using a fab house for the “last” revision, sure, but not during the actual development, that would drive me nuts.

      1. I have to agree that fast turn-around for prototypes is a must when you are getting paid to do it. I can turn a complicated board around with SMD parts in a day with my LPKF mill. The board uses pins to mount on the mill so it can do double sided and multi-layer boards too. I played with conductive epoxies and they work up to a point. LPKF sells an add-on to do plating of through-holes including vias. Any of the free design tools works for design. I use the software that came with the mill to generate the job for the mill. You can do additional editing like adding areas that get milled that are between traces. You have to seperate the work into layers for the different tools. So I use a layer each for traces. A layer for drill holes. I add a layer and outline the board for cut-out. I use some tricks like adding zones that are not used by the autorouter so I don’t get vias under the chips. I usually do a few milled boards before sending a design out to a board house if I am working with a group. Otherwise I might just do it all on the mill. The results are pretty good. It is a little bit of a challenge assembling the boards without the silkscreen.

    4. I think there is not one correct way to do PCB. It always depends what you are doing. I personally use three methods depending on task. Fastest is ‘dremel method’ where I cut boxes on double sized pcb. Bottom is ground, top has the boxes and components. this is good way when you design as you go. Kind of rapid prototyping…
      Second way is to use toner transfer, this method is good when there is too much to dremel. Or you have IC chips or some other stuff that require more precision.
      Third way is use photo sensitive PCB, I use this when IC:s get really small, like 0.5mm TQFP. With these chips toner transfer resolution is not good enough.
      I have never used board house, as I feel I haven’t needed them yet. When complexity raises, have enough vias and board is full of TQFP’s the board house will be good idea. But what I am saying is that it is waste of money and time to use commercial fabrication when dremel or toner transfer could be used.

      Personally I do not like method presented here. It is cool and all but it leaves so small gap between trace and ground that short circuits are likely. Or you can run more rounds around traces but then it gets slow. I have made some PCB with CNC mill and cutter bit, but didn’t like those short circuits.. :-)

      1. In some cases it’s not soo much the quick turn around, it may be more like I don’t want or need 3 of a board and I don’t want to pay for 3 boards when I only want and need only 1 PCB, yet unless that board is 12×12 just to get a single board from a fab house would cost a few hundered dollars, more then it would cost someone to whip up something like this.

  2. Unfortunately for really good results you need a UV laser. Commercial products that do this are $100k+ Perhaps several nitrogen TEA lasers can be ganged together with an optical fibre to get high enough power levels?

  3. Cheap board houses don’t exist in the UK, especially not for one-off projects for the home user – you’re still talking 50 quid plus for a small board which can then have mistakes/mods needed.

    Though this is damn slow, small improvements will make it pretty usable!

  4. Interesting.

    What kind of lasers can be used to do this kind of work and what is the minimum power?

    Coating a board with toner and then fusing it with a laser could work? Or it’s better to transfer it?

  5. What’s the chemical resistance of the plastic stuff repraps (etc.) use for printing?

    Has anyone tried printing the traces onto the board and then etching? Would the plastic have enough adherence to the board to prevent etching?

    Stupid idea?

    1. It’s usually ABS, so I imagine it’d be resistant.

      However, I don’t think you could get traces as thin as you could with a mill or a laser like this. I think the thinnest extruder is about .30 mm right now or about 11.8 mils. You’d definitely need to use the stepper extruder as a dc motor would ooze way to much. Not sure how well it would stick.

      It’d probably be easier to replace the print head with a contraption for a fine resist pen, kind of like MaketBot’s unicorn thingy.

  6. So heres the thing. If you have access to a laser cutter (or rep-laser :) then this method is good. Its what I use. However I have a epilog laser cutter. I used this method for a few things.

    1. To make a prototype board before spending $$$ on a fab house.
    2. Sometimes a PCB is just cleaner than a breadboard. So I use this to create PCB’s for things that I would not normally pay to have a fab house do.

    Great work man!

    Here is a set of my laser etches.

    http://www.flickr.com/photos/rileyporter/sets/72157622690487871/with/4393053680/

    ril3y

  7. I’ve used UV exposure method for years for making PCB for tiny SMD parts. The result is pretty good. It’s not comparable to fab house due to the fact that my home made boards don’t have solder mask or component screen and multi-layer board.

    I’m fine with those limitation, but from time to time I would have problem with under-develop or over-develop when I develop the UV exposed board. Basically sometime the temperature of the developer solution is not at the ideal temperature or if I have the board inside the solution for too long or too short, and I could get a board with thicker or thinner traces. This laser drawing technique seems to solve the problem with consistency. Looking at Ril3y’s result, they seem to be very good. Just leave the work to the machine and don’t have to worry about developing the board in a dim/dark room and worry about whether the result will be good or not.

    -Thanh

  8. Tried something like this using HP laser printer mechanism and UV diode.

    I ran into an infuriating problem, the optics dispersed the beam too much and resulted in a very high power loss and a really lousy spot.

    How about using two laser dots with a fine one in the centre and a larger one on the outside so that you can draw different sized tracks?

  9. Interesting idea. I have used the positive photographic techniques for years. Print on transparency and lay on board,expose,develop and etch. I got good results from this. Plan to move away from this to get away from using the chemicals. This technique should work once he can burn all the way to the metal. Take a look at the comment above from Ril3y and look at his flicker photos. For me I am headed to the milling route. Tired of FECL as well.

  10. Nice! As more powerful laser modules appear to market, maybe we see DIY direct ablation at some point in future ;).

    Rapid availability is key for hobby projects (2-6 hours vs 2-4 weeks starts to really bother if you do small different projects). Toner transfer limit seems to be around 4…6 mil, this allows quite complex boards. Via electroplating with CNC drilling can be done cheaply at home. Usabe silk screen can be done with toner transfer method, solder mask is available in various forms.

  11. I have used heat activated conductive epoxies. They aren’t as good as metal traces but there are some really good heat activated epoxies that you could use as resist. Have you thought about having the laser draw the traces and pads on then wash and etch?

  12. My mill does things like place a dot with the milling bit at the end of each side of a long run and along curves. It also places a dot to start teh drill bits. It uses the milling bit to make the dots. I assume thave issues with the ones on the traces is to make a cleaner cut or make it easier for the mill to change direction? Maybe you could play with the power and timing of the heating of the material to make it cleaner? Adhering the resist might leave cleaner lines than removing the paint too.

  13. I built a tiny engraver using a 250mw red laser and this is exactly the sort of thing I’d like to use it for.

    What’s not clear to me here is, can a “final” board be made this way or is this just a method of preparing a board for etching?

    Second, would a 250mw red laser have enough power for this process?

Leave a Reply to timCancel reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.