Clear flexible printed circuits - first success

[@monkfry]

I met the guy who invented powder coating. He was trying to accomplish something entirely different at the time. Now he's filthy rich and still doing weird things in a much nicer shop. The world could use more people like you. Thanks for posting.

[@androidkenobi]

@¿o? probably for developing insulation, corrosion/abrasion resistant surfaces, and succeeded, but the paint thing then really took off

[@arthurmead5341]

Daniel Gustin?

[@AppliedScience]

There's a really delicate balance between rejecting ideas that require too much future technology, and supporting ideas that will be solve problems in a reasonable amount of time. For example, I'd like to work on brain-computer interfaces that will change how we access information. It's basically an electrode plate that is mounted on the cortex and has bio-stuff to connect it to neurons. Safe to say it will not happen for several decades, and working on fMRI and basic science will help more now.

[@davidwilsonjunior]

AI model to read my dogs thoughts

[@kuntosjedebil]

The Applied Science channel is clearly heading towards having an OLED display in a contact lens. Can't wait to see it!

[@S1rRender]

kuntosjedebil Then Apple will finally have their Eye - Phone

[@rongarza9488]

Mojo Vision

[@ToddFun]

In college, 25 years ago, we made deposited circuits just like this with aluminum deposition. We created circuit masks using thin sheets of CNC'ed aluminum and could move the masks inside the chamber without losing vacuum. We created traces and capacitors mostly. For the dielectric we used stearic acid. We used super clean class and semiconductor wafer substrate. Our chamber had two heater bowls, one for aluminum and one for stearic acid. You get the picture I'm sure. Learned tons working there.

[@RimstarOrg]

Using the same process to make the components sounds very interesting. Looking forward to videos about it.

[@neotoy]

Nice to see a professional of your caliber taping contacts to a battery, gives me hope for my own amateur-hour projects.

[@atotheroh]

Ben, I just have to say that your videos are inspiring and incredibly informative. You have a gift for making science accessible and infinitely interesting. While I'm somewhat convinced that you will one day use your powers to take over the world, I am also convinced that in the mean time, you make the lives of others more enjoyable and increase the intelligence of the human community. Thank you.

[@o.h.8970]

Hey, I discovered your channel today whilst looking for DIY evap chamber advice online. Wonderful channel! I'm currently a PhD student working on thin film electronics and I think what your doing is awesome, even more so that you're doing it DIY style! You could pair this very nicely with your Raman spectrometry setup to gain a lot of information on how your organic layers are doing. Same with the SEM too.

[@svampebob007]

You never cease to amaze me, great idea performed beautifully!

[@wesleybecker834]

I think some capacitor(s) cracked, hence the fast blinking rate (stray capacitance is low capacitance). MLCCs and flexing don't go together, they usually break under the terminal and you can't even see it. They love cracking from unequal heating as well, like soldering on terminal at a time. When there's a clear bending direction, you can lower the risk of them cracking by aligning the cap's length orthogonal to this bending direction.

[@ChrisTacklind]

Ben and Spoif, We've used the toner method for years with variable success. We saw something called "Toner Transfer Paper" in DigiKey and bought some. We finally tried it. Wow! We used to soak the paper forever and struggle to get it off. This stuff transfers all the toner cleanly with just a little water. Digikey Part number 182-1003-ND Expensive but worth it!

[@AppliedScience]

Yes, I was thinking of very similar ideas for cheap, easy-to-apply electronic circuits.

[@Spacefish007]

I made a small PCB back in the days by using sticky tape on thick paper, then spraying EMV-Paint (Paint that contains a lot of copper particles, used for shielding against electromagnetic waves). After that i removed the Stickytape and soldered / glued / stickytaped components onto it.Wanted to use this technique for a LED Ceiling, but it turned out, that the EMV-Paint has a too high resistance with archivable layer thicknesses. A later on used aluminiumfoil and adhesive!

[@rubber20021]

Thanks for sharing this with all. A relatively simple concept in theory, but harder when you approached it... Thanks again!

[@CNLohr]

One of the big steps forward for me (though I haven't made a video on it yet) was when I started using Riston. You can get very fine traces by using a UV source to mask off the Riston. The big difference for me was, you don't need to mechanically remove the mask. You can do it chemically. I haven't printed too many glass PCBs with it yet, instead of toner transfer, but that's changing. P.S. You may find 42/58 Bismuth Solder Paste interesting. It melts below 300F!

