Bit late but hey, in the years when i wasn't alive (before 2000), people often used "iron" method of doing this. Basically, they had printed layout on paper with laser printer, then they put paper onto the copper, heated everything and the ink would melt, stick onto copper and bake onto it. Afterwards, they just etched it. I tried this method many times when i was younger and it worked well. But later in school, we had to use a drawing method, which i got really used to, and here are some tricks: - About the pen. You can actually buy specialised pens for PCBs, they aren't that expensive (i got mine with 0,2mm tip for like 2€), they work the best. Also i've heard the sharpies works good. - Slow that thing down. While its drawing, it actually leaves a trace of LIQUID behind, and if its too fast, it draws that liquid with the tip, leaving indent in the liquid in the middle (the marks at the last PCB). When slow enough, the liquid will dry before that could happen, and it will be more thick (in Z axis..) thus more etcher-resistant, also it won't leave any marks. - You can actually repair some traces/spots that has been broken whiel drawing by using very fine brush and acetone and/or sharpie again. Just underlight that PCB and you'll nicely see, where there's less ink, and just add some more. BUT, dab it, not draw. By dabbing, a literal drop of ink will disperse onto the surface, and there is a big chance of scratching off existing ink while drawing. So, dab. - After everything is done and you like what is drawn, before etching, put the PCB in oven or heat it with a heatgun a bit. The ink will bake onto the copper and will resist etcher a lot more. - Also i don't know what etching solution you used, but use ferric chloride. It isn't as aggresive to the ink as others and does a bit better job imo.
Much faster for me to print with a laser onto magazine paper, iron that to a board, then "rinse" away the magazine paper. The reason magazine paper works so well is because of how thin it is. Once the laser image has been ironed on (the iron remelts the toner, which then sticks to the copper), you just soak the paper for a bit in warm water, and it just breaks down and disappears. But this IS cool.
Meh, laser prints are trash for masking, can never get them dark enough for smd traces... or any traces really, just cussing and filling the missing bits with a sharpie like a pleb, and if you're at it, just do it with a sharpie and a ruler from the get go. Or, you know, just get some photoresist film and do it like a normal person. Edit: Regarding the etching solution, from the ease of use and availability, i prefer the H2O2/HCl solution - I mean, you can get peroxide and hydrochloric acid drain cleaner in any hardware store in any country. Also it's easier to work with, you can see the etching happening
This was neat to watch and very much reminded me of my electronics class in high school! We didn't use a pen, we applied traces via a sticky-backed material. Plus, our solution had an aquarium heater to keep it warm, heh.
Years ago, I used to draw the PCB graphics by hand with nail polish before any pens, home cnc's etc. were even conceived. Worked like a charm, do the results were only as good as your artistic skills.
I saw a film made by Tektronix in the 1960s, showing how they made their PCBs at the time. They used a photosensitive emulsion rather than plotting directly onto the copper, but they also did something else that may be helpful. Because they were making double-sided boards with plated through-holes, they first drilled all of the holes that needed plating, the holes had to be drilled before etching the copper. They also used plated through holes, which means that they couldn't just etch the copper directly, since that would etch away the plating they just put in the holes. What they did was us a negative resist, that is, it covered the areas they wanted to remove, rather than the areas they wanted to keep copper on. Then they used a plating solution that very quickly plates a thin coating of tin on all exposed copper. Then they used an etchant that removes copper but not tin. I think they were using ferric chloride; I don't know how it would work with the copper chloride Thomas is using here (i.e., the DIY etchant made using hydrchloric acid and hydrogen peroxide). In their case, the big advantage of this negative process was that the plated holes weren't affected by the etchant, but in the case of using resist ink as an etch resist, the advantage would be that the ink doesn't have to stand up to long exposure to the etchant. As an added benefit, after you etch, the copper is already tin plated, which is a much better surface to solder to than is copper, and doesn't corrode as easily.
I should add, the tin plating solution is readily available and relatively cheap. I bought a 125 ml (4 oz.) bottle sufficient for making quite a few PCBs for about $20 on Amazon, but the price per ml goes way down if you buy larger quantities. It was called "Liquid Tin", made by MG Chemicals. The board only needs to be immersed for 3-5 minutes to get a good plating, and can be reused.
Ok.. There /has/ to be a HowTo video going around on just what the trick is! Many have tried. Only some seem to get it to work. Maybe it's shared at one of the youtube creator meetups.. I'm going to stick to the hacksaw or grinder for now..
All those soldering points should have etched hole in the middle of it. That small cavity on the copper helps to hold your drill bit at the center until the drill binds to the board. Hand-drilling these boards is really easy when the starting dents are already there. It also helps to see where ever you should make the holes.
So many brilliant tricks. The spring mechanism to keep the pen contacting the PCB, using the bed as an agitator. Some serious out-of-the-box thinking. Thanks for sharing.
The best agitator for etching is using air. I have an old 12V tire inflator that I use. It puts out such a pulsating air flow it is brilliant. I have a big shop air compressor too. But the little 12V hunk of junk just works better. I found that out when I was moving and I'd already moved my big AC but still wanted to make a board. Air entrancing beyond just agitating introduces oxygen into the etch which helps the process out too. Basically if your board is taking longer than 2 minutes to etch you're doing it wrong. You also have to heat the bath up to get a good etch.
when I was young I used the Edding permanent marker and it worked really well also the staedler works well but you should check always the traces before put the PCB on the solution. I remember how I had to re-do the layout tons of times with my marker and ruler. If complicated, when I ended that project my teacher told me that I could use a laser printer, and an iron to get the same result, but always checking the traces.
I've been using laser diode to burn away the paint from spray painted copper clad boards, then etching as in the video. Super fine resolution and less problems with coverage. I believe there's a thread on German reprap forums about this method.
Mind sharing the type of the Laser Diode and what kind of optics you use with us? And what is the smallest size you can focus your laser dot (minimal isolation distance).
I'm using a BDR-209 diode at ~700mW power (400mA) with a three element collimating lens. I've recently bought a single element lens for supposedly better power transmission, and I might rise the drive current a bit for more power... All of this because burning away the paint still leaves me with a bit of residue that I have to scrub off before etching. Not sure if that's the lack of power or if my paint is not ideal. As for the spot size - I currently have it set as 0.05mm in my software. Can't remember how I arrived at this number honestly, but judging by the required overlap between passes, it's definitely 450nm lasers, just be mindful of power density.
Awesome video. Love seeing people use modern technology such as 3D printers and utilizing them to make homebrew custom PCBs. This is what I call ingenuitive innovational progress! I'd like to take these concepts a step farther and see what we can come up with where graphene is the printable substance...
@@Dosbomber My personal prototyping favorite was rub-on decals they used to sell at Radio Shack (yes, I am dating myself). You can still get these, but you really have to search for them.
I've never had ANY luck with any form of "permanent" marker on PCBs, and that's over 20 years worth of making PCB's using various different methods. The only pen that reliably sticks to the PCB and survives etching was the old Dalo 33 pen, which is no longer manufactured, and drew lines about 3mm thick so would be useless for any modern PCB.
never got a stable results. always was some inconsistency in that method. either overheat or underheat, overpressed or underpressed, not enough time, too much time, not enough paper between iron and pcb, or too much paper. also highly depends on an iron you use. ugh. just too much variables to control.
It's so much better, and in my opinion actually easier to print on a transparency, and use a UV lamp to expose it then, develop and etch. And you can do pretty much the same process to put on a UV solder mask.
Halfway between both, if i am going to make a pcb (which i very rarely do when i can have a two layer made so quick and cheap) i spray paint the pcb and then throw it under the laser and have it cut the negatives for the traces exactly as you have. I use a co2 but i think it would work just fine with a diode laser on a 3d printer. There are some inherent hazards related to that radiation which you'll have to decide how you manage but it has been the fastest and easiest method of diying it i have found. Outstanding video, your content gives me tons of motivation, thank you.
@@chaos.corner I started with the iron on method and i think that worked well (assuming you calibrated your printer) but i found it took me a long time start to finish. It's still pretty tedious to make a two sided board and you still have to be careful drilling regardless of which method you choose so it isn't perfect either.
@@giliant True. I mostly work one-sided which can take a little juggling and a few jumper wires but works pretty well. I have been moving to SMT though and that makes things a little trickier since through-hole stuff tends to give you freebies. My printer hasn't required any calibration fortunately.
