[27C3] (en) File – Print – Electronics

good morning every hacker space has a wish a wish that you could make your very own PCB and wait no longer because Jeff here is going to make that happen today and i would say Jeff take the flow away and make it rock okay hi everyone my name is Jeff I going to introduce a new printing machine for printing circuit boards so what is a PCB it’s a printed circuit board it really has two functions so this is a really close zoomed in shot of a circuit board they’re usually those flat green objects you get in your in your PC whatever your motherboard is an example it’s a it’s a flat sheet of fiberglass and then on top of that there’s a patterned layer of copper which here is actually green because on top of the copper is a layer of stuff with solder mask which is a protective coating that stops solder spreading too far and on top of that are the components so the PCB gives you two things it gives you a mechanical support that holds all of your electronic components physically in place and it also provides the electrical connectivity between your components so if we cut through the circuit board you can see all of the different elements so in the middle there’s the fiberglass layer that’s the real mechanical strength of the thing on the top and the bottom of the two patterned layers of copper around the place here and there are these things called vias which are connections between the top and bottom layers okay through the board and they are a a hole that’s been drilled through the boards as a cylindrical hole and then there’s a layer of metal deposited around the inside of that hole that establishes a connection between the top and the bottom on top of all of that are your components so this might be a chip or something and then legs come out of the components sit on the copper and then they’re sold it in place and finally this green material is that solder mask are telling you about this is the protective layer that stops the from oxidizing or getting corroded by crowded in the environment and also stops older from flowing too far when it melts in the manufacturing process so why would you want to make your own PCBs here are some of the alternatives to a PCB oh yeah there are no nice ways of prototyping your electronics so a circuit board is really the the only way I mean the whole of industry has optimized all of the designs of the electronic components that you’ll want to use for mounting on a circuit board and and everything else is kind of a bit of a poor imitation so why not use an Arduino I’ve been sort of subtly bracing the Arduino in all of my marketing merchandise and I really I really have nothing against it I think it’s a great way to get into electronics and it’s been great for sort of artists and people who have no electronic experience but the thing about an Arduino and any other sort of pre-built electronic module the Arduino of course is itself a PCB with components soldered on is that it’s it’s not as flexible as you might want so at some point as you get more experience with electronics you’ll want to maybe design something super miniature that you can’t buy in an

off-the-shelf module or use a chip that’s brand new and no one has modularized it for you yet or you might want to design some analog circuit which is you know specific to a particular job and just isn’t available in an off-the-shelf preassembled piece but why would you want to make your own when you can get a PCB from a you know a quick turnaround prototyping service in about three days for about 50 euros this is yeah pretty much true across Europe and America certainly and cheaper but also a little bit slow if you outsourced to China I think there are two reasons it’s not fast enough and it’s not cheap enough if you’ve got a quick idea and you just want to try it out in a kind of sketch then you don’t really want to wait three days on the whole and also when you’re prototyping I think the real problem with prototyping on breadboards and all of those horrible things that I showed you is that yeah whilst they are still run pleasant and tend to be quite unreliable I think the real problem is that as you prototype you get no closer to your finished object although your circuit design is is converging upon your finished solution you’re still going to need to transcribe that from a sort of bread board layout or a strip would lay out or something into a finished PCB whereas if you can quickly iterate through multiple PCB revisions you’re gradually closed in on your finished product and then your final prototype will essentially be your finished piece EV so how r PCBs made industrially let’s look at that again this is a series of cross section views through the PCB as it’s developing through the manufacturing process so you start off with this laminate of the fiberglass and a thin layer of corp copper on the top and bottom next the holes are drilled through to establish the the places where components with legs are going to end up for example connectors and stuff increasingly rare these days is people are moving towards surface mount stuff but also crucially the vias which establish the connections between the top and the bottom next photo resist is deposited and this is a light-sensitive polymer then the masks which contain the image data for the top and bottom of your design a place on top of the photoresist and exposed to UV light and the areas which were exposed are chemically changed and become resilient and the areas that are hidden by pieces of the masks are not and here you can see that that patterning of the the chemically changed regions and the unchanged regions and then you put it into some magical chemical and that dissolves the unchanged regions leaving in place the hardened areas that have