[@Higgs666]

I've been enjoying your videos. You've been a big inspiration in my construction of an at home low temperature physics laboratory. I'm trying to develop a means to liquify helium on the cheap at home as well as experimenting with low temperature rocket engines. Keep up the good work.

[@Trevorm15]

Wow. Awesome video. I love your ingenuity and drive. Thanks for creating great content!

[@FamilienSoelberg]

Ben you have done it again. I just love the stuff you do. And congrats on the success with the flex.

[@ChrisB257]

Kudos for perseverance Ben - and ingenuity!! Fascinating project. :)

[@worldgeektube]

Is this Ben's full-time work? He has so many skills and so much knowledge with the ability to explain it all. I wish my science teachers had been more like Ben.

[@BotFactorySquink]

We met at MakerFaire 2015 Ben - you saw our circuit printer the BotFactory Squink. I didnt realize at the time that you experimented with flex circuits! We've been able to print on a lot of new materials using our low-temperature metal sintering method, inkjet printing silver nanoparticles onto substrates like glass, thin film plastics, etc. Might be a better and much faster solution than vapor film deposition! In addition, you can print insulating layers to 'seal' the traces and improve flexibility to 1 mm radius bends...

[@AppliedScience]

Sounds great! Is your machine available for sale?

[@avibank5115]

Yea, but last time I checked it cost as much as a small house. Hopefully I'm wrong. But Squink is definitely a major step forward !

[@alexk4709]

Looks like they're currently about $3,199-$4,499

[@tompan6268]

I'm working in Gapel Electronics Co.Ltd,and we're a professional manufacturer of PCB,(prototype and mass production) and FPC in Shenzhen of China for over 10 years. I've checked your website, you have the requirement of PCB and FPC. We'll try best to give you best offer to save your cost. Please feel free to contact me,if you have any new inquiries. Your prompt reply will be highly appreciated. Best regards Tom [e-mail]

[@AppliedScience]

Thanks! I use sodium hydroxide, followed by vinegar to neutralize the base, then isopropyl alcohol. The process requires a bit of scrubbing, but isn't too bad.

[@AppliedScience]

Thanks! Eagle does export PNG, which is how I got this working. It would be nicer to have a vector format (svg, dxf, etc), but the dxf exporter seems broken, and there aren't many other choices. Going through gerbers and CAM, then trying to get back to plain vector format was also was a big waste of time.

[@TaggingTechnician]

I would consider a process that makes the clear plastic substrate "optional" : use it for the circuitry construction and assembly only, with a final step of separating the clear plastic from the product. this strategy results in a more resilient product although it requires additional deposition in every stage of manufacture. well done.

[@AppliedScience]

Search for Chip Quik.

[@activemotionpictures]

Thank you for sharing this insightful video. It's so inspiring! Very well explained.

[@AppliedScience]

It's true that my process for cleaning the glass is pretty terrible. Plasma cleaning would be the best!

[@Falcrist]

This is by far THE coolest comment section I've ever seen on youtube. It makes me ashamed that I don't have anything to contribute.

[@geomonicdotorg]

Two thoughts: 1) brittle failure of flexed traces could be extended using a light flexible substrate applied to the top of the circuit - basically a circuit sandwich. 2) Richard Crewe made an interesting point, as my first thought was, "Wires would flex just fine." I wish I still had my old Perkin Elmer 400, I could whip you out a mask of your circuit with thinner traces for much heavier deposition that mimicked actual wires.

[@ladyborg6]

So cool. If you stuck the components in position on the substrate you could put it all in the metal evaporator and print the track on to the end if the components. If they (and glue) could take the heat, and you stopped it from printing over the whole component so it didn't just short out.

[@akc7100]

My thoughts exactly. Granted you thought of it first lol. The vinyl should hold and shield the components as well

[@davidfowler7019]

I've played around with mirrors and a laser cutter. The mirror backing is an extremely thin silver, then copper, then enamel. I could remove the enamel with the laser fairly easy but soldering to the copper layer was nearly impossible as it would just dissipate into the solder. Might be interesting to try some things with that glue though. I've got lots of ideas but little time to apply...

[@MrBranboom]

Awesome presentation as usual.

[@whidbeyman]

Use silver bearing solder to reduce the tendency of the solder to dissolve the thin traces. You can also use Bismuth based solder, which has a much lower eutectic temperature. Admirable effort, but much easier to plate traces compared to deposition. See also a UK company called Conductive Circuits. They have a very inexpensive process where they inkjet print conductive ink and plate copper on top of that. Cost is about $10 per square foot plus overnight shipping.

