Applied Science - 2018-09-03
I developed a new chemical process to make double-sided PCBs with a laser cutter, and some common chemicals. This is intended to be used on SLA 3D-printed parts, but may work on many materials. Refs: http://infohouse.p2ric.org/ref/29/28004.pdf http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.527.5817&rep=rep1&type=pdf http://shodhganga.inflibnet.ac.in/bitstream/10603/195088/3/chapter%202.pdf https://sci-hub.tw/https://doi.org/10.1016/S0167-577X(03)00424-5 https://library.nrao.edu/public/memos/chem/CHEM_5.pdf https://core.ac.uk/download/pdf/48645791.pdf Support Applied Science: https://www.patreon.com/AppliedScience
Brilliant work as always! Most impressive is how you have managed to keep a wooden tabletop in pristine condition in your lab.
Incredible work! The electroless copper recipe alone is something I've been trying but failing at for a while 🤯
Reading these comments may have induced my first experience as a Patreon snob. It's probably bad but it feels good.
Ground braking, as in lithobraking?
Love you man <3
I always love it when I see one person I subscribe to commenting on another's channel! I love both of you men's channels!
Marco, could you try this on your cheep Eleksmaker to see if it has enough power for the traces?
Hey Ben, great video and investigation. I work for the company that made the triangle widget (Multiple Dimensions AG). This technology is known as "3D Molded Interconnect Device" (3D-MID). You are right about electroless copper plating approaches being closely held secrets. We have 10x the throughput at our facility than our competition and this is one reason why our Swiss company can be globally competitive against Asian competition for these types of products. The "Laser Direct Structuring" (LDS) lasers that we use are expensive and patented by the LPKF company -- they run at 1064nm wavelength from 0.1 to 5 watts which finely ablates and excites the copper-organic dopant that is added to the plastic granulate. This excitation "bursts" the added organics off the dopant and leaves a trace of copper dots on areas that are being ablated by the laser. The combination of ablated plastic areas with seed copper makes these areas attractive to electroless plating. If you or your followers would like to learn more about mass production of 3D-MID please reach out on our contact page : multipledim<dot>com/contact/
@Igor Ovcharov can i use a lemon instead of a lime ? ;-))
@aga
If you can whitewash the walls! That of course!
@aga
You drink too much alcohol and spend too little time on productive labor! ;-))
@aga
Lime in this case, I meant a substance with the chemical formula CaO
@Igor Ovcharov You're wrong, there's no CaO in limes
your channel is such a gem. thanks for your work.
I think the research you did is enough for a degree thesis, crazy dedication!
Rexus King And this is the stuff he does for fun. Pretty amazing.
Have you tried acrylic or polycarb sheet instead of 3d prints for flat parts ?
Re. the direct laser+copper oxide method, how about post-electroplating to thicken up the copper?
There is a UK company that has a process where they inkjet print catalyst, then electroless plate - a few years ago they did a transparent arduino. I wonder if there may be a way to pattern the catalyst to avoid needing a laser- screen-print perhaps?
@SuperAWaC If you mounted an IC you could use ultrasonic bonding to bond the bond wires of the IC to the plated metal. There’s another interesting subject, ultrasonic bonding/welding of metals !!
@goamarty Yeah that's how this company does it:
https://www.eurocircuits.com/Making-a-PCB-PCB-Manufacture-step-by-step/
@Martin Westermeyer Gondonneau Yeah that's actually a really common process, as it saves copper/chemicals and time to only build up the copper where it's needed, here's a detailed walkthrough of a common process: https://www.eurocircuits.com/Making-a-PCB-PCB-Manufacture-step-by-step/
That said they don't detail how they make the electrical connections, but I'd say it's like applied science says: they probably use a sacrificial section of the pcb to short the traces, then break or route it off, or perhaps they actually short them together on the board itself then route off that trace hard to say. They could on simpler boards use a shorting connector that shorts and connects to the traces at the same time I suppose.
@Martin Westermeyer Gondonneau can't see down there if you've been properly answered on that topic, but what you're wondering about the post plating making everything plated together is what the catalyst inkjet printing is about: you add catalyst only where you do want the plating, so you don't coat everything on it like on this video.
Here you also pretty much just add catalyst to the places where you want to plate (the oil avoids it and the laser + cleaning thingy adds it), but on the process OP talked about a printer deals with that part and the rest would be the same as I understand it.
Why not use conductive ABS in one filement and normal ABS in the second trace your Carbon tracks and add thin break offs onto all the circuit nets to keep them conductive then just electroplated copper with copper sulphate. After building up a good copper serface add elecftoless Tin to aid soldering .
Question 1: For makers to practically use this process, they'll need a way to properly dispose of the waste. How do you do that, and what are some things one could do to minimize the volume of waste generated by the process?
