Hacker News Reader

Comment by tomstokes | Hacker News Reader

I love seeing DIY PCB manufacturing projects like this. Using the 3D printer as a plotter is a creative take on the traditional DIY process.

For any aspiring EE hobbyists: Manufacturing your own PCBs is almost never worth the effort. You'll spend days or weeks getting the process and tooling right, and you still have to manually wire any vias that connect the front and back sides of the PCB. These DIY PCB manufacturing projects are fun if you're in it for the experience, but very impractical for getting work done. It's quick and easy to order small quantity PCBs online.

OSH Park is a popular option: https://oshpark.com/#services You can get 3 boards in 9-12 days for $5/square inch. If you need the boards sooner, $10/square inch will get you a 4-5 day turnaround time. You won't save any money by buying all of the gear to DIY etch your boards, and you certainly won't save any time.

View on HN

> Manufacturing your own PCBs is almost never worth the effort

I don't know, for me, since I already had an entry-level 3D printer and Dremel, I was able to mod it to engrave the negative of a PCB onto these copper plates in a few days. Now I can get from design to finished PCB in an hour.

I suspect many people getting into EE might have some of these tools already, and being able to iterate quickly with locally-available supplies is a huge advantage, especially for people new to PCB design.

Edit: I would like to add that my method was rather poorly though out and the method in article is much better. Putting a Dremel on a 3D printer is not a good idea, it was just the first thing I came up with and it happened to work ok-ish for my very basic needs.

It's not a terrible idea, and I've wanted to play around with the concept myself, but I've never really had time or an acute need.

The Snapmaker (1) is a 3d printer with changeable heads. For mine I mostly use the 3d printing head or the low-power laser engraving head, but it also came with a milling head and they sell a higher-power laser head that can cut some materials.

I looked around and people have had trouble trying this with the milling head. The bit usually breaks and the X axis isn't stable enough. Maybe v2.0 will be more sable since it has a second Z axis mount.

That being said, their mounting platform is modular. There's nothing to stop you from fabricating a dremel mount and ripping apart an existing milling head to interface it with the dremel.

[1]: https://snapmaker.com/product

Actually, PCB milling[1] is a thing. I suppose a MPCNC can be used for that. But it still has the huge drawback that you're limited to one layer if you're lazy (like me) or two if you fell like spending the time to add manual vias - though I was told there are tools for that, which work essentially like... Err... riveting pliers? (sorry, non native speaker)

[1] https://en.m.wikipedia.org/wiki/Printed_circuit_board_millin... (derp, I forgot the link and can't edit anymore)

Given that multi-layer PCBs are effectively laminations of individual boards, hobbyists have already made their own multilayer boards [1]. I don't see anything particular about milling as a means of copper removal that would stop you layering up boards?

[1] https://www.instructables.com/id/Flexible-Double-Layer-OR-Mu...

Thanks for the link, nice to know it can be done at the hobyyist-level. But it's still a lot of work I (personally) do not enjoy spending time with. If I had the space I'd probably setup my (hypothetical) MPCNC for milling PCBs, so I can do quick single layer prototypes (maybe some simple stacking using header rows), but anything more complex would still go to a fab.

Those prices and lead times don't seem very competitive. At Sunstone I pay $5/sq in for 1-day service on a 2-layer board. If I was willing to wait a week it's $1/sq in.

I think people who are impressed by 3d printers often aren't very experienced with how easy it is to make or get made the things they are printing. I can draw a board in my EDA program, hit a button at 6pm and come back to work with the boards on my desk at 8am the next morning, if I want to pay for that. Same deal for parts: I can order a box full of crap from Digi-Key at 10pm Pacific Time and the Fedex guy will drop it on my desk 10 hours later. Electronics prototyping is extremely well developed, fast, and dirt cheap.

Let me guess, you live in the US?

Do you have some sort of special deal on pricing? What sort of quantities are you ordering? For oshpark, it's $5/sq in for a batch of 3 boards. I just checked the prices at sunstone and they quoted ~$35 per board for a 1 square inch 2-layer board for 1-day service or $14 per board for 1-week.

JLC PCB is less than $10 for 10 5x5 cm PCBs, including shipping. They just take way too long to deliver where I live.

Not to mention many of their orders are stuck in limbo right now.

All my JLCPCB orders are moving and arriving without an issue but that is not the case with their parts supplier (LCSC) or others.

I just looked at my last order, it was a much larger board. I suppose there are different optimal choices for very small boards or larger ones.

You are right, we live in the golden age of electronics prototyping. I'll throw in plugs for the Octopart search engine for finding parts, and PCBEX for inexpensive prototype boards up to ten layers.

that's actually really cool... sunstone is also US based if you are into keeping it in the US... win-win.

though, i just did a quick quote and they are quoting some seriously high prices so i am not sure where you are getting that number from?

> For any aspiring EE hobbyists: Manufacturing your own PCBs is almost never worth the effort.

