Last chance to enter The Hackaday Prize.

10 ways to etch PCBs at home

There are a ton of benefits for etching your own circuit boards at home, chief among them the ability to design a circuit in the morning and have a prototype in your hand by lunch. There’s always the question of how to etch the board, but [NurdRage] over on Youtube has all the chemistry covered on ten different etchant solutions for DIY PCB manufacturing.

The peroxide-based methods use simple over-the-counter Hydrogen Peroxide to remove all the copper on a PCB. By combining H2O2 with either Hydrochloric (muriatic) acid or Sulfuric acid, you’ll get a relatively easy to acquire and somewhat safe etching solution.

Historically, the favorite etchant for the home PCB manufacturer has been Ferric Chloride and is still surprisingly available at a few Radio Shacks around the US. Another chloride etchant – Copper Chloride – is one of the most reusable etchants available, able to be regenerated by simply bubbling air through the solution. You can actually make Copper Chloride etchant by reducing down the products of an H2O2 + HCl etchant, making this a very good etchant for PCB pros.

In the ‘miscellaneous’ category, [NurdRage] goes over some alternative etchants such as Bleach and HCl, Nitric acid, and potassium nitrate and HCl; the potassium nitrate etchant is fairly similar to aqua regia, so if you’ve ever wanted a gold PCB, this is the way to go.

Balancing the ease of production and safety of all these etchants, we’ll stick with our Hydrogen Peroxide and HCl etchant for now, at least until we move up to CuCl for the best etching machine we’ve ever seen.

Comments

  1. Eirinn says:

    Ferric Chloride has worked well for me. However all the methods of transferring toner has been a total bust. Even followed a guide that guaranteed results, but failed because the type of paper used in the US is different in the EU! It was exactly the same brand, name, wrapping and so on, but the EU version had a waterproof coating duuuurh.

    • Zee says:

      I’ve had similar issues. The only method that does work is permanent marker direct drawing. Toner transfer is a total bust.

      • thexlc says:

        The trick I have found with toner transfer is using really glossy paper (130gsm lumi / core digital gloss) getting it REALLY hot (I use a garment heat press for printing tshirts) and cooling it immediately in water. If I miss the last bit the toner just lifts off again as the copper retains the heat but the paper cools.

        • I once tried toner transfer, before getting fed up with it and setting up a UV exposure arrangement (thanks in part to a thrown-out UV exposure box in the university skip). By far the best is printing on to silicone paper. Tape a small piece of silicone paper to a sheet of normal paper and feed it through the laser printer (usual settings, maximum toner etc). This will then iron on beautifully and release well. With sufficient toner, you don’t get any pinholes.

          It’s also great for doing silkscreen markings, applying decals to wood, or anywhere you need a design transferred.

      • natsfr says:

        There is a special paper called PNP Blue, the cost is high (25$ for 5 A4 sheet) but it’s really effective, I use it when I don’t want to mess with UV.

  2. James says:

    Missed Blondi’s somewhat unorthodox vinegar+peroxide+salt recipie: http://quinndunki.com/blondihacks/?p=835

    As for tonor transfer, I’ve been experimenting with contact book covering polypropylene film (“duraseal” we would call it here in NZ, it is commonly used to cover kid’s school exercise books) with varied results.

    I follow the following procedure
    1. print the circuit on plain paper to establish the position
    2. stick a square of contact film on some paper where the print will be
    3. print it onto that paper with the contact film square
    4. put the print face up on your desk
    5. put your cut-to-size copper board face down ontop of the print
    6. fold the paper around the backside of the board like you are wrapping a parcel, this works to keep the paper in position and pull the image tight to the copper
    7. flip the wrapped board over so that the coppyer+image is topside and apply your iron, just heat it up and carefully slide it around
    8. after it’s good and hot, carefully pick up your parcel without burning yourself and drop it into cold water
    9. let it soak, you’ll see the paper becoming transparent and hopefully will see your tracks starting to show through, give it a very gentle swishing to help
    10. once it’s cooled off in the water, lift it out and carefully unwrap and peel off the film, if your heat was even and the right temperature, and you didn’t smudge it by moving the iron too hard, the film just peels away and the tonor is stuck firmly to the board.
    11. if it didn’t work, wipe the board with acetone and make it nice and clean, perhaps a little scrub, and adjust your technique, temperature… and give it another try.

    it did require a lot of trial and error, but hey, the journey is half the fun. Works better with small boards.