[@JiveFowl]

Have you considered also cutting out spots for your components in the film? It would put the component on the same plane as the bending, and afford the use of adhesive as a sort of structural backing, which would keep the conductive glue from crumbling off. Arguably adding rigidity is the opposite of what you want to do, but it might be a stopgap in the meantime.

[@JS-110]

Very impressive Ben!

[@scholi3039]

Hi. If you can sputter two different metals without venting you vacuum chamber then you can use a thin adhesion layer of Titanium or Chromium (1-2nm is enough). This works very well on glass.

[@naikrovek]

I'd love to see you deposit a thin film transistor.. Holy cow. You'd be printing your own chips soon afterwards.

[@geomonicdotorg]

One last thought: Make a heavy copper deposit onto the flexible substrate using no mask at all, then proceed the old fashioned way, using photomasking and acid etching followed by an additional substrate above for a flex-strengthening sandwich layer with small cutouts for high profile components.

[@RSP13]

That's a GREAT starting point.

[@YSPACElabs]

For getting a vector format for vinyl cutting, you can use gerbv's CLI interface to convert a gerber for the copper top layer from EAGLE into an SVG file. Then you can use Inkscape to convert the SVG to DXF. Silouette Studio will import DXFs, but make sure to set the scaling setting to "as-is" instead of scale to page size. I figured out this method when trying to get a vinyl cutter to cut mylar film for SMD solder stencils.

[@kraklakvakve]

Also works: Eagle - CAM Processor, Device - HPGL, select layers and file(.plt), Process Job. Inkscape - open file.plt, ungroup, Path - Union. Also gives nice polygons.

[@gamemaster1324]

Cool idea! i can't wait to see the result.

[@NoxelectronBlogspot]

Great video and great idea with the integration. You should try to make a touch sensitive capacitor. By depositing Cu on one side of the foil and putting together two foils with the Cu facing outside. So you could make a few capacitors on one foil like a few touch switches.

[@mikeselectricstuff]

Do you use a voltage gradient to make the metal move to the target, or does it find its own way - if so, what makes it move upwards?

[@MeatVehicle]

You have some very interesting ideas there.

[@kraklakvakve]

This worked: Eagle => gerbers => gerbv (viewer), select layer, export SVG => Inkscape, ungroup, Path - Stroke to Path, Path - Union. This produces nice polygons. Not so great for flood fills, though. Flood fills in gerbers are drawn with many many strokes, which gives Inkscape a hard time.

[@queazocotal]

The same exact problem is faced by 'proper' flexible circuits, though solder is lots stronger. The approach there is simple stiffeners. A square of more rigid plastic behind areas with components helps lots. The epoxy isn't the only problem - the track adhesion is not great, and a really strong flexible glue risks ripping it off.

[@cosmiceon]

I would love to see more of this!!

[@feliciahummel6642]

You can export from Eagle into PS by using the CAM processor. You simply select PS as output format, then select the layers. I'm using the CAM processor to create negatives for etching PCBs. You could even convert PS to PDF using Inkscape.

[@omsingharjit]

At 3:38 you are going to fabricate transistor like ic fabrication ??????

[@AppliedScience]

I just saw the stuff on SparkFun's video, and plan to order a roll. Thanks for the tip!

[@phaub]

If you are looking for a way to have better adhesion on glass (you mention this at 3:40), from what I remember, the solution is to first deposit a thin layer of chromium, and then the metal you are interested in.

[@Sedokun]

For high aluminum adhesion (25MPa), glass can be coated with Hypalon (CSPE CSM). It's durable and heat resistant. I'm sure there a lot of other primers exist. Don't forget, that aluminum should be covered with copper to prevent electrochemical reactions. But there are limitations cause thermoelectric effect.

[@MorRobots]

This is awesome, also I would take a look at the code to see if it outputs G code. As for ideas, you should see what it would take to deposit the materials to make an LED.

[@frostbyt1134]

Ben, a suggestion for making much more flexible components would be to use a glue with a higher elasticity, something like polyurethane or silicone. Perhaps some homemade glue with a colloidal suspension of silver or aluminum?

[@DigGil3]

Hey Ben! You could try "sandwiching" the components in between 2 plastic transparencies so to keep them from falling off. These 2 transparencies could even have different circuit tracks to allow more complex connections

[@conradteves182]

How about a glass sheet (or cylinder?) with a circuit pattern between electroluminiscent paint like LumiLor? A little extra circuitry and you could get different parts of the pattern to light up when you want. Great video Ben!

