Why Are You Still Making PCBs?

Few things have had the impact on electronics that printed circuit boards (PCBs) have had. Cheap consumer electronics would not be as cheap if someone still had to wire everything (although by now we’d be seeing wiring robots, I’m sure). Between removing the human from the wiring process and providing many excellent electrical properties (at least, on a well-designed board), it isn’t surprising that even the cheapest examples of electronics now use PCBs.

For many years, the hallmark of being a big-time electronic hacker was the ability to make your own PCBs. There have been many ways that people have tried to bring PCB manufacturing into the hacker’s garage: stick on decals, light-sensitive blank PCBs, and even using laser printer toner (that last one spurred me to write a book on PCB layout many years back). You also see a lot of people using 3D printers or CNC mills to create PCBs. Hardly a week goes by that someone doesn’t ask me how to make a PCB in a home or small business lab.

My reaction is invariably: “Why?” Back in the 1980s, I worked for a company that had PCBs made and our board house was going out of business. So we bought them. They had an array of plating machines, photoplotters, and exotic chemical handling equipment. They were 60 miles from our company, and that was handy because we’d drive over carrying giant rolls of artwork directly to the board house. The cost was high, and with modern-day regulations on dumping chemicals would probably have been higher. The price for tooling was especially high. That first board cost a lot. Even the 100th board was expensive by today’s standards.

I did a lot of boards myself in those days, especially prototype boards that were likely to have issues. But they were never the same as the commercial boards. It is hard to do two sided boards (not impossible, but hard). You don’t get plated through holes, so you have to use wires or rivets to connect the sides. You may not have thought of it, but that copper connecting both sides on a commercial PCB add a lot of strength to the PCB tracks. Just like single-sided boards are easier to delaminate than a double sided board, my homemade double sided boards had the same tendency because they were basically two singled-sided boards back-to-back. You could use some noxious chemicals to sort of plate the board, but it wasn’t as good. And I never found a reasonable way to do solder mask. Silk screen wasn’t worth the trouble, although I’ve used rub on letters and later toner transfer to get a similar effect.

The worst part about making my own boards: the drilling. Those holes really need to line up right (especially IC sockets) and if you screw it up, you get to start over at step 1. Granted, you can go surface mount, but most boards still need at least some holes (even if just for vias). The second worst part was handling all the chemicals. The clear etch wasn’t too bad, but the more readily available ferric chloride stains everything! I still have some marks on my patio concrete to prove that. I never had the gall to try some of the homebrew etches because they had nasty chemicals, too.

Things are different today. You can layout your PCB totally on the computer (or even in your browser). Click a button and you can send those files anywhere in the world. There are dozens (or maybe more) board houses that will produce your board cheaply. Many of these are in Asia, but there are affordable options everywhere now. The only thing you really need is time. You can’t have an idea in the morning and look at a prototype PCB in the afternoon. But if you are willing to wait–and the wait doesn’t have to be that long–you can get beautifully produced boards at a very low cost, even if you are making only one or two boards. You’ll get multiple layers, plated through holes, silk screen, and all the other things you expect in a professionally made board.

Yes, my CNC mill will chip away copper (and end mills) and leave me with a board with no solder mask, no silk screen, and no plated through holes. I still have my tank for heating up ferric chloride (which should have the brand name Stain-it-all). But I never do that anymore. It simply isn’t worth it. You shouldn’t either. Despite the advice in the video below, the best way to get ferric chloride stains out of things is to not use ferric chloride at all!

I’m going to stop short of recommending a particular board house so I don’t get accused of advertising for any particular one. We had a Hackerchat not long ago where several people mentioned their favorites. Ask around. Or leave your favorite (and why) in the comments.

Don’t get me wrong. No one that reads Hackaday needs to be told why someone wants to build something even though they could buy it somewhere else. I do that all the time. If your goal is to learn about PCBs or hack the PCB process, then by all means, knock yourself out. But if your goal is to prototype something and the PCB is just a means to an end, dump that ferric chloride, save the end mill, and find a board house. You’ll be glad you did.


223 thoughts on “Why Are You Still Making PCBs?

    1. We had one of those in college (1982). I wrote a schematic capture program that translated the pin assignments to endpoints. It wouldn’t handle anything but DIP packages though.

      The input was gerber on punched paper tape. :)

      1. I’m switching over because I can only reuse sockets so many times, before the connections get flaky and ruin a whole board. I also can’t get sockets for 26 cents each anymore. And means of getting signals off of the boards are also expensive.

      2. Who has stopped? I’ve been wire wrapping for more that 40 years, and I still build a wire-wrap board to check out a new circuit before I commit to PCB for my customers. Now a days, I make PCB patterns to adapt surface mount parts to standard [0.1×0.3] pinout wire wrap socket spacing and keep a careful eye on ebay to snap up wire wrap panels for future use. Punch out the pins you don’t use, and heat-press them [use the tip of a soldering iron] into perfboard to make up your own sockets.

        1. Everyone who is not doing low tech circuitry, aka “LED toggling fun stuff”.

          Would be fun, so get a usb C bus working with it at good speeds, even a simple sd card or go for higher goals in life: mmwave radar (lets go for around 78GHz).

      3. Because companies stopped making the stuff for us. Wire wrap wire is getting to be difficult to find. There is china junk out there, but nothing like what I have on the huge spool of red from 1979 that I snagged a decade ago at Dayton Hamfest. The old stuff is so much higher quality than what you can buy today..

        1. I have a ton of the stuff!
          I use like a couple of inches when I’m fixing a board or making up a simple harness, but I’ve got probably 50 rolls.
          I’ve also got that sleeve for sliding onto resistor leads and such.
          I’ve considered selling on ebay but there seemed to be plenty of new unused rolls so I didn’t bother.

          1. Wow. Here (in Australia) we have houses that had a product called “Mr Fluffy” that was loose fill asbestos that was used as ceiling insulation. Those houses are being demolished because of the perceived safety risk (Definitely a safety risk but there are better solutions to demolition). So I am guessing from you response that ‘Fluffy Siding’ is a similar asbestos product? What country are you in?

      4. We are all forgetting that most electronics components are now in SMD form. I think wire wrapping simply died because of that. PCB’s got smaller, and less through-holes. So PCB’s became cheaper, and probably that contributed to the rise of PCB houses as well.

        The only through-hole stuff I use is for experimenting on a bread-board. And even for that, I often use SMD chips on a carrier board.

      5. Wire wrap is fantastic for making secure custom links between evaluation boards: you can twist it into twisted pair (hint: use a vice and a drill, and look up or experiment to get the impedance right. And don’t forget the earth return connection! lvds is good, but remember it’s feeding current down one or the other wire continuously), and fairly easily get multiple 320mbit connections between FPGA’s that way. (It’s also good for sharing a system clock around).

        Consider the usual approach of using those “push on” connections: great for a prototype that doesn’t need to move anywhere, but you end up with a rats nest of wires that you need only bump to get a fault – especially when the connectors get old – like old push in bread boards…

        Wirewrap, on the other hand, is known for the reliability of properly made connections. They’re strong too: you’ll usually snap the wire before the wire wrap lets go, as long as you’re not pulling straight up.

        You ought to have a wire wrap tool, and use it for hooking up PMOD’s to esp01 modules… 3D print some clips to let you bolt tiny modules down, and you’ve got a very flexible way to set up complex and powerful (and robust!) systems from a bunch of inexpensive modules.