  3. Juan Felix Mateos says:

    Use CPM6.3 and you will forget toner transfer problems (I do not work for MagicTouch but I really love this paper). I wrote this tutorial (spanish):
    https://docs.google.com/document/pub?id=1MFy5h58DiJeqi9iMTQJrGfTT6UpDUsZ5sjgeuEHEyZ0

  4. Skylark says:

    If, for whatever reason, you need an alkaline etch, you can also try the alkaline solution of ammonia with oxidizer (hydrogen peroxide is an option). It’s not very efficient, but it is a very mild etchant and somewhat selective in the metals it picks up (copper and silver etch quite fast, but for instance zinc and nickel are much slower).

    Make sure whatever you use as mask will not etch. For instance, photopolymer masks are usually stripped with alkaline solutions, so this etch wouldn’t work. On the other hand, a nickel hard-mask (selective plating through photopolymer, strip photopolymer, then etch around the plating) works very well. It all depends on what you need.

  5. qwerty says:

    According to the following video, copper can be etched via electrolysis by using a copper sulphate solution, without the need of dangerous products and/or the risk of ruining sinks/piping . Copper sulphate is commonly used in agriculture, is safer and cheaper than other products, hence my interest. I couldn’t find more references to its use in this context though.

    • James says:

      She is etching a solid copper piece, this wouldn’t work well for PCBs I think because afaik you need conductivity to any area that will be etched away, the etching itself would break conductivity to other parts which you wanted to etch.

    • mental2k says:

      I had an idea about the reverse, though only in the vaguest sense of the word. Namely draw on the circuit with something conductive and plate the the copper onto the board. Never got round to thinking it through any further, but I’d love to know if anyone can further furnish me with the impossibilities/drawbacks of that method.

      • James says:

        The medium you use to draw the conductive traces is still going to be there after plating, between the copper and the board, so that is what is effectively adhering the copper to board at the end of the process. Maybe if you used a strong glue (that wouldn’t fall apart in the process, perhaps a UV cured thing) to draw your traces, and then coated it in graphite powder, dried/cured it and then plated.

        But the bigger problem is that of course all the traces you want to plate need to be connected to the cathode/anode (whichever it is, I’m not an electoplater) so either you have no isolated traces (I guess maybe you could cut the “bridgework” later), or you need to wire up each trace to the power rail individually.

    • Will says:

      Stainless Steel is a bad choice of electrode it tends to put chromate ions in solution that are fairly toxic. Also NOT green and illegal to dump down the sink.

  6. Torby says:

    I’ve used toner transfer for years with the blue paper from Pulsar. Forget the iron and get a laminator. Recently, a friend at avrfreaks put me onto using vinyl sticker material, and I’ve gotten some very nice results with that.

    http://dangerousprototypes.com/2012/09/29/pcb-manufacture-using-oracal-vinyl-film-laser-printer/

    I’ve been using copper chloride, but I’m contemplating going back to ferric chloride as copper chloride seems to take about twice as long to do its work.

    Here’s a recent board: http://outsidetrains.com/mls/muchbetter.jpg

  7. Quinn Dunki says:

    Thanks for the link James! The very simple and effective vinegar/peroxide combo never seems to get any love. It’s a shame because although I certainly didn’t invent it, I use it all the time.

    As for toner transfer, my vote on that one is still “don’t bother”. Print your pattern to a transparency, stick it to a presensitized PCB bought from Jameco, and leave it under a desk lamp for ten minutes. To each their own, but I don’t really understand the appeal of the toner transfer method. UV exposure is easy and basically idiot proof. I’ve literally never had it fail to work first try, and I only started etching boards about a year ago.