been exposed to the UV so at this point there is the image data in a 2d protective plastic layer basically on top of the copper and some exposed regions of copper next is placed into a series of even more magical chemicals which deposit a thin layer of slightly conductive material on all of the insides of the holes and now the whole assembly is electroplated and so that thin layer of copper is grown up to a much thicker layer and the slightly conductive bit through the holes are grown out into a properly conductive piece of bulk copper that establishes that electrical contact then a layer of tin or another more inert metal is plated on top of the the copper and the photoresist that was hardened is stripped away so you’ve now got your patterned copper image on the top of the bottom protected by tin plate and the thin layers of copper that were just there to establish the electrical connections for the electroplating process then etched away so at this point there’s the finished electrical component of the design Lee the top and the bottom layers of copper a patterned and there they’re ready for use and now the solder mask is applied so start with solder mask is applied to the whole PTV completely covering everything and there’s another masking step that happens so this solder mask is the that green protective layer I mentioned earlier it’s imaged in the same way as the copper essentially all the photoresist rather two more masks which now contain the image data for the solder mask layer a place on the board and it’s exposed to UV light again and that chemically

changes some regions of the solder mask and now a different solder mask specific developing chemical strips off the area’s you don’t want so for example the areas where you’re going to solder a components legs onto the board they obviously don’t want a protective plastic layer they need to be exposed copper there’ll be windows in your in your solder mask so yeah so that’s sort of sixteen basic steps and of course that’s the abridged version so there’s a whole series of cleaning steps probably for each of those major steps and i also just sort of conveniently introduced these masks where did they come from they themselves are the output of another photo lithographic imaging process themselves so how is it done by hackers obviously replicating all of those sixteen steps whilst possible and has been done by hackers in there in the back spaces and the basements and whatever you really want to pare that down and I think there are two fundamental steps the first one is producing that H resistant patterned layer of protective plastic let’s say or any other material was going to resist the sort of acid bath that H’s the copper and then the second step is is etching the copper itself so that’s the chemical path which eats away the exposed copper and hopefully won’t eat away the protective layer that you’ve deposited so the most stupid way you could come up with of designing of sorry creating a protective layer it’s just with a marker pen and physically draw it on your board but obviously you’re not gonna be able to produce very fine detail of the sort that you really want in a circuit board and you know if you make a single mistake you have to draw the entire thing again and yeah you end up with stuff that looks like this so it’s kind of it’s obviously a PCB layout and it put really works but it’s it’s not particularly fitty and not particularly repeatable so option two this is seems to be increasingly popular at the moment a lot of hackers are doing this toner transfer so you print out your your design on a laser printer onto shiny paper and then you rough up the surface of your copper board so there’s nice and easy to stick things too and then you place your toner out of the printer face down onto the copper and heat it with a household domestic iron and that melts the toner which is essentially black plastic onto the copper and then you soak off the paper and clean off the paper residue and stuff and you’re left with this resist so step one is accomplished here you’ve got this patterned area of plastic toner stuck to the copper and that looks like a really nice result there and then the easy step step to chuck it into a bath of etch chemical and that eats the exposed copper and you strip off the toner and you have your finish PCB option three photo lifts so a bit like the industrial process but with a few steps removed you can buy PCB material that comes coated in photoresist already and people make their own UV exposure boxes print out their artwork on maybe transparency film with a printer you know inkjet or laser or whatever and then expose it to light develop it and fetch it so another option step 4 option for movie isolation milling so I meant to draw your attention to the the number of steps in each of these processes so I mean this kind of appears on paper to be pretty much the best option there’s just one step you put the PCB in the machine and you mill it out and this is a mechanical cutter that’s eating away areas of the copper that you don’t want but unfortunately the process really sucks it’s really difficult to avoid getting kind of burrs and chewed up areas of the copper that are going to cause problems later on and also you’re never going to be able to get close to the resolutions that are the kind of chemical or optical etching process occur are going to produce so although it’s incredibly simple it’s also quite limited another option availing a couple of people do this this guy modified a a pen clotting machine to move around an optical fiber that’s coupled to a UV LED and then he uses that to expose photo resist coated PCB and then he developed an edge is it so that’s pretty much a three-step process and yeah pretty cool but really slow because probably the output power of the