@@Saartekk Heatbed is not the best idea. You have a large flat area, the liquid surface is losing a lot of heat through evaporation. The simpler method is, use a zip-lock bag. My favorite is using FeCL3 etching, but it works the same idea using other chemicals too. I put a ball of roughly 2cm of FeCl3 in the zip-lock bag, pour in hot water, FeCl3 quickly dissolve. Then insert the PCB. Seal the ziplock. Put the bag flat down on a plate. The agitation is done by simply tilting the plate +10 degrees then -10 degrees.
I would love to see a variation where you attach a laser pointer to the 3D printer and use it to expose a photo etch PCB. The laser could be turned on and off using the fan control. Use your same program for generating the G-code and search and replace the Z changes with fan control commands.
I tired this idea a while ago and had many of the same issues that you had. The biggest issue was getting enough resist ink laid down, and 2 passes would sometimes just rub off the layer below. The best solution turned out to be to just print with plastic and a 0.2mm nozzle. Glass filled PLA worked well for me because it didn't shrink and peel up. The etching turned out really well, but drilling was a pain. The answer to that was to print a jig that holds the PCB on the bed and raises it up, and make a holder for a Dremel that is used to drill out the holes with the printer and g-code. The results were pretty satisfactory, but ultimately I concluded this was really only good for a one-off board. PCB fabs are just so inexpensive today that fabbing my own just isn't worth the time it takes.
I think much cleaner method is using etch resistant paint and 2.5-5W laser to burn out pain and than etch.. plenty of videos on TH-cam and laser is much more precise than pen can and can be easyli attached to 3dprinter and enhance G code as well.. and it's just much cooler to use laser than pen :)
I use the toner transfer method, tight traces & gaps are still an issue but it's relatively accurate. Essentially print the design onto (thin) glossy photo paper with a toner printer, transfer the toner onto the blank PCB with a hot iron, soak the board & remove the photo paper while leaving the toner adhered to the copper then drop it in the etching solution. Its a convenient & low tech way to get it done pretty quickly, once you navigate past the pitfalls the boards can be quite consistent. Most failures are due to the ironing bit not going well... Holes are still a pain to drill, I tap them all first then go at it with the micro drill bits. A pillar drill of some sort would be useful!
I was thinking the same thing. But it should stick really, really well then with resistive plastics and after the etching you should be able to remove the plastic without taking the copper off on the small traces...
As someone who works with metal, its really tough to find an adhesive that bonds (reliably) with copper, I wanna say its pretty unlikely that it would bond well. But that was also my initial thought when I saw the title. Probably worth trying with different types of plastics.
Baking the ink in a stove for a while should have helped to cure it and make it more durable. Also, the etchant that you have used is cute, but it causes heavy cavitation which tends to remove even some harder resists. Ferric chloride is not pretty, but it does a much more delicate job. Really nice idea with the heated etching bath. Looking forward for the second video in the series ;)
@@svofski there's tons of ferric chloride solutions out there. I personally use it because how hard other acid is to find out here and it works really well
@@-______-______- You can 3d print the alignment pins on the bed (with a larger base) and there you have it. Also you can 3dprint better pcb alignment and again, do not remove it from the bed until you are done with it.
You do a good job on creating some original maker-oriented content that shows us that just tinkering around is often what teaches us the most. The "Play more work less" method :-)
A friend of mine who was making custom audio systems back in the university. He was using a laser printer and glossy paper. Print out the PCB layout with the printer on the paper and then using an iron transfer the layout to the copper sheet. After a few seconds in eching solution he got the PCB. I have seen this method with my own eyes. Litteraly 5 mins and you have a PCB without the holes.
Great video. The copper in the spent etchant will be in the form of copper sulphate. Add some rusty nails (rusty ones because they wont have a zinc coating) and the iron will displace the copper forcing it to drop out of suspension as a fine powder (store this up till you have enough to reclaim). This leaves a solution of iron sulphate which is a lawn moss killer and a lawn feed.
I did this a few years ago with a HP plotter... Tip: Put the PCB upside down on the water surface. It will not sink because of surface tension... Like this, the copper sinks down to the bottom of the container and makes the etching much much faster. Fast enough so that the marker stays on the PCB... The smallest components that I was able to solder on such PCBs were the largest SMT chips.
"old process" is an understatement, I was tought to make PCBs this way using a pen plotter connected to a BBC model B back in the early 90s and it wasn't a new technique then. Keep at it, the technique can achieve good results. I echo the comments of others on recommending ferric chloride as the etchant, the more modern alternatives tend to be harsher on the etch resist.
The Fabercaster might work better with "ferric chloride " ? The method presented by Heliox, is quite nice and in the RepRap way of re-using existing systems/component (and might be easier to pull off).
Yes, I built a pen plotter to make PCBs several years ago, and using a Lumocolor 319 S-9 with ferric chloride worked very well. Those pens are a little finer, at a 0.4mm tip.
If by any chance you own a laser engraver, I would suggest covering the copper plate with simple acrylic spray paint and then use the laser to burn off the layout in the paint. I did it several years back and the quality and detail is really good. After etching, excess paint can be easily removed with acetone.
You can use photo-etch boards, where you design your circuit, print it out onto acetate, & then lay the acetate design onto the board & burn the circuit onto the board then etch it with the acid.
The part made to hold the pen seems like it's a little bit over-engineered. Maybe one could design something with a little springy bit of PLA in one part. I thought of this after watching the Veritasium video on compliant mechanisms
the process my father used, was printing out the circuit on a laser printer (with Toner!) and ironing it onto copper, to transfer the print, throwing it in an etching solution and then removing the toner,. Took him a while to get right, but he was making them regularly to build headphone amps
I would like to try marking out blue and a scribe in a similar isolation routing setup, I've got the blue and the scribe but no cnc /3D printer. I do have some stepper motors so it's not totally out of the question..
Just don't look to the 3D printing community when it comes to hooking stepper motors up. They don't have a clue how to do it right. the drives they use can't even push 2 Watts of power. It's pathetic.
The other problem with this is that the center drill hole for the pads is filled in. That might not seem like a big deal, but when you can etch out the hole (even a portion etch), your drill bit will "Self-center" into the hole, making it much easier and accurate to drill
Just use a paper-printer and print in mirror. then iron it on the pcb, put in water to remove paper layer. Then put in etching solution and wait. very precise, perfect result.
thats the way I did make my PCBs like 20 years ago. I've been using an edding marker for this (and obviously not a 3d printer), but it was working just fine.
Thanks for the hilarious video on how not to make PCBs. I first started making my own PCBs 45 years ago and learned the hard way that cleanliness of the copper surface is extremely important. All traces of oxidation, oils, dirt, and whatnot must be removed then said surface should not be touched with bare fingers or else it'll be contaminated again. You should also do your etching as soon as possible after cleaning. For agitation of the etching solution, try using an aquarium air pump and bubbler. Etch with the PCB on edge so that the etched material drops to the bottom of the container and not sit on the unetched copper. Last advise, try using "heat transfer paper" also known as "toner transfer paper" to print the etching pattern onto the PCB. It is inexpensive (as long as you have a laser printer or access to one) and produces a very clean and accurate pattern with very fine detail. Search eBay with either of these terms and you'll find plenty of suppliers. You basically print a positive copy of your etching pattern onto the paper using a laser printer. For best results, set the printer for a very dense print to put as much toner as possible onto the paper. Lay the paper print side down onto the copper side of the PCB, cover the paper with a thin sheet of cotton or linen cloth to protect the paper from heat. Using an ordinary clothing iron set for "cotton" slowly and firmly "iron" the paper. Check the transfer process from time to time by gently lifting up the paper at a corner to make sure the toner is transferring correctly. After that just etch as usual. As far as knowing where to drill the holes, I make sure that the etch pattern I'm using has holes where leads are suppose to go through the PCB. That way when the copper is etched away the hole position is marked and the thickness of the copper trace centers the tip of the drill bit allowing for accurate hole locations. You can also use the paper to transfer a pattern on the non-copper side for parts placement. If you have a color laser printer then you can color code your markings. Keep up the good work. I always enjoy your videos even if some of them make me cringe a little, Lol.
I ruined two toner cartridges using PNP Blue so thanks. Ironing can't get very fine pitch anyways. Toner tends to smear when heated. It is plastic after all.
Hey Tom. Try that process in reverse. Paint on an ink resist. With just a brush or something. Then instead of a pen in your setup. Use a sharp pointed object and scratch away the dried ink of the PCB. Then etch. See how that goes. I think you could get better coverage with the ink. You could even do multiple coats to ensure the etchent won't eat threw.
wow, at some point i realized tom is speaking sooo quickly i actually had to check if i have the speed at 100% :D but great! i love the new format and that way also much more content fits in the video. super thumbs up!