        Resist the urge to put all your smarts into the one chip – instead separate systems by functionality – like how your car has multiple, completely independent hydraulic fluid systems, not one convenient place to fill a “generic” fluid, and you will save yourself much unnecessary pain keeping everything working smoothly. Squeeze everything into one microprocessor, and you’ll be spending an exponentially increasing amount of time trying to keep everything working reliably as you squeeze just one more little feature in… Unless you like that challenge, that is… But shouldn’t you be using your elite computer hardware hacking skills to control/do something that will make a difference?

        Good luck!

        1. Yes but where can you get cheap wire wrap sockets / pins / tools?

          I still have lots of 30 gauge silver wire that I sometimes use to repair broken tracks / vias on PCB’s. I am tempted to make a tool for wrapping standard pin headers.

          Wire wrapping was fast development as you can make changed without starting over with another version of the circuit or PCB.

          It would be easy to make one PCB with all the parts – even SMD and then terminate all the parts to pin headers and wire wrap from there. Would be so much faster and easier.

          1. Ordinary 0.1″ spaced headers – the ones with the square section pins, which conveniently are on just about everything – are all you need to wire wrap onto.

            It turns out that tools fit fine even into places like IDE sockets, and you can even make a connection onto 2mm headers too, if you’re careful about it, and strip the wire shorter.

            Those “wire wrap” headers with the extra-long square pins exist just to give space for multiple connections to each pin — all you really need is about 0.25″ or so, and you can use 0.2″ if you have to.

            The only thing that isn’t optional is that you absolutely need a square cross section to wire wrap onto – the edges bite into the wire and establish the nice, corrosion-free electrical connection.

            So, anything with 0.1″ spaces holes is fair game, so long as it has the usual square male pins soldered in.
            This is perfectly enough to hook up point to point links, and there’s usually enough space to wrap two wires to a single pin.

          2. I just spent hours looking for my wire wrap tool. It one of the ones that looks like a metal jewellers screwdriver. No good I will have to order one.

            Thanks for sharing your experience. Adding wire wrap to prototypes will same me much time and some parts.

          3. Fry’s sells the classic screwdriver wrap tool and an assortment of Kynar wire. The usual distys will carry these as well. Kynar is still widely available as it’s the standard for bodge wires. Sockets can be had from the same, but they’re pricey.

          4. remydyer: Exactly my thoughts. Most of my proto wiring these days is header-to-header, not component-to-component. And most breakout and eval boards have everything come out on square post 0.1″ headers. And in cases where I need to tack a wire to a tiny via or test point, well, Kynar is great for making fine solder connections.

        1. I am going to order a wire wrap tool and give this a go. I will be using standard 0.1″ headers thanks to [remydyer]’s advice. I will get the right angle ones and reverse every second pin to make it SMD. The tools are still made and places like digikey and mouser have them.

          Keywords for google –
          wire wrap strip unwrap tool
          wire wrap tool prototyping

          I even saw a youtube on how to used the innards of a Parker pen refill though this is not quite the same – it will wrap but a proper tool also crushed the wire into each of the four corners of the wrapping post.

          1. Googling with the added ‘prototyping’ filter did the trick, cheers!

            I have already hoarded a lot of wirewrapping stuff from fleamarkets etc, a ‘gun’ working at 42 (!?) volts being one of my newest finds. All this old stuff looks impressively professional, even decades old DIL sockets…
            So impressive I wanted to start properly,’ with these fine specimens :-)

          2. If it’s in the 40+ volt range then it will be 48 Volts as that was the standard for telecommunications equipment and wire wrap was used a lot in telecommunications. The standard wire gauge in telecommunications was much larger though, about 24AWG while the prototyping wire wrap wire was about AWG30 – 31

  1. Dear hackaday please get rid of your newer I’m an edgy teenager writing in angst writers with BS opinions and go back to posting about hacks instead. Recently you have done pretty much nothing but post very opinionated posts instead of actual hacks. some examples and why, the pi/slr lens hack the author made the snide comment “is it practical” then went on to list something else for no reason at all instead of just leaving it. The “This is what a real bomb looks like” post decently authored but at the same time definitely titled and timed to be about the kid with the alarm clock he disassembled. again trying to be edgy. And yes even the kid who disassembled a clock and put it into a case in which 99% of the world who would see that would be initially concerned about what it actually is no matter what his skin color is and the proceeding make a clock articles. Please go back to posting and stop being edgy social warriors because you will lose readers and project submitter including this one.

    1. I am biased because I sell a product through HAD, but I like the new writing going up on the blog. They still post a ton of hacks, but now they mix in a lot of news, opinion and educational stuff that adds depth and variety to the blog. I also doubt they are losing viewers, in fact I would bet posts like these are attracting all sorts of new followers.

        1. Well, you’re paying for it, so you of course have every right to ask for exactly what you want, and nothing else. The gall of them to force you to read articles you don’t like on the internet!

          (To HaD, this is a very useful post for me, as I have been debating between outsourcing and milling).

          1. There’s no such thing as a free lunch, it just means someone else is paying. In this case, the advertisers that pay for eyeballs. HAD ignores its reader base at it’s own peril because without the readers, they have no income.

            The new editorial voice is definitely over-opinionated and the focus away from hacks and more towards social issues rubs me the wrong way, too. Leave the worthless, smug, over-opinionated snark to the commentators!

          2. Same here, I’ve been debating if its worth getting the tools to manufacture PCB’s at home, stick to breadboards for single projects, or get boards manufactured…

            I think this article clears that up for me, learn PCB design and get them made for me!

    2. Opinions from experienced people are valuable, and save time. I’d be pissed (already am from boards coming in from boardhouses) diagnosing connection issues on a homebrew board when I could be programming…

      1. “But I never do that anymore. It simply isn’t worth it. You shouldn’t either.”

        That sort of thing isn’t an opinion. It’s the words of an elitist arsehole who thinks they know what other people want, need or value. There’s nothing wrong with opinions, but these are poorly written opinions.

        1. Exactly. Also, I’m more elite than the writer, since I don’t find it hard at all to do two sided boards. I’ve even done multilayer (ok, just two two sided ones separed by a naked board). Granted that was more out of curiosity to see if it’s doable (it is, if you are in the middle of zombie apocalypse and really, really need to do one). But two sided? Why would it be hard? Start with a two sided PCB, tape the transparencies together as a sleeve (leave margins), expose the sides either one by one or both sides at once, and etch just like one sided board. Drill holes. Cut TH components legs to create vias. Very easy. The multilayer one, now that was just stupid. Perfectly doable, but very easy to make mistakes in the design phase.

          1. I am looking at using coating brass shim (later copper shim) stuck to Kapton tape and using my laser engraver I made to remove the coating, if it works as I expect, I should have no problem doing not only dual sided boards but multi layer boards that are also flexible, and with thin enough gaps between them to easily connect the sides together, wont be plated but still. also as an aside, I have already used it to melt solder and it does it with enough energy, that there is zero smell from the broken down rosin flux core that is normally smelt even on hotter soldering units I have, so should be no problem to solder just about anything that is not hidden behind a board, but popping them from behind through a tiny via hole should make vias even more needed than they already are :)

        2. He got a little cavalier with that line, but it’s a bit overly sensitive to call him an a-hole for it. Making your own PCBs is **usually** an irrational choice, as much as it pains people to hear it. There are a few scenarios where it makes sense 1. if you enjoy it 2. if you can’t wait 3. if shipping is expensive or unavailable to you.