LED is limited but he has to plot the tracks very slowly to sufficiently expose the photoresist to the UV another option lots of hackers and hacker spaces now own laser cutters and people are playing around with using laser cutters to etch PCBs now you can’t etch the copper directly with the kind of laser that you can afford to buy yourself or as a hackerspace even because you just don’t have enough power to erode the copper it just wicks away the heat too quickly so people paint on black spray paint and then etch that off with the laser and then they clean off the board with because it leaves a kind of ashy residue and then etch it another option inkjet technology so people take a an off-the-shelf inkjet printer usually they have to modify it to use a more water resistant ink because obviously the the etch solutions are all water-based and if you have water-based ink the etch is going to dissolve the ink as well as the copper so if we use more water resistant inks and then there’s also a step of baking the ink to get it to really sort of solidify onto the PCB itself so it’s not quite as simple as it might appear just printing an etching but it’s it’s definitely getting close so at this point I should probably just cut to the part of the demo side so my demo is not working a hundred percent and I’ll explain why a bit later on but I have here a PCB that we printed earlier which just just contains a couple of lines now you ought to see it but we’ll show you on the camera in a second so this is just to demonstrate the the etching part of the process remember there’s two steps the printing part and the etching pot so you want to come out probe and if we can switch switch to the video that would be great okay cool so I know if you can see this is this is the PCB that we’ve printed and you can just see at the top here this there’s a couple of lines one’s a little bit fuzzy ones nice and crisp and there’s a few little splashes of wax and stuff around and I’ll explain all of this later but I have to get it in at this point because it it takes about 20 minutes or so to etch yeah so these these two lines that are in wax the moment of the important bits so at the top here this one and this one hopefully they will be revealed at the end so here’s the edge solution we’ve got this this nice insulated bento box here which is going to keep the edge nice and warm the speed of the edge is pretty much just dependent on its temperature so this is about sort of 40 50 degrees and we’ll just chuck with PCB in and Rob here is going to look after the the etch process and we’ll show you what it looks like at the end so yeah it’s nice and simple we just chuck it in the solution and then occasionally sort of swill it around thank you so we switched back to the slides now if possible okay so that is the inkjet method that some people using I’m going to try and do a bit better than that and introduce my process which is based on inkjet technology but with a key change so instead of using an off-the-shelf ink or you know one of these special extra waterproofing I want to use wax so I go to print liquid molten wax and the the waxes are liquid in the printhead and then it comes out and it hits the cold copper surface and content and freezes immediately one of the main problems with the inkjet process which I forgot to mention on the previous slide is that the the ink tends to accumulate into a fairly large mass and then form big droplets and kind of draw together rather than maintaining that perfect image that you’re aiming for my hope is that by printing liquid wax its liquid hits the PCV freezes solid and it’s not going to spread it’s also obviously water-based sorry waterproof rather so it should be pretty resilient against the edge and unlike inkjet inks it doesn’t contain any solid particles which might block up the print head over time so inkjet inks are designed on this absolute knife edge they absolutely mustn’t dry out because if they dry out the solvent evaporates leaves a solid residue in the printhead and gums up your printer but also they have to dry

instantly because when they hit the peak the page then you want them to dry before they spread and blow out your image so this this sort of knife edge of design is obviously a major problem and it’s something that we can completely avoid by using wax because you know we we don’t care about it containing any die we’re not aiming to produce a pretty photograph finally you’re not reliant on a manufacturer’s ink recipe so if you go to the shop and buy you know epson super glossy ultra pro in 2000’s this week and then you print it and you make a PCB maybe next week the stuff that you buy might not necessarily be the same it might be as good for printing photographs or even better maybe but it might not have the same PCB properties because it hasn’t been designed for that so can you actually do this can you inkjet print works and the answer is yes there’s a a while ago I had this crazy project where I had to make this pair of shoes covered in LEDs in no time and there wasn’t there just wasn’t time to outsource the the production of flexible circuits to coat these coat these trainers with so I bought this machine which is called a xerox phaser printer and it inside contains an inkjet printer that prints liquid wax molten wax onto a drum and then it rolls the drum onto the page and i just took flexible PCB laminate which is very much like the laminate i introduced in the beginning of the industrial process it’s a plastic with a thin layer of copper on top I just cut it up into a full sheets loaded into the printer click print and it came out with perfect wax image on perfectly clean copper and then I actually and sold it and ended up with this PCB and it was