Worthiness is something that every hobbyist needs to evaluate themself. You cannot do that for them.

For a hobbyist, time, money, efficiency, quality may all be totally irrelevant.

Yep, this project was basically worth it just for the fun of it.

I love to hear stuff like this. I’m glad you can find your own joy through these kinds of technical arts.

Sometimes it really is just passion that drives us.

Oh yeah, I don't know what I'd do if I stopped finding making things like this fun. It's like being on drugs, except productive.

Hey, you can be productive on drugs... >.>

I guess it depends on the drug.

Yeah I had kind of a weird reaction to his comment.

I was inspired by Ben Eater's YouTube series about making a breadboard computer. Rather than copying his schematics, I designed my own. Took a very long time. Went on digikey and Amazon and dropped probably around $1k in parts and tools. Now I'm building the damn thing and it's very time consuming.

I also recently bought a new laptop for $200 which is many orders of magnitude more capable than my breadboard computer eventually will be. Does that fact mean my time, effort, and money spent on the breadboard computer is a waste? Hell no.

> Does that fact mean my time, effort, and money spent on the breadboard computer is a waste? Hell no.

Well you did spent $1k in parts and tools instead of building theses parts and tools yourself, didn't you? His argument wasn't that it's not worth the experience, in fact he even mention how the experience can be worth it if that's what you want to experience. If your goal is just to get a PCB, then paying for a professionally made PCB will worth it (just like in your case spending $1k in parts and tools will worth it).

I think that just proves his point. You're spending time doing the parts of the project that interests you.

Manufacturing PCBs isn't electrical engineering, it's process engineering. I'm sure there are some process-engineering hobbyists, so if people want to spend their time doing that, then cool. But his point is that if it's the electronic engineering part you're interested in, spending a lot of time tweaking and improving your DIY PCB manufacturing when you can order dirt cheap boards from China that arrive in less than two weeks, that are far and away better than what you could do at home, that generally isn't worth it.

I disagree that it won't save time, if just for the fact that 30 minutes is less than 9-12 days. I usually need neither the quality nor the quantity that these fabs will give me, so being able to quickly make a PCB is definitely worth it for me.

That makes a lot of sense to me. One of the reasons I stick mostly with software is that it allows for much quicker iteration. That increases both my ability to explore a space (when working divergently) and polish a product (when I'm in convergence mode). Hardware is, well, hard.

Iteration speed can make a huge difference in outcome. I read that the average consumer electronics company goes through 3-7 physical product iterations. Apple, on the other hand, went through more than 100 generations of prototype for the first iPod.

Obviously what you've done here is just getting started, but it points in a really interesting direction. A 3D printer with swappable heads for plotting and an add-on PCB etching kit should definitely have a market. That would mean recurring revenue for PCB supplies at a good margin, too.

> That makes a lot of sense to me. One of the reasons I stick mostly with software is that it allows for much quicker iteration.

4-5 day turnaround from OSH park is quite fast. Keep in mind that you usually need to order parts from Digi-Key or another supplier when you finish the design anyway. If you have a mistake on a PCB, it's faster to use a tiny stitch wire than to make a whole new PCB.

On a professional level: You can always buy fast turn PCBs if you need them ASAP, but you'll pay for it. Usually cheaper than paying engineers to fiddle with finicky PCB milling or etching machines, though.

> Obviously what you've done here is just getting started, but it points in a really interesting direction. A 3D printer with swappable heads for plotting and an add-on PCB etching kit should definitely have a market.

Chemical etching is more of a novelty these days. A cheap CNC mill produces much better results without the mess and uncertainty of chemical etching: https://hackaday.com/2018/01/04/guide-why-etch-when-you-can-...

On the professional level, several manufacturers make dedicated PCB mills: https://www.lpkfusa.com/products/pcb_prototyping/machines/ These machines have helpful features to index tools to the surface and align the panels as you flip them over for two-layer designs.

Digikey, Mouser, and Arrow, for the most part, give me next business day delivery for cheap (often free). From Arrow, I think the minimum is only $50 for "free" overnight shipping.

I have extra parts at home, exactly because of this. But I can not order extra PCBs with my future mistakes corrected. So it's either make them myself or try a single design, spot my failure, correct, order a new one, wait a week or two until I have some more free time to mess with it.

Or avoid PCBs completely, what is clearly the best option whenever possible (what is becoming less and less common).

To be fair, modern EDA software helps a lot with not making mistakes. If your schematic diagram and tolerance settings are correct, it is pretty hard to get the PCB wrong. Every time I had a problem with one, I had made a mistake in the schematic (and once the fab house mistakenly bridged pads because my tolerances were too tight).

I wholeheartedly recommend KiCAD if you aren't already using it.

Seconding KiCAD. Their software is fairly intuitive- especially if you're used to Linux- and it's super helpful for making sure that your designs are logical and stuff, and you can also use it to figure out how to best breadboard your designs as well.