          1. Exactly, and he pointed out that a board house doesn’t make sense if you absolutely can’t wait.

            Frankly, I love this kind of article with useful advice. I don’t need the author to tone down his opinions to “just the facts” to be able to learn from it, and without this kind of advice, I could spend months* chasing etching parameters when I could be spending my time actually getting things to work!

            *Not that it takes months to get etching to work, I just don’t have much free time.

          2. Every time I consider making my own PCB, I add up the supplies I’ll need and the time it will take. Inevitably I ship it to OSHpark and save my money and frustration. The only case I could make is if I need to do a large board, but my projects that need a large board are almost always doable with perfboard and point-to-point.

        3. Or, alternatively, what you just said is the words of an ignorant asshole who didn’t even bother to get to the end of the thing they are trashing because climbing onto your high horse, smugly proselytizing and manufacturing outrage in a vain attempt to make yourself feel like you have any self-worth is far more important than being right is.

          Note: I’m being kind by calling you an ignorant asshole. The alternative is that you read the last paragraph and decided to ignore it. Which would make you a lying, manipulative, stupid asshole. That’s far, far worse.

          “Don’t get me wrong. No one that reads Hackaday needs to be told why someone wants to build something even though they could buy it somewhere else. I do that all the time. If your goal is to learn about PCBs or hack the PCB process, then by all means, knock yourself out. But if your goal is to prototype something and the PCB is just a means to an end, dump that ferric chloride, save the end mill, and find a board house. You’ll be glad you did.”

          1. That last paragraph doesn’t change the fact that he comes across as a sanctimonious prick who can’t/won’t for whatever reason and doesn’t think anyone else should either.

            Much of the article is simply attempting to justify that opinion, even to the extent of falsely claiming that “It is hard to do two sided boards”.

        4. Yeah, that’s the problem. Opinionated assholes who are often wrong. I’ve done double sided boards, no problem. No FeCl for etching. A few vias? no problem. It’s inconvenient if you have to drill tons of holes by hand (surface mount is your friend). It depends on your priorities! is the inconvenience worth it? do you need to test several components before a final design in a week? then you may want to do the boards yourself instead of taking weeks and weeks to change the design and wait for new boards and spend a ton of money in very small orders and shipping!

          1. Wow, just wow.

            I won’t be giving up making PCB any time soon to. I was easily able to express a difference of opinion to [Al Williams] without describing anyone with derogatory terms.

            Seriously, I have enjoyed many articles written by [Al Williams] and don’t see why people need to ‘go off’ when they have a disagreement with one article.

          2. @ROB

            Wow, just wow.

            I won’t be giving up making PCB any time soon to. I was easily able to express a difference of opinion to [Al Williams] without describing anyone with derogatory terms.

            Seriously, I have enjoyed many articles written by [Al Williams] and don’t see why people need to ‘go off’ when they have a disagreement with one article.

            Indeed. This one was well below his usual standard, consisting mostly of rather badly phrased opinion which left him looking rather bad. But still no reason to call the man an (arse|ass) hole.

    3. I disagree. I like the new posts. Sometimes I learn something, sometimes I don’t. But keep posting them.

      I have no idea if they are gaining or losing readers, but I’d bet on the former. Anyone who wants to read only the hacks, go for it and just skip the rest. It’s just pixels on a screen so it’s not like it’s filling your house up with unwanted paper or anything….

      1. I agree that editorial articles are a great idea, but maybe they could improve on making them more professionally written and well researched rather than this “top of my head” type journalism. To take this article as an example, he could have written about the advantages of having your board manufactured vs doing it yourself without hyperbolic over-the-top language that sounds very immature. Even the title is ridiculous.

      1. I have ADD Leon. Kind of rude of you to make fun of an illness. I don’t know what that has to do with the article. You should feel shame for yourself and family. I hope God will forgive you. Oh and aren’t you @lehjr1 on twitter

      1. Agreed. There are people building things with the arduino platform these days that could be done with a very simple 555 timer and some ingenuity. Micro controllers are truly incredible. I love them. I don’t need one to blink a light. I might need one if I’m controlling motors or servos. Some of the best uses I’ve seen with those avr’s are 3D printer controller boards. That’s really maxing the speed and memory.

        1. Some people like walking uphill both ways, that’s cool. Working within constraints rather than throwing money at a problem is a great skill.

          But for the love of acrylonitrile butadiene styrene, please stop designing AVRs into 3D printers. A system that needs to maintain communications throughput while doing math on incoming data while maintaining real-time control over motors and potentially dangerous thermal loops is the wrong time to play demoscene coder. NXP makes a dual core ARM that retails in 1x for ~$11 that has all the resources a modern 3D printer could need, but people keep supporting the Arduino / RAMPS / Marlin stack and wondering why their printer stalls when doing high vector density contours.

    4. Due to your poor use of punctuation and generally terrible grammar, it took me several reads to understand your argument. At first I though you were claiming to be an edgy teenager writing this comment in angst.

      Also, HaD has evolved a bit since they started. They used to just post awesome hacks. They still post awesome hacks, but also focus on building a community of education and technological advancement. Its not just a “RSS feed” for the latest cool hacks anymore. HaD has a mission. I for one agree with their mission and love the new direction they have taken. If you don’t like it, get out.

    5. that clock by the 12 year old wasn’t even made by him he took the inside out of an alarm clock and put it in a suitcase. Watch this 20sec video https://www.youtube.com/watch?v=kHk_6Vh4Qeo If someone could embed this for me that would be great as I am not sure how to on hackaday. Also there is this video that explains all the things wrong with the “clock bomb” the giuy goes into great detail https://www.youtube.com/watch?v=CEmSwJTqpgY I tryed to “submit a tip” to hackaday but I think they don’t want to look stupid by posting this evidence as no one has bothered.

      1. Just… that’s not a clock bomb, it’s a clock. Real bombs do not have a huge red display counting down, except in movies. The “actual size” of the clock is not a suitcase, the photo is misleading. It’s more like “Pencil case”. I’m sorry if you think any of that “evidence” is of anything other than a 12 year old happy to get something to work.

    1. to be fair, the quality of those purple boards is *excellent*
      I still order from dirtypcb’s if I can wait, but my experience with the purple board guys (OSHPark) has been excellent.

      I’ll add another one to the pile, we had a 3 day turnaround with advanced circuits for some PCB’s. They had no mask on them, but the copper was tinned nicely. they soldered beautifully.

      1. I have had very good experiences with OSHPark. They always meet their deadlines and are very reasonable on top of the fact that I’ve never gotten a defect.

        It helps that I like purple, regardless of the fact that all of my projects are enclosed or potted.

        1. Board color is no big deal as long as it isn’t black. It’s really hard to trace circuits. If I’m prototyping a board it’s usually inside a project box or 3D printed housing where I can’t see it. If I could get a board any color I like, I’d like to see some orange boards.

          1. I’ve found that out of all the common solder mask colors, yellow provides the best circuit visibility. It tends to be the most transparent, and the copper has high contrast against the laminate substrate.

            I’ve been looking for a boardhouse which can provide clear soldermask, just for a neat look. The mask coating supposedly starts clear, with pigment added for color, so it should be possible. It’s just not a common drop-down option on the western-facing websites I’ve seen.

          2. Yea I’d not like orange myself for the reason that rob mentioned, but yellow or bright blue would be good. Green is also good for that same reason but it is cliche.