Interesting attempt, but getting the pattern right is only half the problem. Drilling holes and routing mount points is the next issue, and then making double sided boards which are increasingly required for anything other than widgets. Very hard to beat the Oshpark or pcbway offerings for this especially with their accelerated processing times for prototypes. But, if you were to make such a system yourself, this is my wishlist: 1. Pattern drilling for all vias, mounting pads, and routed mount points. Has to be done BEFORE copper patterning. 2. Plate through holes for double sided processing. 3. Coat with resist and then laser expose for super precision, registering both sides for double sided. 4. Regular etch. So, we need a milling/drilling capability, and laser resist expose. OR - predrill and then 3D print a conductive material directly onto non-clad FR4. As stated, these are all more difficult than the Oshpark options, but I appreciate the attempt.
I always just use plastic as a photoresist, some of the flexible filaments work quite nicely (dunno if it's polypropylene or TP urethane). By the way, fun fact: if you're drawing a pcb layout by hand, you can use a normal sharpie :)
The question I always wanted to ask: why not print PLA on top of the PCB? No extra tools needed to attach. The only problem could be adhesion, which could be solved by soluble agents.
Back in the days I was getting pretty decent results with a negative laser print on plastic film and hot transfer on copper. Laser toner was super resistant to all etching solutions I tried.
Thomas. Thank You for this video. Probably If You have had cleaned and degreased PCB carefully before plotting, result was much better. Good solution for cleaning is Cif - cleaning crema with microcristals. It removes well corrosion from cooper as well as grease. You need to clean untill water covers perfectlu whole surface of the cooper. Then ink should cover traces it perfectly and etching solution proof. From my experience the best etching solution is B327 - Na2S2O8. It is not so agressive as acids, and not so dirty as FeCl3, but could ignite flamable materials.
Using photoresist and a moderate power 405nm laser would give you a much more resistant coating that would stand up to any etchant. I’ve used both dry film and the bulk “paint” resist sold on eBay with a lot of success. Ultimately the best results I’ve had are with an Epson inkjet printing onto transparencies and photoresist. The resolution is amazing and capable of imaging the finest of traces, without the troubles creating filled areas like a laser which will struggle with solid fills. Refillable carts are essential, and make it very inexpensive to run.
I have no experience making pcbs, but I was kind of thinking that pens would not work. What you really need is something like nail polish. Idea for how to apply ink or nail polish or whatever very fast: make a stamp using a 3D printer. I have a little bit of experience with this, and what I will say is that a "hard" material like PLA does not apply ink to paper very well, because the paper and the bottom of the stamp do not fully come into contact unless you can deform one or the other. PLA works ok, but what you probably want is TPU (haven't tried TPU myself yet, but I intend to do that someday...). I'm guessing the same would be even more true for applying something to a pcb blank, since it's harder than paper, which means that you are really counting on the stamp to deform a little bit.
The Laser toner method (printed on glossy paper from magazine), then transfer on copper board with either hot ironing or nail polish remover is way faster, cheaper and easier. Thanks for trying.
I recommend using a laser printer and transfer the toner with heat onto the pcb. This method is much much easier then the 3D-Printing method. Also it is a lot faster. Don't use a regular iron and paper to do this. Instead use a fuser unit from an old laser printer (you can get them very cheap as an replacement unit for some very old printers, just look around on ebay). Make sure to use a decent quality fuser which you usually do not get when searching for desktop printer fuser units. Instead look for a fuser from large office printers. Also use photo/glossy paper instead of regular paper. They come with a thin layer of plastic (glossy side) which helps to transfer the toner from the paper to the pcb. Know you only need to finde the right temperature (around 190 °C) and roller pressure (maybe you have to modify the fuser unit to fit the pcb) to get really good results. The process is limited to your printers resolution. I can get down to 0.2 mm layer spacing and thickness.
I showed this video to my teenage kids as an example of what hacking looks like. You're awesome, man. I _loved_ how you used the print bed as an agitator!!! (I think you could make this process work with more work on it.)
Your process will work fine. I dunno what you're using as etchant but I've used those permanent markers to correct my pcb made with toner transfer and they work well with ferric chloride. Try using that as the etchant and i think your method will work great
Tip: use a magnifier you wear on your head like clock makers do for drilling holes in PCBs manually and good light. I use a swan necked led light to get the light at an angle that highlights the copper. I get really nicely centered results, ie ICs fit easily. I use a cheap drill press (voltcraft from Aldi) instead of a dremel which has more run out.
A tip for sticking things to mill/engrave/draw on: don’t use double sided tape, it’s a mess to remove. Use regular painter’s tape, the yellow or the blue and stick it both to the table and to the stock then use super glue on the tape and press them together, tape to tape. It will hold the stock as good as double sided tape, but it’s super easy to remove.
It’s funny to see someone make a PCB backwards in so many ways. 1) Always drill the holes first. Thin traces will tear off if you drill the holes after etching. 2) The orientation matters. If you print it in the wrong orientation, you will need to mount the ICs upside down, bending the pins up. 3) Ensure that the temperature of the PCB does not vary by more than 2 degrees F, between the drilling of the board, and the masking of the board for etching. If the temperature varies by more than that, the masking will not match the drilled holes, as the epoxy board will expand or contract beyond allowable tolerances. Also, photo-resist etching is illegal in most countries (US & Canada especially), due the level of toxicity of the chemicals, and the amount of chemicals required per board. For 1 offs, something like this or a CNC is ok, but not optimal. It is fairly inexpensive to take your files to a printer to create 4x positive films, and then take the films to a fabricator, who can also plate and solder mask the boards.
I suggest when plotting to heat up the bed to aid in the ink drying. Also, I've had pretty good success with sharpies doing touch up on the laser printer/iron process. Not sure how hard it is to get a sharpie where you are.
Those black pens are best for when you use UV light, it is also the better way to make your PCB's anyway and you already have the tools to make a mini UV box (I suggest an old flatbed scanner).
You should explore if you could use the same process (hot end mounted pen) for something like calligraphy (think writing out wedding invitations or signing your name on a thank you card). It would be great to see the process you find from scanning a signature to creating an STL file to creating the GCODE etc. Cool video!
I bought a $60 Laser Printer and a $5 magazine to print my PCB diagrams onto and a $6 bottle of nail polish remover to transfer the toner (perfectly) to my board. $10 gallon of Muriatic acid (which should last me nearly forever 😂) from the hardware store and boom, perfect PCBs. The acid eats the copper so quickly that the toner (which is melted plastic so it doesn't corrode anyways) has zero problems with trace definition. Even with traces as small as .254 mm wide.
I know the video is all about using the the 3D printer but the method where you print the PCB on a laser printer then transfer the paper print to the board with an iron, then etch is much faster, more accurate and a lot easier.
I have been using a cold toner transfer method for making boards for several years and have very good results. Then use mt cnc engraver to drill the boards
I ought to add a conductive pen or five to my next pile of materials and parts. Not for anything meant to last, but just to see how practical a fully-drawn PCB would be.
Post etch, it might pull the traces up - the glue that holds the copper on the phenolic board isn't horribly strong. Sanding the PLA away, however, might work.
@@jeffpotts6187 PCB are FR4/G10 not phenolic. Modern Copper Clad is usually very tight. Question is only how well the PLA sticks, and if it resists the acid.
I use "phenolic" as a catch-all term when I don't want to look up specifics about what the copper is hosted on. Secondly, I think you might be OK with Ferric Chloride, but with the other etchants people use I think it would eat away at the plastic. Ferric Chloride is considered caustic, but the others stuff generally IS caustic. With those, I think PLA would break down during etching. ABS and/or PETG, maybe not - but I wouldn't bet on it. I'm not a chemist. That being said, it's less hassle to use a plain old Sharpie.
I did have good results with the method. I used the black Steadler pen with a different pen holder (it also had play though) and I used FeCl3 as etchant. I think the HCl/H2O2 or peroxodisulfate just attacked the ink too much - there is no need to use the hardcore oxidants. For my projects I didn't continue using the method though because the pen tip is pretty coarse (couldn't really find pens with thinner tips) which necessitates large clearences between traces and you can only add the largest smd components without shorting pins. Also, chinese PCB factories ship insanely cheap. But if you want a DIY solution lithography is probably the way to go. You can even add your own solder mask by that method!