    • I agree – although it’s maybe a little bit messier to set up initially, and the presensitised boards are more expensive, it’s a lot more convenient in the long run. I did a test once and managed to get 4 thou tracks even with just a laser print as a transparency. One trick I learned from someone is to double up the transparencies, when using a laser printer – it really helps to ensure a good black out.

      • Quinn Dunki says:

        Yes, I second that- I always print two copies and stack them up. That guarantees there are no light spots in the print. I also use an old picture frame to clamp the transparency to the board. That ensures no light leaks underneath, for the sharpest possible edges.

        4 thousandths is impressive! I’ve been using 10 thou traces, and they’ve been very reliable for me. You’re tempting me to try something smaller now…

        • I wouldn’t actually trust anything as thin as 4 thou ;-) The theoretical minimum is twice the copper thickness, which for 1oz copper is about 2.6 thou, so 4 is pushing it. Board houses seem to usually spec 6 thou as the minimum. Like you I have 10 thou as the absolute minimum, and usually go for 12. You can do tricks like passing two traces between the pads of a 1206 component, which is handy.

          Here’s what I use for aligning and exposing double-sided boards, if it’s of interest:

          http://imajeenyus.com/electronics/20110224_double_sided_exposure/index.shtml

          • Quinn Dunki says:

            Heh, yah, 10 seems plenty thin to me. It’s thin enough to run a trace between 0.1″ spaced through-hole pads, which is a big win for layout, and good enough for my needs thus far.

            Thanks for sharing the double-sided information! I’ve been considering attempting that recently, since my current project may require a board beyond what I can route on one side without dozens of jumpers.

          • When using a laser printer to make transparencies (which, by the way, should be printed on tracing paper, not acetate film), I found that “compatible” toners gave a much less dense coating than HP’s own-brand toner, suggesting there is something in it. That’s partly why I started doubling up transparencies – I’m pretty sure that when I was still using HP’s toner, I only ever needed the one transparency.

          • Quinn Dunki says:

            Interesting- I make mine using the self-print machines at FedEx Office. They are laser, but perhaps better quality than a home laser printer? I print on to acetate sheets directly, and it works well. I double them up because sometimes the printers are a little low on toner or something, and a few areas are a bit light.

        • imajeenyus says:

          This lot – http://www.pcbfx.com/ – have some really interesting toner-based things. They appear to be using the toner as an actual adhesive to stick _additional_ etch-resist material to the board, probably a bit like press-n-peel. They also have coloured sheets which can be used for applying decals.

    • James says:

      The attraction, such as it is, of transfer etching for me is mainly

      1. cost – plain blank copper clad boards are cheap, photo-resist boards are less cheap, and I’m so cheap that I could hang my wallet in a tree to communicate with the birds “cheep cheep cheep”

      2. no developing

      3. no darkroom needed (but of course you can get away with a dim-room and not dilly-dallying, or hobbit blanket arrangement as you say)

      4. no uv box needed (but of course you can get away with the good old sun, or a suitable lamp)

      Of course, the disadvantages of transfer etching are that it gives less repeatable results, and it requires a bit of experimentation, technique, skill, tounge-holding-at-right-angle.

      All that said though, the availability of dry-film photo resist ( http://www.ebay.com.au/sch/i.html?_nkw=dry+film+photo+resist&LH_PrefLoc=2 ) which can be applied to your stock of plain copper clad when you need to do a photo-etch, is perhaps changing the equation.

      Certainly if you need to do 10 mil tracks (indeed 12 mil is touchy enough for me), transfer etching is really not too workable IMHO.

      • Quinn Dunki says:

        That dry film photo-resist is pretty interesting stuff. I haven’t seen that before. I might give that a try.

        Worth noting, the other advantage to a hobbit blanket is that it prevents others from seeing your ring, which you’re really just holding on to for safe keeping and can put it down any time you want to, you just don’t feel like it right now, so stop asking about it okay?

        • Dry film resist is fantastic – getting it applied can be tricky though. You can use a hot laminator – I managed to get away with just an iron. Developing is easy and safe – sodium carbonate. It’s usually negative rather than positive, so you have to invert your board designs (black = copper removed). The Chinese stuff which is around on EBay works well.