Question 2: Did you try Triflow Dry, by any chance? Rather than an oil, it's just PTFE and paraffin disolved in an aliphatic hydrocarbon solvent, and it leaves a dry PTFE/wax film after the solvent flashes (It's fantastic on bike chains because the dry film doesn't accumulate road grime).
Awesome video as usual!
Hats off to your detailed scientific approach! Thanks for sharing!
That is an epic piece of DIY research Ben, thanks so much for sharing!
If you want to save a penny or two, pool algaecides (at least the common ones you get in Australia) are a higher concentration of Benzalkonium Chloride than Wet & Forget and much cheaper!
Good call. The label of "Pool Time Algicide 50%" I just looked up on a US big-box site says it's 50% alkyl dimethyl benzyl ammonium chloride. Looks like "Clorox Pool and Spa Green Algae Eliminator" on a different colour of big-box site is the same.
If you vibrated the part while in the plating bath every X seconds, it should shake off the hydrogen bubbles. Perhaps by mounting the piece on a rigid "stick" instead of a wire, and having a small motor with an off-center weight attached to it (like pretty much all cheap vibration happens). Or maybe a transducer?
Think of it like smacking the side of a glass to shake soda bubbles off.
Ultrasonic bath perhaps?
Wouldn't it have to be an ultrasonic bath that can also maintain the temp and stirring requirements of the copper solution? That's why the transducer idea appeals to me.
I like that idea. I was also thinking that one could bubble nitrogen gas instead.
@n00bkill @Bielanski - I'd be wary of using an ultrasonic bath. The cavitation would likely "clean" the copper off the workpiece as it formed, "clean" the catalyst off the workpiece, or destroy the workpiece itself.
The intention was not to have it constantly vibrating. Doing so might disrupt the copper formation/adhesion (though testing would be required to determine the actual impact).
However, if there is concern that hydrogen bubbles may be keeping the solution separated from the workpiece, "shaking them off" from time to time should help reduce their impact.
My initial thought was something like a 50-100ms pulse, once every 10 seconds. And I have no idea which axis of vibration would be most effective.
I was initially thinking about this from an automated perspective. But perhaps sharply tapping (or having something tap/actuate) the piece downward every 10-15 seconds may do the same job. Though, without automation this would be INCREDIBLY tedious if the bath treatment takes several hours.
That's what I was thinking too. Just scavenge a part from a cheap kids toothbrush after the bristles have gone past their use-by date.
To protect against flux getting under the traces, have you tried to cover everything except contact pads in solder mask, so copper edges are always under the solder mask.
Антон Южаков Maybe it would be possible to somehow do the solder masking with a 3d-printer again? Maybe this doesn't produce enough adhesion however?
Solder masking is usually done with UV reactive compound and a stencil.
I think the problem would be that if you have the equipment to precisely place the solder mask on the edges of a pad, while still having enough solder paste/flux on the pad, You probably aren't doing it at the hobbiest level anymore.
@Overwatch you could do that by mounting a slightly defocused low power UV laser on a 3d printer or laser cutter and then using it to do a raster on photoresist solder mask.
Could you use the copper traces from the copper-bearing epoxy as the starting point to further plating? It seems like that would be a good shortcut around the moss kill steps. Also, it might be less susceptible to lifting from the flux.
I wonder how stable it will be, could it become conductive (in areas where weren't traces in first place) in matter of time especially in warm environment.
It would probably need to be extremely hot for it to become conductive
Very good idea. I wonder if there will be a follow-up video.
You can actually deposit a thin, 1~10 micrometer thick or if you can slow down a bit later even thinner copper layer onto your substrate, then cover it with photoresist, where you can create rather large aspect ratios, i.e. the wall can be up to e.g. at least 98 degree steep with the right photoresist, and then subsequently just electroplate your traces into the voids you created with the exposure. Then you remove the photoresist and etch the whole pcb a little bit to just barely get all the non-plated parts of the original thin copper layer off of the pcb, but not so much that you etch under the traces. Basically this allows you to get traces being >90% copper within a square enclosing the cross section, and you can reach that metric down to about 3~5 micrometer side length of the square. You should be able to stack up on top of this via spin coating more resist to fill the voids between the traces once again, and then depositing another base layer for further electroplating. Though, these techniques seemed to use less amateur ways of depositing a thin base layer, iirc startign by sputtering some metal and then continuing by wet chemical deposition (basical electroless plating, but restricted to metal surfaces or so).