          3. FYI: https://forum.pjrc.com/threads/24950-Teensy-3-1-Changes-To-Green-PCB
            >Over the last year and a half of making Teensy 3.0 and 3.1, we’ve been working on improving the quality and production yield. With any PCB, a small percentage fail. We test every Teensy throughly, so any shorts or opens are detected. The bad ones are saved for diagnosis and then we just toss them. (no, do not ask if we’ll sell the bad boards at surplus prices, because we won’t ever do that)

            >It turned out some of the failures could be traced back to the black solder mask. What seemed like it ought to be the same, just a different color, in fact isn’t the same at all. The black material just doesn’t have quite a high a resolution as the green. I was very surprised to learn this, but it’s true.

        1. I’ve done 7 bpards through dirtyPCBs now and all of them have come out great! I made my CNC mill 5 years ago to produce my own boards but in the last year and a half I’ve hardly touched it. The boards are just so much easier to do through dirtyPCBs and I can have a much tighter layout.

    2. Faster turnaround, tighter tolerances, and better quality results. Why do you miss batchpcb when their successor is better in every metric that matters? And what do you have against purple boards?

    3. How does one learn how to design a PCB? I am working on a 1, maybe 3 off Raspberry Pi HAT. Its literally just 5 parts, 2 of which are resistor and cap. I’ve tried online options and Eagle, and I am just not understanding how to lay out the traces, especially where they will overlap.

        1. I second the KiCad recommendation. I’ve used Eagle and KiCad, and I prefer KiCad, mainly because of the lack of limits.

          While there are guidelines and methods for laying out traces, I’ve found that it’s mainly an experience thing. After you’ve done it a few times, you get a feel for how to position components and run traces. Practicing with a small project with just 5 parts is a good exercise as you can easily rip up the traces and move things around without having to reconnect, say, 34 components.

  2. ” But if your goal is to prototype something and the PCB is just a means to an end, dump that ferric chloride, save the end mill, and find a board house. You’ll be glad you did.”
    I will be glad I did… next week. Nah, I want my boards now. Couple hours beats waiting for board house every time – if the board is within my process capabilities.

      1. Or citric acid, table salt and hydrogen peroxide. Cheap, available in any store + pharmacy.

        Also – in my case – any Chinese board house = ~1 month wait. If the post office or customs feel bitchy – 2 months.
        Local are pricier, but it would be about a week and going to rather inconvenient part of the city.

  3. I will name names! Sunstone is *awesome* if you need boards fast. If you get the file uploaded in the morning, you can have a board less than 24 hours later. You’ll have to wait an extra day if you need silk or resist.

    I’d also like to add that for simple boards, a CNC cutter like the Othermill can be great. If you can design with mostly SMD components and plan your sides carefully, then though hole plating is not an issue. Really great for vreating on-the-top tools like breakout boards and programming fixtures too.

      1. Exactly! No one is stating their country so you have to assume that they’re in the US and in that case the info is useless to me as I am elsewhere.

        So – good article and good intentions but for anyone not in the US … useless info.

  4. I have used a few different places in the past, but my current favourite is DirtyPCBs. Price is right; you have options that you don’t have at other places such as panelization, colors, etc; the interface is nice and simple and made by and for hackers. Yeah the wait is annoying (I am waiting for my latest batch of boards right now, in fact), but it’s not the end of the world, and I could always fork over some more money for DHL if I really wanted them fast. (The manufacturing does not take that long, in my experience; just the shipping is slow).

  5. Well one great reason is, someone turned up at my house on Saturday morning with an idea and left with a functioning prototype a few hours later. Also they’re good for fitment testing, mistakes, unforseen things. Then i’ll send it to a board house.

    It’s fun for people that haven’t done it before, the problem is when you’re an old hand at something, you’re often jaded about it.

    Lets outsource everything then we can just sit and binge watch netflix waiting for DHL to arrive.

  6. I still use a PCB mill for quick and dirty stuff like fixtures where I need to hook up a connector to a couple of through hole pins. Stuff like that a PCB mill is good for, but I couldn’t agree more that for layouts of even minor complexity (fine pitch, lots of vias, two layer) just order the boards! The one thing that is still worthwhile to do yourself is hand solder your own parts, it is expensive to setup pick and place when you only need a couple of protototypes.

  7. Some of us just enjoy making something by hand that could easily be mass produced. I’m sure once I can do it easily I might get bored of the creative process, but right now the “can I make a good quality LQFP board?” is as much of a fun challenge as designing the circuit or coding for the microcontroller. I don’t delude myself that the price or turnaround time makes much of a difference. Once I don’t enjoy it then I’ll stop.

  8. My 2 cents. I still make PCBs at home because unlike north america/Europe, getting PCB done at a fabhouse is still an expensive proposition. And good luck if you want them within a day or two, you will shell money like there’s no end.

    I’m yet to find any decent and pocket friendly PCB service in India.

    1. PCB manufacturers in my country charge $75 for any prototype. So I tested manufacturers from China, which is cheaper, around 10-20USD depending on size. Downside is waiting times. Still this is too much for me, because I earn about 300USD per month, and more than half of it I spend on bills (and food isn’t cheap either). So I must etch my boards at home.

      Pro tip: use sodium persulfate instead of ferric chloride. It’s slower, stains clothes, but won’t destroy anything else.

      1. I am using Ammonium Per-sulphate now and it works well. It needs to be heated to 70 degrees Celsius though.

        I used to use Ferric Chloride but it is hard to get here now as it has been declared a dangerous chemical for transporting.

        I never had any issue with stains. Ferric Chloride is actually quite dangerous to contact so I was very careful with it – never any drips or spills. I treat Ammonium Per-sulphate with the same care.

        I still think the inhouse PCB is ok for some projects. A quick breakout board or a single sided PCB (maybe with links on the other side) perhaps for power regulation / control.

        But double sided PTH with fine pitch SMD is for the board house.

        1. I’ve used Ferric Chloride, messy and hard to see the board etching; Ammonium Persulfate, hard to get and hard to use; but now I use 50:50 Muriatic Acid (from Lowes/HD) and Hydrogen Peroxide (from wherever) at room temperature. Remember AAA, Always Add Acid. You can see your board etch and it doesn’t stain anything, except from the green copper color that comes from etching.
          I like 2-hour prototypes, but Sunstone does a real nice job for production stuff.

  9. You can put it even on a higher level. Not only you can outsource the pcb manufacturing, but the PCB assembly, too.
    With Seeeds Open Part Library and assembly service it’s very easy to get a one off PCB assembled without all the hastle sourcing, chipping, quoting. While I asked 4 companies for a quote for two weeks – I got my PCB assmbled and delivered in 4 weeks.
    This makes it very simple if you are using components they offer in their library.

    The price is a different question, but I would not have done it faster or cheaper for a single piece.

    There is only one draw back, if a component is not in their list, then you still have to do it traditionally. But BGA & Co are anyway not on their targeted cheap PCB road track.

    I would wisch other assembly houses would do the same. Open your list of parts. which are keeping anyway. Let the people use parts from this and give a website to calculate the price from a BOM. Reject all parts, which are not matching unavailable and call the job done.

    1. You can put it even on a higher level. Why not just describe your idea to an engineer and let him design and implement the circuit? You can have it for a very good price and you will save a lot of time.

      1. And why you’re at it, why not just throw some money at people with ideas instead of thinking one up yourself?!
        There is a point where outsourcing production makes sense, but it’s far beyond anything we’d identify as “hacking” or even prototyping.

  10. Well, if someone can recomend some good and reasonably priced board house in the Americas, that would ship to Brazil, I can give them a try. Otherwise, the reason for making our prototoype boards in house is so that boards do not take 3-4 months to arrive, and get lost in the mail never being delivered.