[@BigAdam2050]

I still wonder if all these projects, ideas, builds are working towards something. In which case, its probable going to be horrifyingly amazing thing.

[@GeorgeGraves]

You can use a "print to PDF" printer driver to get a vector out of any program. I use the Adobe software that seems to come with CS4. But I've also used "do PDF" - both worked great.

[@bikefarmtaiwan1800]

Great idea Ben. Thanks for the excellent video and thanks to the watchers for interesting comments below! I wonder if after sputtering you could electroplate copper ( or copper friendly) traces in a simple copper sulphate bath to improve the conductivity.

[@RC_Engineering]

Are you talking about the evaporation? In a vacuum there is no air to block the aluminum, so it just flies in a straight path under the effects of gravity. Particles move pretty darn fast, so many of them fly right to the target.

[@cavemanelectronics9611]

you are truly a modern magician.

[@AtlantideVFX]

Great work! I love watching your videos.

[@franklynch7267]

In terms of making your template, there are some videos on screen printing designs on youtube using photosensitive emulsion paper (used to get patterns onto t-shirts etc). The template is simply printed on an inkjet printer and then used to exposes the photosensitive emulsion paper. This leaves you with an easy to create template. hope this helps you.

[@aaro1268]

You could try laminating a copper foil (or deposited circuit) within a plastic transparency on both sides. Ideally, the components would be made out of the wire itself, as solder and through-hole components appear to be fragile in heat and flexing. By punching holes in further layers and using conductive resin, you should be able to make multilayer flexible PCBs.

[@OskarNendes]

Nice. Always finding interesting solutions here :)

[@Bigman74066]

Maybe you can cut the traces directly into copper foil and stick it onto the transparent foil. you can then solder components onto it (if the adhesive can withstand such high temperatures)

[@markbell9742]

Hi Ben: Poor adhesion of Al-Glass, suggests contamination. Aluminum on a well coated mirror should not peel off when scotch tape is pressed on and pulled off. I just tried this (several times even from the aluminized edge) on a 40 year old diag mirror with zero Al lift-off. Assuming you’ve super-duper cleaned the glass you could try adding a shutter so the 1st stuff is not deposited allowing a degas/clean-up phase, during which the vacuum should (might) get better as the Al acts as getter. Mark

[@DJ027X]

Component pops off, circuit keeps working. That's skill

[@RukusDM]

always epic stuff. Keep the vids coming. Great work.

[@samykamkar]

Love it!

[@PureChaos69]

Oh my gosh this makes so much sense, thank you! I tried making home made co2 with baking soda and conifer (that does make co2 right?) and trapping it in a container then freezing it.

[@shpixi]

...how about painting with conductive glue straight on vinyl negative?

[@110110pab]

i have seen printable solar panels and wondered if they would have the same problem, ..ie: break if flexed too much. or if they might be able to be printed on to a boat sail. your printable board would certainly offer a way to connect it , BTW, i love this channel. keep up the good work .

[@JamesSleeman]

Rather than the vinyl cutter which wouldn't be very feasible for more complex circuits you could probably use dry film negative resist (plenty of it on ebay), just stick it on, run through a laminator (or a cool clothes iron) to bond, expose, develop in sodium carbonate solution (20 to 40g/L) and then do your deposition, strip in stronger/hotter sodium carbonate.

[@samfosteriam]

I've been playing with cutting the circuit directly out of copper tape and putting that on a plastic substrate (I envy your vinyl cutter) Polypropylene is sufficiently heat-proof that I'm able to solder components to it if I'm quick. I've not had any luck with the conductive adhesives I've used so far - they are brittle, dont form a good bond and dont conduct very well :(

[@iamdisccomp]

Instead of vinyl, you just need to find copper sheets of desired thickness with adhesive on one side. You can then cut and apply to whatever substrate you want in one step.

[@totallynotabot151]

Silhouette Studio imports DXF but is somewhat picky about the precise settings. There is also an SVG option but that requires their Pro update.

[@AppliedScience]

I'm pretty sure that my glass surface was dirty. Cleaning with plasma is the best way, and I don't have that set up yet.

[@gigicaly]

why not drill little holes in the points were the epoxiglue is applied and apply also from other side for a better grip?