This is a really good start. First off, I think you need 3 linear bearings on the toolhead to fix the binding issue. This could easily be turned into something functional with a few changes. But that being said, milling is probably a better option, just due to the lack of toxic chemicals.
what toxic chemicals? fecl3 with the residual copper ions is able to be neutralized with sodium bicarbonate and can be safely disposed of... I fail to understand how people don't get that...
Hi Tom. It is easy to make a mount for a small drill that you could use to drill the pcb. Sure a printer tends to lack the rigidity to mill but you can drill only easily
The pads have no holes in, good luck drilling! Easy to do by hand with the right tool as they centre the drill bit, I used to do it for pocket money. Measure actual ink width and adjust g-code generation to suit.
Toner transfer works well with ferric chloride. It also seems that an Edding works good with ferric chloride, but I didn't really try that. And you can export the PCB in a way so you have circles instead of solid dots as Pads, so you know where to drill.
You can see the pen running out of ink because half is sort-of ok, and the other half is poor. Also solvent based ink tends to dissolve it's own lines if it touches deposited ink. Maybe a new pen and a smaller PCB and it might work. I put a 0.4mm holed pad on each THT pad as a guide to locate the drill bit accurately, etch and tin the PCB, and the step left in the centre of the pad is enough to guide the drill when I drill them.
Thomas, Could an FDM 3D printer with a small enough nozzle directly deposit a plastic etching mask layer to a copper-clad board? Would any common filament materials like PLA, ABS or PETG even survive submersion in FeCl during etching? You'd have to be careful removing the filament layer when you're done to avoid lifting copper pads, but luckily these plastics have a much lower melting temperature than copper or especially fiberglass, likely meaning you could burn it off without too much work. Maybe there is a second chemical bath that could dissolve only the plastic, leaving the copper and FR4. I'd really love to see this tried out in the future!
Hmm have you tried using a wavelength laser that will only burn where the ink is drawn? You could use the ink to prevent scorching on edges with a laser that normally happens with laser etching.
You could always 3d print a drilling template with all the holes in the correct position, you'd need to trim the board accurately, or add in large hole markers so the two could be bolted together. That said, you are right, it doesn't seem like the best way to make them. I am pretty sure the original RepRap boards I've seen were mostly acid etched though.
Hey, here's a crazy idea: Why not print the tracks in PLA or another plastic that the 3D printer can handle directly on the PCB? I mean, that's somewhat the whole idea with PCB Toner transfer... toner is a mixture of plastic and pigments that gets melted on paper (or a PCB in our case) providing a pretty nice etch resistant mask. I think it would be quite fiddly to get this working, and you'd need a very flat PCB surface to begin with... anyway... I think this would be an awesome idea for a video!
I tried once to heat a recipient with etching solution (ferric chloride), which worked well. The problem is that later I found some odd corrosion on the tool head. Maybe I should get the tool head higher. But maybe if I going to repeat this, I will build a heat bead with a mosfet and an arduino. Which will have the advantage that if I knock it over, it won't be over the 3D printer.
If you print a paper template and use a drill press that can handle very small bits, you can do a reasonable job drilling your own holes in custom-etched PCBs. Double sided with some alignment work too, but since you can't really plate vias that easily on your own and the alignment and setup steps are a lot longer, it's not really worth the effort, IMO. I wonder if there's a liquid extruder head (maybe a syringe controlled by a stepper) that would be suitable to apply solder resist. The data for the layer is part of the gerbers that are generated, and they do sell liquid solder resist for touchup work, so if you connected it up to your printer head and could control the flow out of the syringe or other mechanism, you could probably do a decent job applying it.
Thomas, don't give on 2D drawing PCB designs onto copper just yet. I bought a pen in Japan called a Uni Prockey Twin, made by Mitsubishi Pencil Co., part number PM-120T, that has a tip on each end, one fine and one two or three times thicker. Don't know the diameters but I do know that it worked fine as a repair pen for a PCB I made recently using the nail-polish-remover-toner-transfer method. I suggest you outline the drawing using the fine tip first, then fill in the voids with the fat tip. This could preserve resolution without making wave patterns in the voids. I don't know if there's enough ink in the pen for a decent sized PCB but it won't hurt to find out.
I personally use Diptrace for layout and GBR export (it's kinda similar to Eagle), CopperCAM for the g-code, and bCNC to send the G-Code to the Arduino controller. For me personally, bCNC is the only tool to use because it has a built in probe height mapping function. I put in the tool I'm using to route out the traces, and using the Probe pin (A5) connected to the PCB via an alligator clip with another clip to ground connected to the tool, then let bCNC map out the surface. It's a slick process and makes perfect depth cuts. For the holes, I quit using a drill bit and switched to tiny endmills, they make cleaner holes that drill bits in PCB material. I can knock out a decent size PCB in less than 25 mins on the router, small PCBs take only a few minutes, plus I get to tinker with machines =)
Spray the board with lacquer first (preferably white). Draw on the lacquer and dissolve the undrawn parts away with something like isopropyl alcohol. Etch with ferric chloride. Also, your drawing technique from what I can see in the video is using a dabbing motion with the pen to fill in copper areas whereas the pen should stay down for each entire track (or as much as possible). The raster file seems to make no difference between track drawing and area filling. Again this may just be my view of the video file being wrong. Does it draw a continuous outline for each copper section first without lifting the pen?
Hi! I can see you have problems with the etching sollution when using pens. Only pen that I find that worked was the Edding brand. Staedtler pens were too weak against HCl and H2O2 solution. You could use Ferric Chloride or Sodium persulfate, with latter I did achieve good results, but the "resist" paint has to be good.
Bit late but hey,
in the years when i wasn't alive (before 2000), people often used "iron" method of doing this. Basically, they had printed layout on paper with laser printer, then they put paper onto the copper, heated everything and the ink would melt, stick onto copper and bake onto it. Afterwards, they just etched it.
I tried this method many times when i was younger and it worked well. But later in school, we had to use a drawing method, which i got really used to, and here are some tricks:
- About the pen. You can actually buy specialised pens for PCBs, they aren't that expensive (i got mine with 0,2mm tip for like 2€), they work the best. Also i've heard the sharpies works good.
- Slow that thing down. While its drawing, it actually leaves a trace of LIQUID behind, and if its too fast, it draws that liquid with the tip, leaving indent in the liquid in the middle (the marks at the last PCB). When slow enough, the liquid will dry before that could happen, and it will be more thick (in Z axis..) thus more etcher-resistant, also it won't leave any marks.
- You can actually repair some traces/spots that has been broken whiel drawing by using very fine brush and acetone and/or sharpie again. Just underlight that PCB and you'll nicely see, where there's less ink, and just add some more. BUT, dab it, not draw. By dabbing, a literal drop of ink will disperse onto the surface, and there is a big chance of scratching off existing ink while drawing. So, dab.
- After everything is done and you like what is drawn, before etching, put the PCB in oven or heat it with a heatgun a bit. The ink will bake onto the copper and will resist etcher a lot more.
- Also i don't know what etching solution you used, but use ferric chloride. It isn't as aggresive to the ink as others and does a bit better job imo.
we still use iron btw
@@dhupee Yep, Iron with a laser toner printer is so faster, more precise and cheaper.
@@alexandrevaliquette1941 if you have laminator it's better
Much faster for me to print with a laser onto magazine paper, iron that to a board, then "rinse" away the magazine paper. The reason magazine paper works so well is because of how thin it is. Once the laser image has been ironed on (the iron remelts the toner, which then sticks to the copper), you just soak the paper for a bit in warm water, and it just breaks down and disappears.
But this IS cool.
Meh, laser prints are trash for masking, can never get them dark enough for smd traces... or any traces really, just cussing and filling the missing bits with a sharpie like a pleb, and if you're at it, just do it with a sharpie and a ruler from the get go. Or, you know, just get some photoresist film and do it like a normal person.
Edit: Regarding the etching solution, from the ease of use and availability, i prefer the H2O2/HCl solution - I mean, you can get peroxide and hydrochloric acid drain cleaner in any hardware store in any country. Also it's easier to work with, you can see the etching happening
This was neat to watch and very much reminded me of my electronics class in high school! We didn't use a pen, we applied traces via a sticky-backed material. Plus, our solution had an aquarium heater to keep it warm, heh.
I've heard of the ancient technology of sticky-backed traces, but have never had the honor to witness it myself.
@@MadeWithLayers Can you name some sites to download the example pcb design files. I'm a Mech Engineer and need it in dxf or dwg format
Years ago, I used to draw the PCB graphics by hand with nail polish before any pens, home cnc's etc. were even conceived. Worked like a charm, do the results were only as good as your artistic skills.