incredibly simple so we know that one you can inkjet print works this is doable because you can buy a machine that does it and to wax is an effective etch resist so I plied some producing machine that’s kind of like the MakerBot of of PCBs I suppose to try and just simplify this process and democratize it um so here are my goals two steps print H no around no sanding of PCBs no cleaning just two steps feature sizes smaller than or point two millimeters which is the low end of industrial processes I think this project is useless if you can match some at least some of the capability of industry because then your designs are you know can be carried both ways from from the the mass-produced world into into paccar world and vice versa if you can keep the same design rules essentially I’d also like to be able to print on both sides and align them index them together but I should point out that I have no solution yet for her connections through the board those buyers that I mentioned earlier so yeah that’s one for the community if you’ve got any ideas that would be great so let’s choose a printhead to use that’s the really the heart of the the inkjet printer the the device which sucks in ink from a tank and squirts it out onto onto a surface in a really accurate way so two types of inkjet technology one is the piezoelectric type so here there’s a there’s your ink tank your ink cartridge and there’s a system of plumbing basically that carries the ink up to the nozzle on the top of the nozzle is a thin diaphragm of stainless steel or silicon silicon rather and on top of that the piezoelectric element and a Bolton is applied to the pizzo and it bends up would switch and it does this slowly to draw in ink from the reservoir and then it sharply accelerates downwards with an opposite voltage pulse and that expels a droplet of ink and the other method is the thermal method so very similar plumbing but instead of that piezoelectric membrane there’s a heater and the heater quickly dumps energy into the ink it vaporizes and under the pressure its own vapor it expels a droplet I want to use the Pizzo method because the thermal method has very short lifetime the these power resistors essentially have such high currents pump through them that they age very rapidly and this is why everyone in the consumer market except epson used this technology and they sell their printhead as part of the cartridge whereas epson sell a piezoelectric printhead which is part of the machine because it’s got that longer lifetime so I’d like to use a piezoelectric methods because longer lifetime and also you get a little bit more control over the ink formulation you don’t have to have something that’s

easily vaporizes and all of that so yeah I use pretty much the crappiest chiefest printer you can buy this is this is really the bottom of the range this prints for peak elisa droplets which are 19 microns in diameter that’s about this diameter as well as the diameter of your hair and on the nozzle plate there are 30 30 and 30 nozzles of cyan magenta and yellow and then a 90 nozzle strip of black and obviously by having more nozzles you can print more at once and sort of parallel eyes the process and speed it up and you get all of that for 40 euros which is absolutely insane so if you strip our part all of its apprentice you’re not interested in you get to this print head so here are the four needles that plug into the four cartridges cyan magenta yellow and black and on the bottom the bit we’re really interested in this nozzle plate the the Assembly of microfluidics and piezoelectric gubbins and Wizardry that actually does the printing if you take that out you can see what it looks like there’s a nozzle place itself that’s got two rows of nozzles that are barely visible and certainly not visible in this photo and then on this flexible circuit board in the middle that little black blob is the control chip and that is what we’re interested in talking to and i’ll i’ll talk about the reverse engineering of that in the in the coming slides flip it over and you can see the four ports which allowing into the printhead so that’s I magenta yellow and the one on its own here is the black one and they connect to the system of vessels that leads up to the ups the ink cartridges this PCB on the left is purely a connector it’s completely done there’s no electronics on there the only electronics we’re interested in again is that black blob in the middle I meant to delete this slide this is an inside view under a microscope and the nozzles are hilariously small you can see that row of nozzles there it is kind of a bit of a crappy image so now we need to obviously heat up this whole printhead and we can’t heat up the whole plastic assembly of my tiny plastic tubes because plastic such a good insulator that wax is going to freeze in the tubes so I started machining this sort of printhead reservoir block to replace all of that plastic gubbins so there’s the four holes for the four colors of ink although probably I’m only going to use one ever so the ink goes into the top and then it comes out with awesome in these four smaller holes here around the edges of this piece which coupled onto the onto the pinhead itself to supply the heat there’s a power transistor the top right which is essentially just shorted out and on the other side is a temperature sensor that measures the temperature so we can control the temperature by delivering current to the to the power transistor and measuring the temperature rise on the heater and yeah that works so now we’re looking at the bottom of the printhead and you can see those that heater the transistor and the temperature sensor and there’s a droplet of liquid wax dribbling out so that part is sorted at least the easy bit so we’ve got