  11. I average about 1 board order a week through iTead Studio. With the exception of Chinese holidays, they are turning orders from clicking send to DHL ringing the bell in about 8 days. And the order that just arrived this morning is for a commercial product, 4 layers, .8mm thick, ENIG, black solder mask, and I’m pushing their DRC. So not typical hobby specs.

  12. There’s a price comparison site for PCB manufacturing. It is http://PCBShopper.com and it includes 25 of the best-known manufacturers in North America, Europe and Asia. You enter your board’s specs and it immediately gives you prices and delivery times.

    Users can also read and write Amazon-style reviews of the manufacturers.

  13. I’m not a fan of the increasing number of provocative editorials on HaD. Everyone has opinions; what’s significant about this one to elevate it to article status? The headline is a disingenuous invitation to discourse; by the end, the tone turns discouraging and offensive to anybody who doesn’t share the author’s priorities.

    Home PCB manufacturing isn’t the obvious leg up it used to be, when domestic board houses were the cheapest option and a prototype would still incur a high per-unit charge on top of a stiff setup fee. But fast prototyping on a shoestring budget is still valid for plenty of scenarios, even if they’re not relevant to you. Over the years my spare time has dwindled, while disposable income has increased. It’s easy to back-burner a project for a couple of weeks until a $20 PCB order arrives. If I were a teenager with little spending cash, I would definitely be making my own boards.

    I’m not sure it’s responsible to pressure novices toward a 3-copy OSH Park run for a single board which will probably need rework or respin, multiplying costs and adding discouraging wait times.

    Many of the process drawbacks the article cites are valid, but I have to say: don’t like ferric chloride stains? Quit splashing it around! I use a small acrylic tray and not more than a couple tablespoons of etchant for a typical board. When the board is finished, the etchant goes back to a storage bottle. I rarely spill a few drops, and then only into the sink where it’s quickly diluted with water as everything is washed after use. A liter of etchant lasts for years. I evolved this approach for tinning solution, which is somewhat more toxic than FeCl3; but the same careful, sparing approach is appropriate for any chemical.

    1. You know, now that I look at it, most of my younger cohorts share the kind of blase’ attitude expressed in this piece. It’s almost like digging into the nitty gritty of electronics is no longer as fun, but more of an annoyance, because “there’s somebody else who does that”. People just want what they want, and they figure the “details” can be outsourced to the manufacturer of a product being used for a hack. I think part of this is simply that electronics is evolving into something much more complex than arranging individual parts in a circuit on a board. We now have amazingly cheap and complex ICs, tiny computers that can do what giant boxes of parts did 20 years ago, and hordes of advanced reusable software, so just the thrill of operating at a higher level and getting results that put people seemingly farther faster is driving some of the thinking. However, if no one ever learns the micro, component-level stuff, who will? Someone in the country selling these PCB services, no doubt, who is more than happy to sell their technologically advanced product/service. My concern is that as a whole, we’re losing these abilities as a populous. Sure, the internet will tell us how to do it if we ever need to learn, but we don’t want to learn right now – there’s a system I’m going to put together with a RaspBerry Pi, an ESP8266 with a huge library I downloaded in 3 seconds. It’s good that people are advancing by learning higher level concepts, but I hope a segment of our ranks keeps the lower level stuff just as alive and flourishing, despite it’s seeming “boring” nature. The thinking that outsourcing will offer “freedom from the details of making something” is actually replacing the joy of learning the details. Sigh. Perhaps its a sign of the times.

      1. This could potentially create more job opportunities for the people who do the “boring” stuff, as you put it. It’s like the drone thing but on a far larger scale. If it’s easier for people to get involved then that means someone’s doing the hard work. More people spending money in this particular sector. Of course that opens it up to a bunch of buttholes and morons but that’s the price you pay for having a large market.

      2. I think there is one difference between people just starting out now and people 20, 40+ years ago in the US: the median income just don’t buy much space anymore unless you are lucky enough to find work in the right areas. So while many might have been able to own a house with space for a small workbench in the garage or office, that will be many years off for most people now. That means that people really have to Choose their Battles on which tools and projects they can tackle.

        That’s just the dilemma I have right now when I came in to read this article. I would like to do some board fabrication myself, but the space that everything requires takes could mean other tools I need a lot more right now; like a small drill-press (which is pretty well a prerequisite to easily making boards anyway).

        I get people’s point on the tone of the article, but I also I think it’s important not to overlook changing circumstances in how projects might be approached.

        1. I’m still cutting my teeth on electronics and come at it from a software programming background. Just got an Internet-controlled lamp (via an ESP8266 and SSR) up and running last night. As someone trying to make the leap from breadboards and perfboard to PCBs, I’d say another difference that the last 20-40 years has made is the massive range of components available today. Looking for a new component is the most daunting part of the game for me. Frankly, Digikey is a nightmare for someone new to electronics, not because there is anything really wrong with Digikey’s website, but just the massive range of parts. The learning curve one must climb from, “I need a transistor for this project,” to “My transistors are headed toward me in the mail,” is a steep one.

  14. “But I never do that anymore. It simply isn’t worth it. You shouldn’t either.”

    Please don’t speak for me. You have no idea what I value, and a quick hint: waiting 3+ weeks for a circuit board I could have whipped together in 2 hours is not something I value.

    I still make my boards because I have something that no other board house can offer… despite your assertion otherwise.

      1. Kind of the point really. If you want to make assumptions about your hypothetical readers then maybe it helps to understand your hypothetical readers. You know the kind of people who hack together things, who do anything they need to do to get something done now with what they have on hand.

        I represent the hypothetical reader on this site more than the writer seems to realise. And dismissing any form of construction method generically for any reason without knowing the application should make people turn in their hacker card and leave in shame.

  15. Well, I’m going to try this again but without any links… Apparently, that gets posts thrown out w/o warning here? I’ve been using DFRobot [dot] com for all my PCB fabrication with great success. They’re super cheap and don’t charge an arm & a leg for different colors like some others do. Their support is also unbeatable.

  16. anything over 10 boards and I’ll consider a board house. I only do 5 or so at a time, so gunning up the CNC mill is fine for me. I try to avoid surface mount if I can because I’m better with thru hole. Using the cnc to cut the holes and mill the traces gives me a great result, and I’ve figured out how to do two sided boards with no problem.

    Still trying multi-layer with my scraps….success eludes me there.

  17. Err.. What?

    The absolute minimum cost to get a single 50x50mm board from say dirtypcbs.com comes in at around £18 here in the UK.

    I can get a 100mmx150mm sheet of copper clad board for £1.50 including postage. The parts for an entire project often come in at less than £18. I can get a whole Raspberry Pi A+ for £18

    Maybe I’m missing something, if anyone knows where I can order a single small custom PCB for £1.50 I would love to know.

    1. Of course it is impossible to get cheap PCB prototypes from any board house. There has been lately a very strange approach on Hack-a-Day: Do not hack it, buy it! And recommending a solution 2-10x the price of DIY solution. Nearly everything in here is kind of a toy that can be replaced with a ready made alternative that you can just buy and use. I don’t understand what’s the point anymore. Is this a site for hackers and makers or *consumers*? I personally do not have to be reminded that I do not have to make something by myself but can buy it.

      1. “Don’t hack it, buy it!”