I saw a film made by Tektronix in the 1960s, showing how they made their PCBs at the time. They used a photosensitive emulsion rather than plotting directly onto the copper, but they also did something else that may be helpful. Because they were making double-sided boards with plated through-holes, they first drilled all of the holes that needed plating, the holes had to be drilled before etching the copper. They also used plated through holes, which means that they couldn't just etch the copper directly, since that would etch away the plating they just put in the holes.
What they did was us a negative resist, that is, it covered the areas they wanted to remove, rather than the areas they wanted to keep copper on. Then they used a plating solution that very quickly plates a thin coating of tin on all exposed copper. Then they used an etchant that removes copper but not tin. I think they were using ferric chloride; I don't know how it would work with the copper chloride Thomas is using here (i.e., the DIY etchant made using hydrchloric acid and hydrogen peroxide).
In their case, the big advantage of this negative process was that the plated holes weren't affected by the etchant, but in the case of using resist ink as an etch resist, the advantage would be that the ink doesn't have to stand up to long exposure to the etchant.
As an added benefit, after you etch, the copper is already tin plated, which is a much better surface to solder to than is copper, and doesn't corrode as easily.
I should add, the tin plating solution is readily available and relatively cheap. I bought a 125 ml (4 oz.) bottle sufficient for making quite a few PCBs for about $20 on Amazon, but the price per ml goes way down if you buy larger quantities. It was called "Liquid Tin", made by MG Chemicals. The board only needs to be immersed for 3-5 minutes to get a good plating, and can be reused.
3:11 ahhh... the good old TOT method. I never got it to work
Yeah, that's an excellent effect. ToT is ahead of his time.
Yep, you have to watch the "Clickspring Time Travel" vid!!
@@OzFaxFlyer I did back then. Also the wintergatan one and french guy cooking... Great minds and stuff ..
Ok.. There /has/ to be a HowTo video going around on just what the trick is! Many have tried. Only some seem to get it to work.
Maybe it's shared at one of the youtube creator meetups..
I'm going to stick to the hacksaw or grinder for now..
@@ElectraFlarefire I got it working on spaghetti. but nothing harder than that. at least not in 3 parts. I wish there was an open source wiki for that
All those soldering points should have etched hole in the middle of it. That small cavity on the copper helps to hold your drill bit at the center until the drill binds to the board. Hand-drilling these boards is really easy when the starting dents are already there. It also helps to see where ever you should make the holes.
I drilled hundreds of boards with centre marks in a plain ol' drill press in my younger years. We drill a stack of 3 at a time.
3:34 for a split second i thought Tom was pouring superglue on the bearings hahah
Oh, it wasn't only me. :-)
Thought the same
It can help to reduce the strain on the moving parts...
Me to
superglue IS the best lube just add a pinch of unicorn fart and mix vigorously! I read it on the internet, it simply has to be true!
So many brilliant tricks. The spring mechanism to keep the pen contacting the PCB, using the bed as an agitator. Some serious out-of-the-box thinking. Thanks for sharing.
The best agitator for etching is using air. I have an old 12V tire inflator that I use. It puts out such a pulsating air flow it is brilliant. I have a big shop air compressor too. But the little 12V hunk of junk just works better. I found that out when I was moving and I'd already moved my big AC but still wanted to make a board. Air entrancing beyond just agitating introduces oxygen into the etch which helps the process out too. Basically if your board is taking longer than 2 minutes to etch you're doing it wrong. You also have to heat the bath up to get a good etch.
Someone's been watching This Old Tony.
yepp, i tought the same
when I was young I used the Edding permanent marker and it worked really well also the staedler works well but you should check always the traces before put the PCB on the solution. I remember how I had to re-do the layout tons of times with my marker and ruler.
If complicated, when I ended that project my teacher told me that I could use a laser printer, and an iron to get the same result, but always checking the traces.
I've been using laser diode to burn away the paint from spray painted copper clad boards, then etching as in the video. Super fine resolution and less problems with coverage. I believe there's a thread on German reprap forums about this method.
I use this method, too. Beautiful boards!
Mind sharing the type of the Laser Diode and what kind of optics you use with us? And what is the smallest size you can focus your laser dot (minimal isolation distance).
I'm using a BDR-209 diode at ~700mW power (400mA) with a three element collimating lens. I've recently bought a single element lens for supposedly better power transmission, and I might rise the drive current a bit for more power... All of this because burning away the paint still leaves me with a bit of residue that I have to scrub off before etching. Not sure if that's the lack of power or if my paint is not ideal. As for the spot size - I currently have it set as 0.05mm in my software. Can't remember how I arrived at this number honestly, but judging by the required overlap between passes, it's definitely 450nm lasers, just be mindful of power density.
Awesome video. Love seeing people use modern technology such as 3D printers and utilizing them to make homebrew custom PCBs. This is what I call ingenuitive innovational progress! I'd like to take these concepts a step farther and see what we can come up with where graphene is the printable substance...
Standard Sharpie permanent markers are typically used for masking PCBs. They're inexpensive, have lots of tip types, and the ink holds during etching.
I've never had any luck with Sharpie markers as etch resist. :(
@@Dosbomber My personal prototyping favorite was rub-on decals they used to sell at Radio Shack (yes, I am dating myself). You can still get these, but you really have to search for them.
I've never had ANY luck with any form of "permanent" marker on PCBs, and that's over 20 years worth of making PCB's using various different methods. The only pen that reliably sticks to the PCB and survives etching was the old Dalo 33 pen, which is no longer manufactured, and drew lines about 3mm thick so would be useless for any modern PCB.
Much better to laser jet print pcb artwork onto cheap photo paper then iron on to clad board never fails good results with good high resolution
My dad used this method, worked quite well. He used a modified laminator instead of an iron.
never got a stable results. always was some inconsistency in that method. either overheat or underheat, overpressed or underpressed, not enough time, too much time, not enough paper between iron and pcb, or too much paper. also highly depends on an iron you use. ugh.
just too much variables to control.
It's so much better, and in my opinion actually easier to print on a transparency, and use a UV lamp to expose it then, develop and etch. And you can do pretty much the same process to put on a UV solder mask.
Did it this way in school about 15 years ago. Worked a charm.
I like the direction you're going explaining the origins of the RepRap project and its focus on replicatable machines.
Rep Rap is why academics should not try to do anything practical in nature.
This video is not sponsored by jlcbcp
Should it be?
Toner transfer is still my go to when i need a modestly complex prototype in a few hours
This is the kind of video jlcbcp needed.
Halfway between both, if i am going to make a pcb (which i very rarely do when i can have a two layer made so quick and cheap) i spray paint the pcb and then throw it under the laser and have it cut the negatives for the traces exactly as you have.
I use a co2 but i think it would work just fine with a diode laser on a 3d printer. There are some inherent hazards related to that radiation which you'll have to decide how you manage but it has been the fastest and easiest method of diying it i have found.
Outstanding video, your content gives me tons of motivation, thank you.
I mostly use the iron-on method but I've been wanting to give laser a try.
@@chaos.corner I started with the iron on method and i think that worked well (assuming you calibrated your printer) but i found it took me a long time start to finish.
It's still pretty tedious to make a two sided board and you still have to be careful drilling regardless of which method you choose so it isn't perfect either.
@@giliant True. I mostly work one-sided which can take a little juggling and a few jumper wires but works pretty well. I have been moving to SMT though and that makes things a little trickier since through-hole stuff tends to give you freebies. My printer hasn't required any calibration fortunately.
I love the 3dp as pcb agitator :)
Jup! Very clever! Why not?? It has a heatbed included for the acid solution... very very clever!!! 👍👍👍
@@Saartekk
Heatbed is not the best idea.
You have a large flat area, the liquid surface is losing a lot of heat through evaporation.
The simpler method is, use a zip-lock bag. My favorite is using FeCL3 etching, but it works the same idea using other chemicals too. I put a ball of roughly 2cm of FeCl3 in the zip-lock bag, pour in hot water, FeCl3 quickly dissolve. Then insert the PCB. Seal the ziplock. Put the bag flat down on a plate. The agitation is done by simply tilting the plate +10 degrees then -10 degrees.
I would love to see a variation where you attach a laser pointer to the 3D printer and use it to expose a photo etch PCB. The laser could be turned on and off using the fan control. Use your same program for generating the G-code and search and replace the Z changes with fan control commands.