this heated block we’ve got to couple that onto epson’s printhead and you’ve got this aluminium block with four holes and there’s printhead with four holes and you got to make them together and this is not easy I came up with this really elegant solution and I was really proud of myself i took double sided sticky tape and then laser cut it into these little precision gaskets and then stuck these gaskets on the back of the printheads that they just expose the four holes and then you bond it onto the onto the printhead but it it really sucks it’s really leaky and not very much of the wax actually ends up where you want it in the printhead so then I thought I’d do what epson do they deposit a liquid gasket of silicone rubber and then that seals and cures and forms a watertight layer so i made this stencil on the laser cutter this is a just a thin plastic membrane and you can see there’s a big blob of silicone at the top there squeeze you the silicone over the stencil peel off the sensor then you’ve got these two precision islands of shaped silicone liquid and then glue the printhead on top whoa what’s going on there alder spoiler oh my goodness hang on yeah sorry about that yep yeah gaskets yeah yeah yeah but this is still sucked because the the the silicone I used was

an air curing silicone and once you’ve assembled this contraption there’s no air left and so the silicone never really set and it could never withstand the pressures that I needed to prime the nozzles the first time around so third approach laser-cut rubber gasket made out of bicycle inner tube and then mechanically clamped the print head down onto the onto the rubber gasket and that still really sucks but it just about kind of works it’s it incredibly leaky and I think that’s the the major reason why my demo today is incredibly sketchy so yeah if anyone’s got any good ideas I think really the solution is the silicone but with two parts in the cone or air curing silicone anyway we sought out the plumbing partially how we’re going to do the electrical interface to actually control this thing now this is obviously reverse engineering task because epson aren’t just going to tell you how their couldn’t heads work so the first bit of any reverse engineering task is to read around to see how much you can get away with not doing yourself and there’s two previous reverse engineering attempts that I found of epson technology one a guy called volcan sahin who reverse engineered a much older epson head a few years back and used it to print a particular type of ink for PCB contraction actually and he’s got a few details on his website but they’re a bit vague and he’s a bit shady shady about it there’s another project really cool called possum which is a an academic project somewhere a bunch of researchers reverse-engineered an epson printhead and use it to print DNA arrays so the next thing patents patients are an excellent source of information it turns out don’t read any of the text it’s mostly this word and is completely impenetrable but some of the diagrams are quite good and this is actually a pretty pretty detailed description of the electrical interface and it’s just in the epson patient you can just download it and read it and you can see the sort of key elements of the interface here there’s a this analog magic voltage waveform that physically actuates the Pizzo membranes and then there’s this digital interface of the bottom that clocks in the data so this is a system overview of the of the printhead there’s that magic square wave voltage pulse there that actuates the peso’s and then that is connected to the various nozzles with this multiplexer chip that was the little black blob in the flexible PCB that I showed you earlier and you shift in the beta on this digital bus here and that basically sets some switches and then depending on which which which is the press when you give it that voltage spike some of the nozzles fire and your image data comes out okay I use this thing to start reverse engineering the the the digital interface this is a this is the open bench logic sniffer which you can buy for about 40 euros or something and I interposed that between the printer with its default epson motherboard and everything and the printhead itself so I’ve spliced it in the middle here and i’m watching the data go by between the printer and the printhead and there’s a decent amount of information in the patent about the structure of that data but i had no idea which physical pins on the device corresponding to which signal and so by attaching the logic analyzer and looking at the shape of the traces and reading the patent i’m able to work that out so then I built my own board here this is I didn’t print this myself obviously there’s a bit of a chicken and egg problem so I outsource this and this contains the digital to analog converter for producing that magic trapezium shape wave pulse and also this digital bus which connects the print head and you can see the print head off at the top there there’s also this development board off at the left which is doing the PID temperature control of the of the nozzles and it of the printhead block rather and it’s keeping that a flat 70 degrees give or take so here’s the here’s the magic waveform coming out of my device I basically copied all of the timing of the off the epson hardware by just looking at it on an oscilloscope and these these are quite rapid so this rise this sharp rise here the one that actually does the sort of physical expulsion of ink is about a microsecond or probably much less a microsecond virtually the rise time so it it’s quite a high-speed situation and it works which is pretty cool so here we’re looking at the bottom of the nozzle plate and these wisps are streams of wax droplets and I’m just cycling through one nozzle after another and so there this is this is a cloud of each each pulse as a cloud of a thousand i think