And if you can get a way to see the exact positioning of a test projection with e.g. red or yellow light, to figure out how many pixels you need to shift the 405nm led/laser projection, you could attach a miniature dlp beamer, or the guts of one (weight/size/binary-high-frequency-pattern-mod) and use a rather cheap, low resolution but mechanically stable/no-wobble gantry to stitch a slightly downscaled (2~10 times smaller than on the dlp, if you care about reaching the limits of your optics/etching) projection via some stepping over the board. Even if you don't downscale (1:1) or even upscale 2x on some dlp chips, you can get etch-factor-limited trace widths with common 35 micrometer and at no more than 2x downscaling with a low resolution dlp even for 17 micrometer (aka, half-ounce) copper. This is a bit better than normal chinese pcb fabs deliver, with very minmal turn around time, but unfortunately the benefits are just for <5pcs runs and/or semi-additive processes (the previously explained electtroplating between the photo resist).
Excellent video sir! I worked in the PCB industry for many years. We were able to make our own eless Cu baths, but could never match the stability and speed of the proprietary stuff.
One quick note, we always had a fluoroboric acid dip called an accelerator between the catalyst and eless bath. I was given to understand that this removes the exposed tin from the colloid exposing the platinum to the eless solution.
Thanks for the awesome work you did to make this process work for the home gamer!
Awesome and inspiring cutting edge stuff Ben. These experiments must have taken a lot of time. I truly appreciate your dedication to openly advancing the state of the art!
Awesome! This is amazing work thank you for sharing!
As always, facinating. Gives me faith in society that we have smart dudes like your self pursuing STEM for professions and just shear curiosity.
But for every smart person like Ben, there are X number of criminals, brain-dead drug abusers and people leeching off the system and not contributing. Unfortunate truth...
yeah and i wonder what happens if the all people are "equal" (kill whites or mix them, see europe crisis) Its the declared goal of the leftist ( google EU doctrine replacement Migration).I think the movie ideocrazy describes it very well.
https://www.youtube.com/watch?v=CXSz0bA9CiE
Hidden lesson on the video: Toothbrushes make great pointers. I love your videos. You're a great teacher.
i wonder if you could get copper traces on a business card...
You could always order a thin PCB with the right size and use that. Of course that would somewhat limit your aesthetic choices.
the whole point would be to do something crazy like make a pcb out of a regular business card
A guy here on Youtube that goes by the name Mitxela made Stylophone business cards. He had the PCBs printed by a commercial service though.
@SuperAWaC might be doable using conductive silver ink. Soldering it would likely be destructive though. Even if you managed to not put it on fire, for example by soldering under an inert atmosphere, would still likely char the paper. This could make the final product unsuitable even as a circuit board because of conductive carbon formation. You might still be able to pull it off with some plastic material having the outward appearance of approximately glossy paper.
@Gameboygenius : If using conductive "ink"/paint traces why would you want to solder anything use conductive resin or thermoplastic.. (Compatible with the traces of course.)
"It seems like surfactants and sort of controlling surface affinity is a pretty big deal."
XD
This is really great! I always wondered if I'd ever be able to do via's in a makerspace without expensive equipment. Definitely try this out as soon as I get the chemical supplies.
This is great news especially in a post Radio Shack world.
As usual on this channel, the video is so clear that there is not much to ask regarding the process shown.
16:45
Science: "Activate test 66"
Ben: "Yes milord"
You're endurance is amazing. There had to be times through this project when you considered giving up. I am also impressed with your ability to source some of these chemical components from commercial products. Excellent work!
Great video as usual, have you tried silicone based oils? in my experience , after been burned by a laser, it turns into very hydrophilic silica nanoparticules that sticks very well to the surface.
aetius31 would that actually help? I guess a cleaning process is already done so it wouldn’t hurt, but the current oil seems to work fine.
@zach burke The point is that it could make a 3d template for the copper so the layer could be thicker and/or mechanically stronger.
aetius31 so it’s a question of if it gets removed in the cleaning process?
Always something new and interesting you teach about.
Your dedication to thoroughly exploring a topic is most impressive. Thank you for sharing your findings.
So might be possible covering FDM printed part by this solution (epoxy + copper oxide), lasering it, then using galvanic method to thicken/improve this copper traces?
But i doubt this is practical.
There are also many more plastics than PLA.. PETG and PC come to mind.
You could use Argon instad of the Air for the bubbler. Nice Vid.
Argon or nitrogen defeats the point.
He put the bubble in there in order to introduce oxygen. little point in having an argon or nitrogen bubbler.
It was to cause disturbance and knock off bubbles, not to have oxygen in there afaik.
@[Whatever name goes here] No, he said the oxygen was to stabilize the solution. The stirring action was supposed to be an added benefit by turned out to be a negative.
lol The plating solution looks like the blue demonstration water they dump in to Depends on the commercial
@Gordon Wedman Yes, but the negative was the oxidation of the copper, from oxygen.
Stupid question, wouldn't it be easier to sputter some copper on a piece of plastic and then put the whole thing in a copper sulfate bath to grow your trace according to your spec ?