I tired this idea a while ago and had many of the same issues that you had. The biggest issue was getting enough resist ink laid down, and 2 passes would sometimes just rub off the layer below. The best solution turned out to be to just print with plastic and a 0.2mm nozzle. Glass filled PLA worked well for me because it didn't shrink and peel up. The etching turned out really well, but drilling was a pain. The answer to that was to print a jig that holds the PCB on the bed and raises it up, and make a holder for a Dremel that is used to drill out the holes with the printer and g-code. The results were pretty satisfactory, but ultimately I concluded this was really only good for a one-off board. PCB fabs are just so inexpensive today that fabbing my own just isn't worth the time it takes.
I think much cleaner method is using etch resistant paint and 2.5-5W laser to burn out pain and than etch.. plenty of videos on TH-cam and laser is much more precise than pen can and can be easyli attached to 3dprinter and enhance G code as well.. and it's just much cooler to use laser than pen :)
Yeah, but then you're using a friggin laser.
Thomas Sanladerer and the problem with that is .... what? That’s what they make laser goggles for!
Careful, copper is really good at reflecting the beam back.
A sharpie would do the job
I use the toner transfer method, tight traces & gaps are still an issue but it's relatively accurate. Essentially print the design onto (thin) glossy photo paper with a toner printer, transfer the toner onto the blank PCB with a hot iron, soak the board & remove the photo paper while leaving the toner adhered to the copper then drop it in the etching solution. Its a convenient & low tech way to get it done pretty quickly, once you navigate past the pitfalls the boards can be quite consistent. Most failures are due to the ironing bit not going well... Holes are still a pain to drill, I tap them all first then go at it with the micro drill bits. A pillar drill of some sort would be useful!
Wondering about printing plastic onto the copper then acid etching
That's quite a sound idea. Especially with a smaller nozzle.
I was thinking the same thing. But it should stick really, really well then with resistive plastics and after the etching you should be able to remove the plastic without taking the copper off on the small traces...
I was thinking the same thing...
As someone who works with metal, its really tough to find an adhesive that bonds (reliably) with copper, I wanna say its pretty unlikely that it would bond well. But that was also my initial thought when I saw the title. Probably worth trying with different types of plastics.
So far PLA and PETG don't wanna stick. I'ma try other plastics
Baking the ink in a stove for a while should have helped to cure it and make it more durable. Also, the etchant that you have used is cute, but it causes heavy cavitation which tends to remove even some harder resists. Ferric chloride is not pretty, but it does a much more delicate job. Really nice idea with the heated etching bath. Looking forward for the second video in the series ;)
And ferric chloride is simple/easy to make at home, so bonus points there.
@@LightGameFrameworks I never tried preparing ferric chloride at home. Would you like to share a recipe?
@@svofski there's tons of ferric chloride solutions out there. I personally use it because how hard other acid is to find out here and it works really well
Thanks to help us avoidning the errors. Way to go: Laser etching over spray painted board.
You can drill some registration holes to align the top and bottom layers for a proper dual layer construction.
@@-______-______- You can 3d print the alignment pins on the bed (with a larger base) and there you have it. Also you can 3dprint better pcb alignment and again, do not remove it from the bed until you are done with it.
You do a good job on creating some original maker-oriented content that shows us that just tinkering around is often what teaches us the most. The "Play more work less" method :-)
A friend of mine who was making custom audio systems back in the university. He was using a laser printer and glossy paper. Print out the PCB layout with the printer on the paper and then using an iron transfer the layout to the copper sheet. After a few seconds in eching solution he got the PCB. I have seen this method with my own eyes. Litteraly 5 mins and you have a PCB without the holes.
Great video.
The copper in the spent etchant will be in the form of copper sulphate. Add some rusty nails (rusty ones because they wont have a zinc coating) and the iron will displace the copper forcing it to drop out of suspension as a fine powder (store this up till you have enough to reclaim). This leaves a solution of iron sulphate which is a lawn moss killer and a lawn feed.
Just printing a transparency and exposing a photo resist is still quite easy, and gives great results.
Sharpie permanent markers work for the etch resist. Have never heard of any other brand working for making printed circuit boards. Good video, Tom!
I second this
I did this a few years ago with a HP plotter... Tip: Put the PCB upside down on the water surface. It will not sink because of surface tension... Like this, the copper sinks down to the bottom of the container and makes the etching much much faster. Fast enough so that the marker stays on the PCB... The smallest components that I was able to solder on such PCBs were the largest SMT chips.
"old process" is an understatement, I was tought to make PCBs this way using a pen plotter connected to a BBC model B back in the early 90s and it wasn't a new technique then.
Keep at it, the technique can achieve good results. I echo the comments of others on recommending ferric chloride as the etchant, the more modern alternatives tend to be harsher on the etch resist.
The Fabercaster might work better with "ferric chloride
" ? The method presented by Heliox, is quite nice and in the RepRap way of re-using existing systems/component (and might be easier to pull off).
I also think it would be better to do it with ferric chloride :)
I use ferric chloride and when I see people use this other method on TH-cam, it just looks like a lot of faffing.
Yes, I built a pen plotter to make PCBs several years ago, and using a Lumocolor 319 S-9 with ferric chloride worked very well. Those pens are a little finer, at a 0.4mm tip.
If by any chance you own a laser engraver, I would suggest covering the copper plate with simple acrylic spray paint and then use the laser to burn off the layout in the paint. I did it several years back and the quality and detail is really good. After etching, excess paint can be easily removed with acetone.
You can use photo-etch boards, where you design your circuit, print it out onto acetate, & then lay the acetate design onto the board & burn the circuit onto the board then etch it with the acid.
The part made to hold the pen seems like it's a little bit over-engineered. Maybe one could design something with a little springy bit of PLA in one part. I thought of this after watching the Veritasium video on compliant mechanisms
the process my father used, was printing out the circuit on a laser printer (with Toner!) and ironing it onto copper, to transfer the print, throwing it in an etching solution and then removing the toner,.
Took him a while to get right, but he was making them regularly to build headphone amps
I would like to try marking out blue and a scribe in a similar isolation routing setup, I've got the blue and the scribe but no cnc /3D printer. I do have some stepper motors so it's not totally out of the question..
Just don't look to the 3D printing community when it comes to hooking stepper motors up. They don't have a clue how to do it right. the drives they use can't even push 2 Watts of power. It's pathetic.
I like the direction youre going with your chanel not only doing 3D printing stuff
The other problem with this is that the center drill hole for the pads is filled in. That might not seem like a big deal, but when you can etch out the hole (even a portion etch), your drill bit will "Self-center" into the hole, making it much easier and accurate to drill
Just use a paper-printer and print in mirror. then iron it on the pcb, put in water to remove paper layer.
Then put in etching solution and wait. very precise, perfect result.
thats the way I did make my PCBs like 20 years ago. I've been using an edding marker for this (and obviously not a 3d printer), but it was working just fine.
Thanks for the hilarious video on how not to make PCBs. I first started making my own PCBs 45 years ago and learned the hard way that cleanliness of the copper surface is extremely important. All traces of oxidation, oils, dirt, and whatnot must be removed then said surface should not be touched with bare fingers or else it'll be contaminated again. You should also do your etching as soon as possible after cleaning.
For agitation of the etching solution, try using an aquarium air pump and bubbler. Etch with the PCB on edge so that the etched material drops to the bottom of the container and not sit on the unetched copper.
Last advise, try using "heat transfer paper" also known as "toner transfer paper" to print the etching pattern onto the PCB. It is inexpensive (as long as you have a laser printer or access to one) and produces a very clean and accurate pattern with very fine detail. Search eBay with either of these terms and you'll find plenty of suppliers. You basically print a positive copy of your etching pattern onto the paper using a laser printer. For best results, set the printer for a very dense print to put as much toner as possible onto the paper. Lay the paper print side down onto the copper side of the PCB, cover the paper with a thin sheet of cotton or linen cloth to protect the paper from heat. Using an ordinary clothing iron set for "cotton" slowly and firmly "iron" the paper. Check the transfer process from time to time by gently lifting up the paper at a corner to make sure the toner is transferring correctly. After that just etch as usual.
As far as knowing where to drill the holes, I make sure that the etch pattern I'm using has holes where leads are suppose to go through the PCB. That way when the copper is etched away the hole position is marked and the thickness of the copper trace centers the tip of the drill bit allowing for accurate hole locations. You can also use the paper to transfer a pattern on the non-copper side for parts placement. If you have a color laser printer then you can color code your markings.
Keep up the good work. I always enjoy your videos even if some of them make me cringe a little, Lol.