wax droplets coming out of the printhead so yeah so here’s the complete assembly in the thing i brought with me today you can see the the mechanical assembly is scavenged from an even older epson printer head at epson printer honor that’s a bit easier to play with business called Stefan motors and things rather than surveys there’s my couple of gentleman balls mounted on here and you can just see the print head up at the top right on the mounted on the carriage and that goes back and forth and does the printing so results this is a lego brick I didn’t print this you’ve probably seen these before but the little blob on the top is about three millimeters in diameter and we’re looking at this down a microscope and if we just switch the focus slightly you can see that on the copper plate below there’s some very fine tracks here so these are little tiny stacks of wax droplets so it’s by a high resolution also you could do 3d you can just stack up drop sits on top of each other basically so so these are stacks of about a thousand droplets each and this is one of these stacks really close up in this insane focus stacked image so this is a composite of lots of different images taken at different focal lengths if you can see that it’s like an icicle it’s obviously one complete piece of wax and you can see it’s got this kind of blob blob Ian look to it but those blobs aren’t individual droplets they’re actually sort of just a kind of freezing zone I think in these hairs coming out the side here are kind of on a closer scale so the size of the droplet so they really are tiny and there’s potential for some pretty high res 3d printing stuff but we’re not interested in 3d for now we want to do 2d circuits so I started trying to print a large patch of copper of wax rather and then etched into copper so here’s a here’s a patch this essentially a load of lines printed right next to each other really really dense and I etched them and I was really surprised they actually came out almost as individual lines I was hoping that they’d just agglomerate into a single mass but the resolution was sufficiently high them they almost came out on their own so I think these lines are about more point one they probably less about more point naught five millimeters wide yeah this this patch on the top right of this PCV is what I was aiming for so I reduced the spacing between the lines and printed thicker wax and stuff so i got this solid patch which i was really happy with because it proves that you can actually make a thick nice extra system player this you know going to be able to print ground planes and stuff on your PCB there’s a single pin holes defect in the middle that’s actually massive but other than that it’s a perfectly solid plane so raster images that’s what we really want right just about see this raster image emerging from the Machine one line at the time i’m only using a single nozzle at the moment because of a variety of problems so it’s really slow but it certainly works there you go there’s the finished image and h it in the bath and you end up with a pattern board now this is this is a little bit crappy there’s lots and lots of pinhole defects in here and so it’s really not very good as a circuit board so I change the settings a bit and produce the thicker layer and stuff and this is actually pretty pretty perfect so very few pinhole defects in this this is the second attempt at a raster image so I started to try and push the push the resolution see what kind of features i can create and yeah they’re getting pretty dense so what’s next better priming if the wax nozzles is what I want because at the moment nozzles only print for a short while and then end up failing and I think that’s not sorry catastrophic be failing forever but they accumulate a layer of wax on the printhead itself and that prevents droplets from leaving and they essentially just kind of stack up in this in this growing blob of wax and this is there’s probably about a million factors that go into this but that’s the key problem I need to solve next and I think the real problem that is causing insufficient priming is this crappy seal between the the print head and the reservoir block so yeah suggestions on the postcard the there’s still some still some craziness in the in the electrical in the digital interface with the printhead that I haven’t quite got nailed yet and there are some nozzles

which furiously fire when I’m not expecting them to and it’s also in the printhead interfaces there’s a kind of greyscale feature where each individual pixel is composed of several sub pixels and I don’t really understand how that works and that’s the cause of several strange interfacing but hopefully I got that nailed I am reading some more patients and digging out the logic analyzer again now even even epson have problems with wax and rather ink accumulating on the nozzle plate so they they actually have a squeegee blade mounted in the printer so your your nozzle scans up and down for a bit and it prints your image and then it stops moves over the squeegee squeegees itself moves over this thing called the spittoon which i think is a hilarious name and it purges wax of ink rather into the spittoon for a while to clear out one of the nozzles there’s squeegees itself some more and then it goes back to printing you might have noticed that you also DD PPG around our we re so I need to do that for the memory stage so yeah 3d would be something I really like to do so this is a fin this obviously this is a commercial 3d inkjet based wax printing machine used in the jewelry industry for making wax positive molds that are then cast into into physical