        I think this is the elephant that has been in the room for the last couple of years as we’ve witnessed the “rise of the maker” culture. “Hacking” doesn’t mean anything anymore. It’s all about buying premade stuff, using ready-made services, and haphazardly slapping together a bunch of libraries to copy how someone else made something.

        I’m not sure why anyone is even surprised at this kind of article coming from Brian or the resulting response he gave when faced with reasonable criticism, this is what he does.

        Oh and this article is now featured on slashdot, now with inflammatory commentary on the way HAD readers have responded. I bet that brings a lot of clicks in, gotta make that money somehow.

    2. What’s your time worth? If you’re getting paid minimum wage, 3 hours of your time will add up to £18. Personally, I expect more than minimum wage, and would rate the value of any boards I produced myself at a lot less than those I got from a board house in any case.

      1. Well everyone’s time is worth a different amount, which is what makes the blanket “You shouldn’t either” statement ridiculous.

        Regardless of what my time is worth though, electronics for me is a hobby that I enjoy in my spare time. If I had a choice between buying an expensive 100 page novel and a cheap 300 page one, I would be unlikely to choose the shorter more expensive book because my time is valuable.

      2. What’s your waiting time worth? Or your customer’s waiting time? If I have a design in-hand, that can be easily produced in-house, I can have a board finished in a couple of hours or less. The best I can do with a random board house (Dirt Cheap PCBs) is over a week and adds $65 to the cost.

        I’m just a hobbyist, so I’m my own customer. As someone new to electronics, I screw up a lot. It usually takes me at least a couple of tries to get anything right. Adding three weeks of waiting and $130 in out-of-pocket expense hurts. For prototyping a simple design, I would have a hard time justifying a PCB service.

        On the other hand, once the prototyping is done and there is no rush, waiting 2-9 weeks may not be a big deal. Certainly, paying as little as $14 is nice… Actually, that’s amazing. In fact, discovering that $14 figure has convinced me to take the plunge, starting with some PCB business cards.

        Were I doing electronics as a business, or as part of a business, and had more work than I could handle, I would leverage a PCB service like crazy.

    3. For many of the cheap board houses (e.g. Elecrow, dirtypcbs) you get “8-10” boards for your $10+shipping, not one, just because of the way they manufacture. The boards will be tinned, solder-masked, silkscreened and drilled.

      Obviously if you don’t need the other 7-9, that’s waste, but if it’s something you ever wanted to make more of, it’s not. Plus the time spent doing something more interesting that drilling all those boards may make it worthwhile. I only have a few hours each week to pursue my project, and spending it on something more creative than drilling holes is appealing.

      I didn’t see a mention of pcbshopper yet – good for shopping around.

  18. Yeah, from the summary I was hoping for an insightful advice on some genuinely refreshing alternative technique, like smartly cheapened (semi-permanent) breadboarding or some all-around better prototyping process, not just this “order your PCB” — most of us are kinda’ familiar with that idea already, but thanks anyway.

  19. I make my own boards because it’s a challenge and I enjoy accomplishing things – which is the whole point of it for me. There’s a shorter turnaround time, you can address rev2 design issues immediately and not after a 50 panel run. You don’t have to pay ridiculous fees for not having the “right” 5000 reel for the one-off pick-and-place they use. You don’t get board house manufacturer marking killing the aesthetics of the silk-screen. If you are doing one-off prototype boards, you’ll be spending ten times the cost of doing it yourself. NOTHING that the production houses do is outside the capability of the motivated hobbyist or even the motivated limited producer.

    I have learned to never fully trust others to do the work that you are perfectly capable of doing yourself. They ALWAYS screw something up, somewhere, at some point.

    What the hell happened to artistry? It’s all “give me my ready-made product”… it’s sad.

    1. “You don’t have to pay ridiculous fees for not having the “right” 5000 reel for the one-off pick-and-place they use.”

      What does that have to do with ordering PCBs?

      “You don’t get board house manufacturer marking killing the aesthetics of the silk-screen.”

      Right – because there is no silkscreen, or soldermask.

      ” If you are doing one-off prototype boards, you’ll be spending ten times the cost of doing it yourself. ”

      Only if your time is worthless.

  20. I like the foray into editorials, but they could be a bit higher quality.
    This one is pretty lame, because it doesn’t really discuss WHY people shouldn’t do home-etching any more. There is an abundance of reasons, but there are also some pretty good reasons in favour of home-etching.

    One of these reasons is physical size. Cheap board houses heavily penalize large circuit boards, whereas at home, this is just a matter of ensuring your process is able to handle it (exposure, etchant tank etc.). There are some cases where you need a large, but sparsely populated board, a use case perfectly fulfilled by home-etching.

    Time-to-board is another reason. No matter what they do, a Chinese board house can not produce and ship a simple PCB faster than me. I can have an idea today and have a somewhat working prototype tomorrow, even on a hobby budget. Quick turnaround PCB manufacturing is expensive.

    Another reason is the sheer challenge of it. It’s incredibly difficult to layout a high-density board that can still be made at home. I requires a lot more skill than 2-layer plated-through layout. Of course, this is a drawback for beginners and needs to be acknowledged. If you’re doing a high-density design, it might be far less hassle to go for professional production.

    Homemade PCBs may also encourage people to finally use reasonably modern components. Since through-hole packages are such a pain to use due to the amount of drilling involved, it might finally get hobbyists to stop fearing SMT. SOICs are a walk in the park to solder and about half the size of DIP parts, 0805 resistors should pose no challenge to anyone (capacitors, on the other hand, require a bit of skill to solder due to cracking). Sure, you probably won’t be using DSBGAs on your homemade PCB, but then again, cheap manufacturers don’t support solder-mask defined pads anyway, so you won’t be using them on cheaply manufactured boards either.

    And then, of course, there’s exotic board materials. You can buy photo-sensitive bendable (0.3mm thick, FR4, not Polyimide) PCBs for a reasonable price. You can buy photo-sensitive metal-core PCBs for a reasonable price. These are somewhat special applications, where no cheap PCB manufacturer will be able to supply you with a professionally made board due to the lack of volume. They’re only cheap because they pool everything, so if you need something out of the ordinary, you’ll have to part with quite a lot of money.

    Of course, homemade PCBs are a pain to produce, so I tend to go for cheap Chinese PCBs most of the time. However, this decision is made BEFORE I begin layouting, so I know which design rules I need to follow, what can easily be done and what’s hard.
    Of course, there are some ways of making PCBs at home that people should consider completely outdated and not recommended. Isolation milling is one of them: You’ll need extremely expensive equipment, it takes ages, doesn’t even produce the layout you specified unless you’re insane and rub out all the isles, mills are expensive and the resolution is severely limited, so it’s hard or next to impossible to use fine-pitch components. Toner transfer is about as cheap as it gets and if you have a laser printer, you just need chemicals, which makes it extremely attractive to hobbyists just starting out. You can also start over again at any point in the process before etching, which can be quite nice. However, it is insanely hard to get consistent results with fine details, so I see less and less of a point in toner transfer.
    Photo-exposure, once you’ve figured it out, gives consistent, high-resolution results. While I prefer 12 mil traces, it is not an insurmountable problem to go down to 6 mil for critical sections. Yield will obviously decrease, so I wouldn’t recommend it too much. Drill restrings need to be really large because you’re usually drilling after etching, wheras board houses drill before etching. Drilling after etching has the big risk of tearing copper from the substrate, so restrings need to be large to take up the strain.

    tl;dr: editorials: yes, but please ensure reasonable quality. home-etching: yes, for speed, materials, special applications. professionally-made boards for everything else

    1. Good point about oversize boards. I have a project on the table right now, that has about a dozen chips and connectors that fill 4 square inches, plus about 30 square inches of capacitive touch sensors. This would cost me $170 from OSH Park, and dirtypcbs can’t do it at all. My solution? Use OSH Park for the high-density section ($20), and add a home-made single-sided board for the touch panel. Also, the touch panel section is the part most likely to need revision, since it’s the main user interface. Best of both worlds.