I ruined two toner cartridges using PNP Blue so thanks. Ironing can't get very fine pitch anyways. Toner tends to smear when heated. It is plastic after all.
Hey Tom. Try that process in reverse. Paint on an ink resist. With just a brush or something. Then instead of a pen in your setup. Use a sharp pointed object and scratch away the dried ink of the PCB. Then etch. See how that goes. I think you could get better coverage with the ink. You could even do multiple coats to ensure the etchent won't eat threw.
wow, at some point i realized tom is speaking sooo quickly i actually had to check if i have the speed at 100% :D but great! i love the new format and that way also much more content fits in the video. super thumbs up!
Interesting attempt, but getting the pattern right is only half the problem. Drilling holes and routing mount points is the next issue, and then making double sided boards which are increasingly required for anything other than widgets. Very hard to beat the Oshpark or pcbway offerings for this especially with their accelerated processing times for prototypes.
But, if you were to make such a system yourself, this is my wishlist:
1. Pattern drilling for all vias, mounting pads, and routed mount points. Has to be done BEFORE copper patterning.
2. Plate through holes for double sided processing.
3. Coat with resist and then laser expose for super precision, registering both sides for double sided.
4. Regular etch.
So, we need a milling/drilling capability, and laser resist expose.
OR - predrill and then 3D print a conductive material directly onto non-clad FR4.
As stated, these are all more difficult than the Oshpark options, but I appreciate the attempt.
Vias are the bane of all homebew PCBs I think.
I always just use plastic as a photoresist, some of the flexible filaments work quite nicely (dunno if it's polypropylene or TP urethane). By the way, fun fact: if you're drawing a pcb layout by hand, you can use a normal sharpie :)
The question I always wanted to ask: why not print PLA on top of the PCB? No extra tools needed to attach. The only problem could be adhesion, which could be solved by soluble agents.
Back in the days I was getting pretty decent results with a negative laser print on plastic film and hot transfer on copper. Laser toner was super resistant to all etching solutions I tried.
Thomas.
Thank You for this video.
Probably If You have had cleaned and degreased PCB carefully before plotting, result was much better.
Good solution for cleaning is Cif - cleaning crema with microcristals. It removes well corrosion from cooper as well as grease.
You need to clean untill water covers perfectlu whole surface of the cooper.
Then ink should cover traces it perfectly and etching solution proof.
From my experience the best etching solution is B327 - Na2S2O8. It is not so agressive as acids, and not so dirty as FeCl3, but could ignite flamable materials.
This is what I'm looking for. However, you save me time and trial for thing that may go wrong. Your work is highly appreciated!
In school we used a tiny drill mounted to a rail and drilled the holes with that. Cheap and it worked perfect
Using photoresist and a moderate power 405nm laser would give you a much more resistant coating that would stand up to any etchant. I’ve used both dry film and the bulk “paint” resist sold on eBay with a lot of success. Ultimately the best results I’ve had are with an Epson inkjet printing onto transparencies and photoresist. The resolution is amazing and capable of imaging the finest of traces, without the troubles creating filled areas like a laser which will struggle with solid fills. Refillable carts are essential, and make it very inexpensive to run.
I have no experience making pcbs, but I was kind of thinking that pens would not work. What you really need is something like nail polish.
Idea for how to apply ink or nail polish or whatever very fast: make a stamp using a 3D printer. I have a little bit of experience with this, and what I will say is that a "hard" material like PLA does not apply ink to paper very well, because the paper and the bottom of the stamp do not fully come into contact unless you can deform one or the other. PLA works ok, but what you probably want is TPU (haven't tried TPU myself yet, but I intend to do that someday...). I'm guessing the same would be even more true for applying something to a pcb blank, since it's harder than paper, which means that you are really counting on the stamp to deform a little bit.
Back in the late 90's, when I was hand-etching PCBs, I used a Staedtler *overhead projector* (fine tip) marker, and it worked pretty well.
The Laser toner method (printed on glossy paper from magazine), then transfer on copper board with either hot ironing or nail polish remover is way faster, cheaper and easier.
Thanks for trying.
I recommend using a laser printer and transfer the toner with heat onto the pcb. This method is much much easier then the 3D-Printing method. Also it is a lot faster. Don't use a regular iron and paper to do this. Instead use a fuser unit from an old laser printer (you can get them very cheap as an replacement unit for some very old printers, just look around on ebay). Make sure to use a decent quality fuser which you usually do not get when searching for desktop printer fuser units. Instead look for a fuser from large office printers. Also use photo/glossy paper instead of regular paper. They come with a thin layer of plastic (glossy side) which helps to transfer the toner from the paper to the pcb. Know you only need to finde the right temperature (around 190 °C) and roller pressure (maybe you have to modify the fuser unit to fit the pcb) to get really good results. The process is limited to your printers resolution. I can get down to 0.2 mm layer spacing and thickness.
I showed this video to my teenage kids as an example of what hacking looks like. You're awesome, man. I _loved_ how you used the print bed as an agitator!!! (I think you could make this process work with more work on it.)
Here's a thought for using a 3D printer for PCBs: Use a laser to burn away the copper? That might not work great but...
Your process will work fine. I dunno what you're using as etchant but I've used those permanent markers to correct my pcb made with toner transfer and they work well with ferric chloride. Try using that as the etchant and i think your method will work great
Tip: use a magnifier you wear on your head like clock makers do for drilling holes in PCBs manually and good light. I use a swan necked led light to get the light at an angle that highlights the copper. I get really nicely centered results, ie ICs fit easily. I use a cheap drill press (voltcraft from Aldi) instead of a dremel which has more run out.
A tip for sticking things to mill/engrave/draw on: don’t use double sided tape, it’s a mess to remove. Use regular painter’s tape, the yellow or the blue and stick it both to the table and to the stock then use super glue on the tape and press them together, tape to tape. It will hold the stock as good as double sided tape, but it’s super easy to remove.
It’s funny to see someone make a PCB backwards in so many ways.
1) Always drill the holes first. Thin traces will tear off if you drill the holes after etching.
2) The orientation matters. If you print it in the wrong orientation, you will need to mount the ICs upside down, bending the pins up.
3) Ensure that the temperature of the PCB does not vary by more than 2 degrees F, between the drilling of the board, and the masking of the board for etching. If the temperature varies by more than that, the masking will not match the drilled holes, as the epoxy board will expand or contract beyond allowable tolerances.
Also, photo-resist etching is illegal in most countries (US & Canada especially), due the level of toxicity of the chemicals, and the amount of chemicals required per board. For 1 offs, something like this or a CNC is ok, but not optimal. It is fairly inexpensive to take your files to a printer to create 4x positive films, and then take the films to a fabricator, who can also plate and solder mask the boards.
The black traces floating off were kind of artistic. Farewell RAMPS!
I suggest when plotting to heat up the bed to aid in the ink drying. Also, I've had pretty good success with sharpies doing touch up on the laser printer/iron process. Not sure how hard it is to get a sharpie where you are.
Those black pens are best for when you use UV light, it is also the better way to make your PCB's anyway and you already have the tools to make a mini UV box (I suggest an old flatbed scanner).
Ferric chloride will attack the ink less than the acid you used I suspect.
I can use almost any permanent marker with ferric chloride
I was thinking he was gonna use that instead of the stuff he used
That works for sure. My father was repairing incomplete photolithography images that way back in the 80s...
I made PCBs with permanent markers (aka "laundry markers") with ferric chloride in the '70s. I wasn't the first.
You should explore if you could use the same process (hot end mounted pen) for something like calligraphy (think writing out wedding invitations or signing your name on a thank you card). It would be great to see the process you find from scanning a signature to creating an STL file to creating the GCODE etc. Cool video!
I love how 3D printers can do so many non 3D printer-y stuff
Nipuna Gunarathne like removing tomato’s from a salad
Who wants them in his salad tomato gets stabbed
Why not just printing a 0.2 mm layer of ABS or TPU on it?
abs? abs does not stick to anything cold, and barely to anything heated
the marker soaks in to the metal slightly, plastic is to thick and also will not stick.
I bought a $60 Laser Printer and a $5 magazine to print my PCB diagrams onto and a $6 bottle of nail polish remover to transfer the toner (perfectly) to my board. $10 gallon of Muriatic acid (which should last me nearly forever 😂) from the hardware store and boom, perfect PCBs. The acid eats the copper so quickly that the toner (which is melted plastic so it doesn't corrode anyways) has zero problems with trace definition. Even with traces as small as .254 mm wide.