          I tried it briefly for photochemical machining of thin steel sheet and it worked really well. Baking the resist after developing gives a fantastic hard film.

          Liquid resists/home-made resists – avoid at all costs unless you absolutely must coat something non-flat. Spray on resists are so hard to get a uniform coat with, and the only home-made resist which works is dichromate/gelatin (or variants thereof). Short shelf life, messy processing, and serious disposal problems (it’s hexavalent chromium after all).

          This guy – http://dalibor.farny.cz/some-progress/ – has used dry-film resist to do some
          etching of stainless steel letters for discharge tubes.

          As for etchant, I have to say I’ll always swear by good old ferric chloride until I actually need to use something different. It’s reasonably gentle (board in 10-15 min or so) and very controllable. Slow etchants are better for fine detail.

  8. Rajstennaj Barrabas says:

    I just recently sorted out what it takes to do toner transfer in the US.

    In summary: It can be done, but you need the ability to choose from a number of options at each step.

    In other words, you need access to a couple of laser printers, a couple of methods of heating, a couple of types of likely paper, and so on. At every step you try out different options and use the one with best results – over the course of an afternoon I tuned the process to get good results. About 20 attempts in total, reusing the same pattern on the same board (ie – stop and reset before etching).

    Printers: Some Brother laser printers use toner which is outright incompatible with this process, but some reportedly work. (My Brother printer won’t transfer at all). HP printer toner will work, but the incredibly cheap HP printer models (P1000 series) are marginal. Get your friends together and try out everyone’s device and see which does the best transfer, then use that.

    Transfer: Techniques which use an iron will work with difficulty, it takes skill and you will have misfires. Fortunately you can get a laminator from WalMart for $25 (Scotch thermal laminator – model TL901 or similar). The rollers will apply even pressure at the right temperature, so it’s easily worth the money.

    Laserjet ink fuses at around 105’C, and the laminator is around 110’C, so you might want to up the temperature setting of your laminator. I took the plastic cover off of my laminator and mounted the guts openly on a board. Then I replaced the 160’C thermal fuse with a 210’C fuse (available from eBay) and put a 50K pot in series with the temperature sensor.

    My unit now laminates at 140’C, I put a board through 10 times for perfect transfer.

    Paper: All shiny paper has a “clay” coating which will take and transfer toner. The problem lies with the underlying paper, which may or may not dissolve in water. High gloss inkjet photo paper run through a laser printer (yep – it works & won’t hurt anything) will transfer fine but will not dissolve easily under water. High-gloss magazine pages will work, as will newspaper inserts and dead-tree catalogues – but you have to find one that dissolves easily.

    Try all types of paper you have access to and find the one that works best. Pages from the MSC catalogue work for me, so now I’m on their mailing list.

    (Also check the ink used. Red ink in my catalog pages will transfer like toner, so I have to be careful to use pages that don’t have red images or graphics. Not difficult, but one more tweak to the process.)

    It takes an afternoon of dedicated experimentation. Keep a notebook, try all your options, and methodically eliminate the ways that won’t work. Write down your final process.

    Lots of people have had success at this, there’s no reason why you can’t either.

  9. raaab says:

    I would not call a mix of hydrogen peroxide and acid (sulphuric of hydrochloric acids) safe. Depending on the concentrations of peroxide and acids used it can be an extremely powerful oxidiser, and should be treated with caution

    • Rajstennaj Barrabas says:

      I would not call large quantities of water safe. Depending on the depth and temperature, it can provide an exceedingly dangerous environment, and should be treated with caution.

  10. JimBob says:

    Etchant is not my problem – it’s transferring the artwork to copper clad which is. I have bought photo-sensitised boards in the past, and the whole process was simple, easy and fast. The problem is that I have to order the blank boards, which may take up to a week to arrive. Surely there is a chemist out there who can come up with a homebrew way a coating could be applied to blank boards.

  11. Mike says:

    Best paper for toner transfer for me is hp presentation paper.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 91,130 other followers