I disagree and think saying it's intuitive is doing people a disservice. It's not intuitive at all, but it's more intuitive than the alternatives and it won't take you very long to learn how to use it.

If you don't get it immediately, don't worry, that's normal. Just watch some tutorials and you'll be good to go in a few hours.

Okay, you're right. I'd forgotten that I'd A) had a failed start with it and B) only gotten past that once I had to use it for a school project. That being said, once you figure it out, it's really powerful. It just takes a bit to get there, and I'd forgotten the beginning of my learning curve.

Sorry for the confusion, and thank you for reminding me of how hard it was for me personally to start.

No problem, and I agree, it's very powerful and it's fun to use. It doesn't take very long to get a feel for what's going on, and it's very rewarding afterwards, as you know.

"Manufacturing your own PCBs is almost never worth the effort."

It depends on the target device. For analog or simpler digital circuits prototypes where single side boards can be used it's still very convenient, with the only annoyance being the etchant treatment and disposal. In the old days I made things easier by adding one intermediate step in which I redrew the schematic as the ICs and keyed parts were seen bottom up, that is, ICs had the 1 pin up right, etc. Also the schematic wasn't arranged functionally, like an opamp drawing for example, but it rather showed how the pcb was going to look from solder side. From then on, copying it on the actual pcb became a breeze. As for drawing on the pcb, I did all by hand, using both transfer sheets with pads where necessary and thin (1mm or less) normal water resistant markers for connections, buses etc. Faber Castell and Staedtler were my favorites, as I found those specific for pcb drawing were a lot worse quality wise, not to mention like 4x more expensive. This method would of course not scale with circuits complexity, more than one layer and or series production, still I enjoyed a lot being able to design, etch and solder even moderately complex circuits in one evening or a week end in "mom's basement" (actually dad's attic...:). With time I also trained myself to draw my schematics directly that way, so I could get rid of one step between drawing on paper and the final drawing on the pcb. I miss those days...

> You'll spend days or weeks getting the process and tooling right, and you still have to manually wire any vias that connect the front and back sides of the PCB.

Fancy! At school we had a UV box & bath, and made single-sided PCBs. I assume you can use the same process just with boards that have copper both sides. It was great for demystifying the process, and I suspect economical in a school setting.

(Aside: I then studied EE (+CS mix) at university, and there was none if that. Many of my peers without the opportunity to take Electronics at GCSE & A Level would be nonethewiser. Maybe that's fine, lines have to be drawn, but seems to me like an easy extension of existing lab assignments.)

As a hobbyist, if I had the room I'd certainly want the kit. Nothing beats the immediacy for prototyping , or turning around fixes/modifications. The basic stuff isn't prohibitively expensive, we're not talking even hundreds of units before break-even.

I do it at home (PCB mill) for RF circuits on Rogers 5880 and similar materials. At work though (more RF), I just order it. Seems to always be $3k-$5k per run no matter how many boards and what material. You are buying the PCB lines time yourself rather than sharing the cost.

I had twelve boards recently from OSH Park for a whopping $3.37 total, which includes shipping. They were small boards.

PCB Mills create a lot of fiber glass powder. That thing is not healthy to inhale.

You can buy PCB of other materials such as compressed papper (F2?) but they are not easy to get hold off.

My mill has the cutter fully surrounded by a vacuum manifold, and the vacuum itself is bag and HEPA filtered.

The compressed paper PCBs have horrible properties. The RF laminated all have specific properties for electrical performance; dielectric constant, low loss tangent, low surface roughness, homogeneity, controller thermal expansion, thermal conductivity.

Professionally- agreed. However I look at it as a learning exercise, which is a wonderful way of calibrating one's sense of time/money quadrants.

If you don't have time, but have money- send it to overnight PCB service. If you have time, don't have money- do it yourself. If you have both- send it to regular PCB service.

Now the last one is the least desirable quadrant to be in- if you don't have neither time nor money and this is the first time you're doing this- expect to find either (or even both) of those, at least until you have gotten the equipment and gone through the process a couple of times.

I've seen hobbyists use JLC PCB in the past for low quantity runs and the quality always seems astounding with relatively quick turnaround times (several days). https://jlcpcb.com/

Looks like they do SMT mount now too which wasn't the case last time I looked at their offering.

I've used them many times and have always been extremely satisfied. The downside is the three week wait times.

FWIW PCBWay generally gets my weekend orders to me within the week, and costs about the same as JLC ($5 for 5 boards plus $15 shipping). In some cases I've put an order in on Friday night and gotten them on Tuesday.

Where do you live?

Have had several orders with this timeline. I’m in California.

Yeah, shipping in the US is much faster. Here in Greece you have to pay 3 euros minimum on shipping charges and shipping from anywhere in the EU will take at least a week.

Reply on news.ycombinator.com