I know the video is all about using the the 3D printer but the method where you print the PCB on a laser printer then transfer the paper print to the board with an iron, then etch is much faster, more accurate and a lot easier.
I have been using a cold toner transfer method for making boards for several years and have very good results. Then use mt cnc engraver to drill the boards
In the old days, I used Edding 3000 as the pen by hand. No problems with it. I wonder if you can try it with Edding 3000.
I ought to add a conductive pen or five to my next pile of materials and parts. Not for anything meant to last, but just to see how practical a fully-drawn PCB would be.
Could you use a weight instead of a spring to hold the pen down? It would give a constant force regardless of the height of the printer head.
Fun to watch. I really should finish printing my MPCNC parts.
I would like to see an attempt printing the traces with PLA on the copper clad, then trying to etch. Follow up by pealing the PLA off.
Post etch, it might pull the traces up - the glue that holds the copper on the phenolic board isn't horribly strong. Sanding the PLA away, however, might work.
@@jeffpotts6187
PCB are FR4/G10 not phenolic.
Modern Copper Clad is usually very tight.
Question is only how well the PLA sticks,
and if it resists the acid.
I use "phenolic" as a catch-all term when I don't want to look up specifics about what the copper is hosted on.
Secondly, I think you might be OK with Ferric Chloride, but with the other etchants people use I think it would eat away at the plastic. Ferric Chloride is considered caustic, but the others stuff generally IS caustic. With those, I think PLA would break down during etching. ABS and/or PETG, maybe not - but I wouldn't bet on it. I'm not a chemist.
That being said, it's less hassle to use a plain old Sharpie.
@@jeffpotts6187I see, you are not a chemist, you use catch all terms, and yet you know PLA won't work. Ok.
I did have good results with the method. I used the black Steadler pen with a different pen holder (it also had play though) and I used FeCl3 as etchant. I think the HCl/H2O2 or peroxodisulfate just attacked the ink too much - there is no need to use the hardcore oxidants.
For my projects I didn't continue using the method though because the pen tip is pretty coarse (couldn't really find pens with thinner tips) which necessitates large clearences between traces and you can only add the largest smd components without shorting pins. Also, chinese PCB factories ship insanely cheap. But if you want a DIY solution lithography is probably the way to go. You can even add your own solder mask by that method!
This is a really good start. First off, I think you need 3 linear bearings on the toolhead to fix the binding issue. This could easily be turned into something functional with a few changes. But that being said, milling is probably a better option, just due to the lack of toxic chemicals.
what toxic chemicals? fecl3 with the residual copper ions is able to be neutralized with sodium bicarbonate and can be safely disposed of... I fail to understand how people don't get that...
Apparently, you can etch PCBs with vinegar, salt, and hydrogen peroxide.
Hi Tom. It is easy to make a mount for a small drill that you could use to drill the pcb. Sure a printer tends to lack the rigidity to mill but you can drill only easily
I use ender 3 with a laser driven off the fan ttl output. Etch pcb spray painted black. Takes longer but i can use .4mm traces and double sided
The pads have no holes in, good luck drilling! Easy to do by hand with the right tool as they centre the drill bit, I used to do it for pocket money. Measure actual ink width and adjust g-code generation to suit.
Toner transfer works well with ferric chloride. It also seems that an Edding works good with ferric chloride, but I didn't really try that. And you can export the PCB in a way so you have circles instead of solid dots as Pads, so you know where to drill.
You can see the pen running out of ink because half is sort-of ok, and the other half is poor. Also solvent based ink tends to dissolve it's own lines if it touches deposited ink. Maybe a new pen and a smaller PCB and it might work. I put a 0.4mm holed pad on each THT pad as a guide to locate the drill bit accurately, etch and tin the PCB, and the step left in the centre of the pad is enough to guide the drill when I drill them.
Great video, not exactly the result you want but you (and we) learned things.
Thomas,
Could an FDM 3D printer with a small enough nozzle directly deposit a plastic etching mask layer to a copper-clad board? Would any common filament materials like PLA, ABS or PETG even survive submersion in FeCl during etching? You'd have to be careful removing the filament layer when you're done to avoid lifting copper pads, but luckily these plastics have a much lower melting temperature than copper or especially fiberglass, likely meaning you could burn it off without too much work. Maybe there is a second chemical bath that could dissolve only the plastic, leaving the copper and FR4.
I'd really love to see this tried out in the future!
Hmm have you tried using a wavelength laser that will only burn where the ink is drawn? You could use the ink to prevent scorching on edges with a laser that normally happens with laser etching.
You could always 3d print a drilling template with all the holes in the correct position, you'd need to trim the board accurately, or add in large hole markers so the two could be bolted together. That said, you are right, it doesn't seem like the best way to make them. I am pretty sure the original RepRap boards I've seen were mostly acid etched though.
Hey, here's a crazy idea: Why not print the tracks in PLA or another plastic that the 3D printer can handle directly on the PCB? I mean, that's somewhat the whole idea with PCB Toner transfer... toner is a mixture of plastic and pigments that gets melted on paper (or a PCB in our case) providing a pretty nice etch resistant mask. I think it would be quite fiddly to get this working, and you'd need a very flat PCB surface to begin with... anyway... I think this would be an awesome idea for a video!
Wow! Thomas must have made a zillion videos by now - impressive!
I tried once to heat a recipient with etching solution (ferric chloride), which worked well. The problem is that later I found some odd corrosion on the tool head. Maybe I should get the tool head higher. But maybe if I going to repeat this, I will build a heat bead with a mosfet and an arduino. Which will have the advantage that if I knock it over, it won't be over the 3D printer.
If you print a paper template and use a drill press that can handle very small bits, you can do a reasonable job drilling your own holes in custom-etched PCBs. Double sided with some alignment work too, but since you can't really plate vias that easily on your own and the alignment and setup steps are a lot longer, it's not really worth the effort, IMO.
I wonder if there's a liquid extruder head (maybe a syringe controlled by a stepper) that would be suitable to apply solder resist. The data for the layer is part of the gerbers that are generated, and they do sell liquid solder resist for touchup work, so if you connected it up to your printer head and could control the flow out of the syringe or other mechanism, you could probably do a decent job applying it.
Thomas, don't give on 2D drawing PCB designs onto copper just yet. I bought a pen in Japan called a Uni Prockey Twin, made by Mitsubishi Pencil Co., part number PM-120T, that has a tip on each end, one fine and one two or three times thicker. Don't know the diameters but I do know that it worked fine as a repair pen for a PCB I made recently using the nail-polish-remover-toner-transfer method. I suggest you outline the drawing using the fine tip first, then fill in the voids with the fat tip. This could preserve resolution without making wave patterns in the voids. I don't know if there's enough ink in the pen for a decent sized PCB but it won't hurt to find out.
Should have done this with photosensitive PCBs. Would have worked perfectly and you can use whatever pen works the best with the printer.
I personally use Diptrace for layout and GBR export (it's kinda similar to Eagle), CopperCAM for the g-code, and bCNC to send the G-Code to the Arduino controller. For me personally, bCNC is the only tool to use because it has a built in probe height mapping function. I put in the tool I'm using to route out the traces, and using the Probe pin (A5) connected to the PCB via an alligator clip with another clip to ground connected to the tool, then let bCNC map out the surface. It's a slick process and makes perfect depth cuts. For the holes, I quit using a drill bit and switched to tiny endmills, they make cleaner holes that drill bits in PCB material. I can knock out a decent size PCB in less than 25 mins on the router, small PCBs take only a few minutes, plus I get to tinker with machines =)
Spray the board with lacquer first (preferably white). Draw on the lacquer and dissolve the undrawn parts away with something like isopropyl alcohol. Etch with ferric chloride.
Also, your drawing technique from what I can see in the video is using a dabbing motion with the pen to fill in copper areas whereas the pen should stay down for each entire track (or as much as possible). The raster file seems to make no difference between track drawing and area filling. Again this may just be my view of the video file being wrong. Does it draw a continuous outline for each copper section first without lifting the pen?
Hi! I can see you have problems with the etching sollution when using pens. Only pen that I find that worked was the Edding brand. Staedtler pens were too weak against HCl and H2O2 solution. You could use Ferric Chloride or Sodium persulfate, with latter I did achieve good results, but the "resist" paint has to be good.
Check Edding 792 ...
When are we (you and the community) going to start building the ideal 3D printer?
Probably never as there isn't really such a thing, everyone has their own use cases and specific requirements.
To etch FAST (