  21. We have been some years researching about how to make a low-cost robot which could cover most important tasks of PCB making at home. It was difficult and time consuming but the actual prototype: TwinTeeth can paint the circuit with a UV laser, drill vias and holes, dispense solder paste, 3D print covers and knobs and much more. We are still researching about how to improve it and also we are designing more toolheads. Anyway, I think we don’t should ask why we are still making PCB’s. Better questions could be: why we are still making PCB’s by hand? and more far away: why we are still making copper PCB’s?. We are in the silicon age not in the copper age and we should already have in our garages a wafer silicon printer to make our own chips or a “workcenter” near home where we could send our designs to print them in silicon.

  22. Why build electronic items at all when there is so much ready made kit already out there.

    Come on, why is this kind of crap showing up on a site like Hackaday. So ordering from a boardhouse is more practical. We get it. Carry on, use a boardhouse or make your own. It’s your own choice, not this guys!

    Reasons I might make my own board:

    1) http://americanradiohistory.com. Check it out, several decades of electronic project magazines. Do you know any boardhouses that accept PCB patterns in the form of PDFs? Nope.. me neither.

    2) It’s ready today. All these people doing a process of wait for a board.. build.. test.. fix design error.. repeat.. I don’t get you! You are waiting days at best for each iteration! Wait.. hold that thought.. Why are people prototyping on PCBs in the first place? Seriously, invest in a breadboard and through-hole components. Sure.. some things only come surface mount. Make breakout boards. Seriously.. prototype on custom PCB.. that’s not a prototype! Well.. maybe if it’s a microwave frequency RF project. That’s about all I’ll give you on that one. Oh.. and for RF at sub-microwave frequencies.. learn about Manhattan construction. You don’t need a custom PCB. Save that for the finished product.

    3) Because it’s DIY. If we wanted something pre-made we could get it from Best Buy!

  23. There is exactly one reason why I make boards at home: Time. I use OSHPark all the time, but it is 10-14 days before I get my boards back (unless I am paying for super swift service). Sometimes I just have an idea and want to prototype it. It is very satisfying to go from idea to working prototype in an hour…

  24. The electronics hobbyist will continue to manufacture boards at home for a good long while after boards cease being used to assemble commercial products. My own collection of well-worn fine point sharpies attests that the age of homebrew handmade boards is far from over, but admit there are good manufacturers out there that will gladly do a fantastic job and part you from your green “while you wait” and dazzle with all sorts of wonderful colors and thru-holes. Time indeed is coming when those sharpies won’t be good enough, but it may be that the board houses won’t be of any use then either.

    I’m just waiting with $$$ for some genius or a garage full of kids to provide us the “killer product” that eases our homebrew pcb mfg pain, and becomes rich.

  25. USELESS article. I doubt that there’s anybody out there who isn’t aware that you can have PCBs built for you. What might make a difference for some of the people who still roll their own, is information about where you can get them done cheap, and just HOW cheap. Leaving this up to the commenters is lame.

    To others: if you are so picky about the esthetics of your finished board, do you really think you can get a nicer look from a PCB mill than the nice, pretty purple solder mask that OSH Park supplies? And really: at $5/sqare inch for three copies from OSH Park, and even lower prices from dirtypcbs (dot com) for larger boards, it’s really hard to justify making them in the shop. If I need something faster than two weeks, I can dead-bug/manhattan it (and I can wire it just as fast as I could do a decent job of laying out the PCB), and version 2 can be done with a cheap PCB of much higher quality than I could produce.

    I can imagine that sometime in the early 1900s, there were hackers out there blowing glass and building their own vacuum tubes, and even today there are hackers out there playing with silicon wafers to make their own transistors, and others STILL crafting Nixie tubes and triodes in their basements. Kudos by the thousands to those folks, but making your own PCBs today is kind of in the same category – a great challenge for its own sake, but NOT very sensible if your objective is getting a circuit running.

    I have made enough of my own PCBs to know better, and therefore agree with the author, but the article is still pretty useless.

  26. Why would I make my own boards? Well, I ordered boards from Seeed studios and was supposed to get my boards within 10 days. it took over 30 days. And the boards had a mistake that I made, so I will have to fix that and get another batch made again. I will make the next batch at home. I can’t wait for someone in China to not do what they say.

  27. Who…who actually makes PCBs themselves and didn’t already know this? I make my own thanks to the combination of it being much cheaper and much faster. The second one is a biggie; I’m paranoid and like to be damn sure the thing works before I drop the cash and sit around waiting to have a proper board made. The first plays into it because, near as I can tell, my options are either “have Elecrow make them for cheap and wait 3 weeks” or “have one of the US boardhouses make them and pay more for the boards than everything on them, then wait a week anyway”.

    Really now, creating a resist mask and etching aren’t difficult processes. I’ll grant that drilling is a pain by hand, but I have a CNC router. Personally, I don’t bother with the mask and just give it a spray of conformal coating after assembly and testing. Even if I did want the mask, the process is almost identical to what I use for the etch resist (UV-cured resin). The only reason doing it myself sucks is the plated through holes. That is currently the _only_ reason I’ll wait on boards instead of DIYing it. The second someone figures out a cheap way to do that, I’ll probably only ever order a board if I outright can’t produce it for some reason.

    I’ll grant that I’m perhaps a little more invested in making them than most, but the fact is that, if I can’t get them in a few days for what I deem a reasonable price, I’m ALWAYS going to either make them myself. If not, then at least make the first one myself to verify the design. The only exception to this would be a board with a lot of vias.

    Not to say I wouldn’t PREFER to have someone else do it all the time, because it’s tedious and uninteresting, but it’s just rarely worth it for one reason or another.

  28. Couldn’t agree more. I understand making them if that’s a step you legitimately enjoy, but for me I’d far rather have a proper working PCB for a reasonable price, and work on other stuff while I wait for it.

  29. I normally just build my circuits on perf board. Occasionally I will still etch my own PCB, but not very often at all. About the only time I would bother doing that is if I am building circuit out of a (normally older) electronics magazine. The only reason that I would ever use a PCB service – and the only reason that I did the one time that I did – is if I need multiple copies of a board. If I only need one board for my own personal use – which is 99.9% of the time – I cannot see any reason to send off for, spend money on, and wait for a professional board.

  30. My problem is finding good, easy to use, PCB design software with a decent component library. I have tried several. I can hand design a simple PCB (and most of mine are simple) and make a batch in much less time than it takes to create custom components (due to limited library components) and design the PCB using CAD. Does anyone have any suggestions?

    1. If you’re running Windows, I’ve heard praise for DipTrace both from a UI standpoint, and for ease of part creation. They have a pin-limited (but otherwise fully functional) freeware version.


      (Disclaimer: I use Eagle under Linux and have no personal interest in or experience with DipTrace.)

      1. Where is a good “How to use Eagle” tutorial? I just installed it today, because I’m working on a custom Raspberry Pi HAT. The circuit is fairly simple, but I’m just not figuring out how to use the software.