[@AboveDisturbing]

Holy hell, I would wager you could use this to make extremely lightweight flight controllers for drones. Anything to reduce weight would be awesome

[@OoFREDIEEEoO]

I think is possible to make really elaborated circuits with a Electron gun deposit, that would be a cool application!! :)

[@qwertyy37]

I don't know about the feasibility of this idea, but in regards to improving flexibility, maybe the circuit can be deposited with the substrate already in a flexed position. That way, when the circuit is flexed, it may be less likely to break. I imagine you would need a curved backing plate to give it a try. Also, if you would like to have aluminum on a substrate with adhesion issues, depositing a seed layer of a metal that does stick well first will allow you to add on Al on top.

[@ThatWoodWorking]

This video interested me a lot, and am looking forward to seeing the development of it. I was thinking about your method and i think it is possible to miss a step, if you create a conductive transparent sheet which you could put into the vinyl cutter to cut the circuit out with, for example graphene based film, i am unfamiliar with how it is made, but i thought it might help. Thanks

[@PhattyMo]

Yet more good stuff! I'm wondering if you could perhaps make some lower-value capacitors by layering another sheet of plastic on there..or perhaps depositing on both sides of a single sheet? (don't know how you would make a via to connect to the other side though.) If you could control the amount of aluminum deposited,you might be able to make some lower-value resistors,maybe?

[@CookingWithCows]

It would in any case be difficult to determine if the measured temperature is constant throughout the bottle, so I don't think if printing the thermo couple onto the glass would be very reliable. It is something that must be explored in experiment :) I believe sticking a simple thermo sensitive (colour changing) material onto the glass would have the same accuracy (see those fish tank thermometers that you just stick on the glass)

[@preben01]

There is someone who made a gerber-to-cameo printer script that is super awesome to use. Then you dont need to trace it by hand. It basically sends the gerber directly to the cameo. Google "cameo pcb stencil" and select the link from the dangrous prototype forum. Its a bit of hassle to configure for first time use, but works flawlessly!

[@eddraper]

Thanks for the uptake in video production... You seem to have your fingers in a lot of pies - all of which are interesting.

[@whatever11ization]

Have you considered running the completed circuit through a laminating machine (probably a little modified) in order to bind components, i expect the glue would require a little more care as would temperatures

[@crookedninja5]

very cool... like to see you build a pill size transmitter in a choice of red or blue colors!

[@davidyeefrach4554]

Brilliant. Very creative.

[@AppliedScience]

I remember seeing your glass PCB video a while back, and being super inspired. There is just something so cool about having a clear electronic circuit. As far as I know you were the first with a glass PCB. Where do you buy your photoresist?

[@t76d777f6d]

Hi there, it would be interesting to use your methods to create a "planar magnetic" headphone driver. I've been trying to research methods for doing this but have found it difficult to find any journals or design guides for this type of work.

[@-vermin-]

Rather than conductive glue you could use z-axis conductive tape. It's very thin and flexible. It conducts from one side to the other but not laterally along the tape.

[@RhetteLawe]

Just some thoughts: You might consider chilling the metal backing plate. There's a few different ways to do this, but i think a peltier effect would probably be best, given the low pressure situation that you have going. you can get them quite cheap, and, as you clearly know what your doing with the electronics, modifying them to fit your needs would be easy. After that, overhead transparencies are usually a type of cellulose, which might not be the best. I would consider using an acrylic type of plastic. It has more thermal stability then cellulose, and actually has some good optical properties. The only thing that might make that less then desirable would be it's typical rigidity, but it can be blended with a more flexible plastics, like low density polyethylene, to avoid some of that. You do have to experiment a bit though. plastics are not like metals. they don't always blend nice.

[@BlueJay137]

Can you try sticking the electronic components on the transparency sheet first, then cover the parts in a way so that they don't short out, then run the deposition, peel the vinyl. and hopefully,this would make it so that the metal deposits right on the leads of the component no soldering is needed and it would be much more flexible.

[@benkasper4035]

Awesome, simply so cool, keep it up 100%

[@AddisonPhilips]

Again, completely awesome!! I want to be just like you when I grow up, LOL. You are doing this stuff in your shop there, like the guys at Bell Labs and Intel did over the years, on a commercial scale. Have you noticed any dimness from having deposited metal on the chamber walls? It happens. I think your work is very worthwhile. Keep uploading! - Phil

[@scianiac]

I've had good luck using Eagle's EPS vector output, I've brought those over to Inkscape for solder stencils and such.

[@RandyLott]

Unfortunately, ceramic capacitors are very brittle and can break easily. Other types are more ideal for this. Cool video!

[@nelumbonucifera7537]

I eagerly await your first flexible thin film diodes and transistors. I was quite surprised to hear you mention OLEDs. How will you source the organic materials?