I think the end goal here to develop an accessible method for making (3d printed) plastic parts with integral circuit traces rather than a plastic part with integral circuit traces . Assuming that that's the case, then sputtering's probably off the table; AFAIK, 3D printers and laser cutters are fairly common in hackerspaces (and other similarly themed places), but I suspect that sputtering setups aren't.
Probably not. Sputtering onto plastic is pretty difficult, the adhesion is really bad (depending on plastic). Really you want to sputter Ti first, then place a layer of Cu on, either by sputtering or electroplating.
That copper oxide process at the end is really interesting for not requiring a bath. If you modded a 3d printer with a laser as a second head, you could put down copper in places inaccessible to a bath.
Hey, Ben just out of curiosity have you ever seen the channel ROBRENZ ? Amazing machinist and some electronics content too. Seems like a channel that you may enjoy, I certainly do. Thanks for all the awesome original content you create.
How about indexing the 3D prints so after you apply the copper to the print, you can put the part back in the printer and continuing to print more plastic on top? This could generate multiple layer PCB's and also potentially be used to create a solder mask that could be used to assist in keeping the flux from getting under the copper if the copper lands were a bit larger than the mask, so the mask covered the exposed edges of the copper.
26:21 That sounds like some kind of armor against energy weapons.
I frikin love tri-flow. Use it on everything. Its a PTFE suspension in synthetic oil.
When I was < 10 in the 80s, my dad bought cheap as tea bags that, oddly, came with a tiny (~4cm) ceramic animal. Saturdays were spent drinking tea, and plating and etching circuit boards. One saturday I asked if we could plate one of the tiny animals. Sure enough, a few weeks later my dad had obtained the supplies from his job that we'd need to plate the tiny animals. FWIW he worked at a chemical manufacturer and supplier in silicon valley. We used a plating process very similar to this... however we cleaned the animals in a process using Dawn dish soap and ammonia, and had good results.We plated a dozen or more animals with copper over a few weekends. My dad had a background in chemical engineering. Once we ran out of animals to plate, my dad used some of the chemicals to make explosives. The copper made brilliant blue-green sparks and flames. A few of the explosives were "lit" by dripping ammonia on them. It was a lot of fun! I took the copper animals to school to show off what my dad and I had been up to on the weekends. Many of my classmates thought it was pretty cool, of course quite a few thought it was stupid and lame, or just didn't believe me. My teachers were impressed and a bit concerned. It was a lot of fun... until... One teacher asked how we were disposing of the "dangerous" chemicals. I remember being asked specifically if we put the chemicals down the toilet. That's when I talked about using some of the chemicals to make the explosives/fireworks. "Some of them you can pour ammonia on and they explode!" My teachers went from mostly impressed if a bit concerned to extremely, feverishly, concerned that my dad was making extremely dangerous chemical mixtures that were going to blow-up his apartment, or create highly toxic gases, and kill me. So concerned were they that they got police and/or child services involved.
Thank you so much for sharing the electroless copper solution!
You learn new things every day. Thanks for sharing! :) If you could make electrical contact with each trace, could you then electroplate them to give them a better finish and better conductive properties?
I was wondering what you'd been up to. Not disappointed, this is awesome stuff.
To Applied Science: Thank you for such a well detailed and valuable presentation!
Very cool results, thanks for sharing the process.
This is incredible! Thank you for working at this--I loved it!
Ben, I don't have a question but I wanted to say thank you for your inspiration. It's fantastic motivation to make more stuff!
I really love this channel! Would be great to be in the lab with you, so much knowledge and wisdom here!
This video just made my day. <3 Amazing work
You, Sir, are absolutely awesome! I'm starting to understand that the DIY SEM is only the tip of the iceberg... literally awestruck, I am. Thank you!
0:06 this looks awesome)
Kacper Cieśla - 2018-09-03
You madman. I'm amazed that you can dive deeply into so many different topics with so much dedication.
DrTeddyMMM - 2018-09-05
Edison took the discoveries of people working for "him" and called them his own...Edison he is not, more on par with Tesla, let everyone enjoy the secrets of discovery...
Kenny Phillips - 2019-06-16
@DrTeddyMMM Actually, if I hire you to help find, for example, the best filament for an incandescent light bulb, and you find it, it's still MY invention.
DrTeddyMMM - 2019-06-25
@Kenny Phillips Give credit where credit is due..
Do a little research, Edison was a dick...exactly why when Tesla worked for him Tesla told him to fuck off and left his employ...
aga - 2019-07-12
Edison was basically a ruthless Businessman, not an inventor.
In his time, he met Nicolai Tesla, who was without question a Genius.
The guy making this channel is more like Tesla than Eddison.
BioDOC - 2019-12-23
@Kasper Cieśla you're so right, he's a real all-rounder 👍
the great thing is what he does, he doesn't just do it superficially
you just can see that there's a lot of knowledge behind it