        For instance, placing through hole points for an ATTiny85. I found an Atmel Library, but they still don’t show up on the list when I click “Add”. Also, running traces, especially from the outer row of pins, and when they need to overlap.

        1. Unfortunately I don’t know of a tutorial I can personally endorse. I started using Eagle in 2002 and remember a steep learning curve. Eventually I became very proficient.

          I’ve used SparkFun’s design for manufacturing (DFM) tutorial and it was straightforward and informative. They have several other tutorials on getting started with Eagle, and I imagine those are of similar quality.


      2. I tried diptrace and found that its library is really strangely arranged – all the components were named with opaque IDs that didn’t appear to be for any specific supplier or manufacturer. It might have an extensive library, but it was really difficult to find things in it… I was hoping for more, because I’ve seen lots of positive mentions of it too.

        1. I don’t know how long ago you tried it, but the library currently is pretty good. There are a bunch of ‘generic’ libraries at the beginning, followed by a large number of manufacturer specific ones. If you’re looking for a part by a specific mfg, you need to look in that latter section.

          As with any PCB CAD, you’re going to spend a lot of time designing your own schematic components and footprints, though. Its footprint editor is substantially better than Eagle’s.

          1. I just downloaded it again to have a look. Nope, still has a library full of components named “2-175674-9” and similar. One of them probably is a DE15 pcb-mount VGA socket, but how would you ever know? That seems to be a TE Connectors part number, but that implies that I need to first find Amphenol, TE, Tyco etc’s part numbers for the same thing and search for all of those in the library, rather than just find whatever is already in there and buy one to match. It seems very laborious.

          2. Some of the generic libraries have parts like that, yes. But there’s also libraries called “Amphenol”, “TE”, or “Tyco” – which is where you should look if you have a particular manufacturer in mind.

          3. And how about the other way around? I want to search by function – I need a 2×16 100mil header and I don’t really care who makes it. It seems I have to research it in someone’s catalog first. Even having found the little description string, entering substrings from that into the filter box doesn’t seem to bring them back up.

          4. You can add the generic symbol from the “Con Symbols” library, then set its pattern to one of the premade generic header symbols from the corresponding pattern library.

          5. Since I whined so much yesterday, I just wanted to add that I tried creating my own patterns and components in diptrace last night, and it was almost a pleasant experience.The UI for tweaking traces (and wires in schematics) is also very nice compared to Eagle!

    2. KiCad. It’s Linux-only, but free. I’ve found component libraries for it that cover everything I can think of. I was involved in PCB layout for several years (up to 8 layers) a decade ago, and learned PowerPCB at that time, so I wasn’t a complete newbie, but when I tried learning Eagle and KiCad, I found KiCad much more intuitive. It could be just because it’s more like PowerPCB, but its also nice that KiCad is completely free, with no layer, component count, or board size limits.

      1. I have tried KiCad but the standard library seems limited and I couldn’t find good additional libraries. Any suggestions on where to look? For example, I need a TO251A layout and a 2×32 by 0.1 pin spacing female connector layout and I couldn’t find either for KiCAD.

  31. Living in lovely South Africa with a postal service that manages to delay packages for 6 months, ordering everything is simply not an option.

    Especially if I am prototyping. One mistake and your professionally made, beautiful board is junk.

    I make them at home, it takes me an hour to print, etch and drill if it is complicated.

    Until I get a local service comparable to OSHPark (Never) I will continue to make my own.

    Agree with some posters up top, HaD seems to increasingly want us to buy things rather than make them.

    One of the best feelings I had was when I saw a Tellymate online and could not afford the thing (College student) and made it myself within a week!
    Learned a ton.

    I like the editorials, but please keep in mind that HaD readers are tinkerers at heart.

  32. Wow. It is pretty clear that one or two of you have caught the author in bed with your wives. Agree or disagree, there’s no reason to name call either the author or the crass troll comenters. What I find funny, though, is the indignation. A man on my TV right now is telling me to buy a Ford truck. I don’t want one, but it is ok that he wants me too. I’m secure enought that I wont’ call my TV station to complain about them trying to make me buy a truck. Grow up. HAD is supposed to be a community, but this is a sad set of comments on all sides for a community.

  33. The only reason I sometimes do my own PCB’s is time, sometimes I just feel like working on something that day and I can’t wait a week (minimum) to get back my fabricated pcb’s so I go ahead and etch my own. There is also one other option, sometimes I re purpose old PCBs that we’re designed for something else but have similar architecture. A bodge here and there can get me the result I want.

  34. *watches vid*
    /Subscribe, Podcast, buddy of mine/

    *Step into my office! Why cuz you are f’in fired. 7 minute abs not 6!*

    Citric acid? Okay. Citric acid in Cola? [Al Williams] Get down here! Tell us what are the two acids in Cola. If you don’t say Phosphoric and don’t revise the nonsense the hipster said we will be ಠ_ಠ and VERY sad you let this poser into the blog.

  35. Message taken from the article : stop making stuff and buy more.

    I come to the blog, bypassing every other part of HaD for the projects not self righteous commercials for one
    anti-maker service or another. I am not a ymajorvoice on here and probably never will be, and that’s fine. “why are you still making PCBs” is as like asking “why are you still baking cakes” or carving or casting metal or any number of craft work, simply because you(the author) are jaded and dissatisfied with the skill and art of a process does not mean you should use HaD as your soapbox, that would be like me coming here to discuss the quality of amateur midget porn and why you shouldn’t direct it yourself.

    Tl,dr this artical should be on the home page and not the projects blog.

    p.s to the author you do come across as nothing more than a whiny teenager complaining about having to put any WORK into anything. Any article telling the community to STOP MAKING cause its easier to buy than make deserves to get tar&feathered

  36. To the large cadre of scandalized “alpha-makers” who whine about this article; are you so wrapped up in declaring your perspective that you don’t see what amazing click-bait this has yielded for HaD? If you genuinely don’t like something on the Internet, the surefire way to help it into obscurity is to ignore it. But instead you’ve provided me and many others with a compelling reason to read the article (due to the comment count, not the title). Thanks! Oh and p.s. – great Op-Ed! I don’t agree with it but who cares? Get back to work.

  37. I make my own boards because 8c/square inch and a 12 min. turn around time, and volumes from 1 to 5.
    If my needs were outside that, I’d send out for them, esp where a turn around time of 30 days, and a price of $1.78/square inch are ok.

    1. sometimes the problem is not the value/square, but the price of the shipping or the time to arrive months (plus the delays if the customs clearance process intervenes, then the secret and extrapolated customs fees and respective bureaucracies..)

  38. Some of us seem to be taking this too personally. The whole bit about “I don’t do this and you shouldn’t either” isn’t meant to be taken seriously. I took it as a way to add personality to a relatively dry topic: fabbing vs shopping this out to a board house. I realize the audience, but come on, guys.

    Clearly, there are an abundance of reasons to do your own boards. All the article is saying is that we don’t have to work as if that’s the only option. It’s not a personal attack on your relevance; it’s a light hearted attempt to start a discussion about who you have gone to in the past to get your boards done.

    Nothing is wrong with replying “no, I’ll make my own for obvious reasons”. But some of us are replying like the author told us that are skills are useless and we should all be put down and retired with our soldering irons and copper plates.

  39. I still make my own PCBs because finding a place to do a one-off is practically impossible and I WANT MY BOARD NOW NOT FOUR FREAKING DAYS LATER.

    Also, if you don’t roll your own, you are not a hacker by any sense of the traditional meaning, and should give up writing articles while claiming to be a hacker.

Leave a 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.