metal objects and yeah resolution is completely nuts and actually the droplet size of this printer is much much larger than the droplet size of my printer which well it’s kind of actually a problem because i don’t want super high resolution droplets I just wanted to work at the moment so it would be much easier if these nozzles are a little bit bigger in a little bit friendlier but anyway even with bigger nozzle you can achieve completely insane 3d resolutions so yeah I’m going to show you around the printer now I’m fortunate how I’m going to printing because of this crazy plumbing situation but hopefully we’ve got some lines on a PCB rob’s just taking it out of the etch solution now and I guess if anyone has any questions you can start sort of thinking about them and shipping them in so can we switch to the video please the camera yeah hua gei cool it’s okay so this has come out of my focus on it will shuttle it on the table I try not to jiggle too much uh-huh uh-huh yeah so there’s two lines there which is pretty cool try and see if I can get the light on them there yeah there we go lines excellent thank you rob for your catching help now i’m going to show you around the printer as well just a little bit so um thank you crap out of the way so yeah so we’ve got the the printhead here and you can move that up and down only on the axis that Stefan motor controlled and if I take this piece out this is the the printhead assembly so on the bottom here you can see this this is the nozzle plate and if it’s I try out to jiggle too much you might even see mirror sorry yeah there you can see two rows of nozzles there and just over here you can see that’s a nice accumulated wax KU which is which is the problem of the day so you can see the two rows in other words that’s that’s a tee blaq nozzles in a light now sorry 90 black nozzles in a line and 87 odd color nozzles and then on the other side you can see the ports which also reservoirs that stole the wax and then there’s the heater board there yeah so that’s pretty much it and then there’s just the control electronics down here and the heater electronics there cool does anyone got any questions if you have any questions could you please come up to the microphone and the only asked question in the microphone please thank you go ahead turn it’s working no yes hello my name is axel Ruth thank you for a nice presentation is there a secret formula for liquid icing the wax because I think wax is normally solid like a nick xerox printers there are solid blocks do you use solid blocks yeah yeah just solid weiss and then melted just melted yeah simple as that i’m actually using wax from a craft shop that was sold for making candles which a terrible mistake i should be using some wax supplied by you know sigma-aldrich or something that’s got an exactly known chemical formula rather than just like super glossy candle wax but yeah literally

just a block of real wax without there any special additives I have one more question sorry wax is like melts in heat that’s why this what you use but you also most etchings things there are to it with a heated to solution like a 50 degrees yeah box over there doesn’t the wax melt in your cheeks solution that’s a good question yes the answer is yes if you if you turn up the solution of the temperature of the solution to hive and the wax melts the this wax melts at about 65 degrees Celsius and it’s being printed at 70 degrees celsius and the etch solution works well it says it works best at 50 low that’s probably just a safety measure and I suspect it works better and better and better the hotter and hotter and hotter you make it but so yes so I’m etching at 50 degree C which is below the melting point of the wax thank you cheers we’ve some we’ve done some some ceiling with a two-component glue so with google plus and I’m if you’d put in on a very very thin inlay of that take care that you don’t don’t plug up the holes maybe thats that’s worth a try cause you only have one try feel put them together you never get them apart again yeah or perhaps we could you should think about that idea definitely thank you i think a two part is that an epoxy or yeah yeah so your friend epoxy and hard now that you mix together and put it on there yeah great thank you I think definitely a two-part thing is going to work out on the air hearing stuff yeah one way you can look is motor or a ceiling stuff like there’s a black silicone which you can use to see the motors together and there they have stuff like that maybe look in that shops maybe you find something you like okay right yeah that’s a very similar problem right assembling cylinder heads onto engines yeah yeah did you did you look into any other printing media to achieve the same result possibly without being without having to heat it first such as well my water water proof marker in or stuff like that okay so did I look at alternative inks basically no I’m really excited about printing wax because you it there’s nothing to dry out and solidify and gum up the print heads pretty much so yeah I’m kind of scared of because my printer lacks any kind of self-cleaning mechanisms and stuff I’d really prefer to use an ink that is liquid and then it’s a solid you know nothing that can dry out and clog up the print heads and the other advantages of I mean obviously solvent based inks would probably be better in this regard but water-based inks are no use a tool because the water they said she’s gonna even so yeah at the moment no but I was I just thought yesterday actually might be fun to try and print immersion tinning solution so this is a chemical which you place on copper and it deposits tin onto the onto the copper and tin is an excellent etch resist but that is again a water-based solution and so if the water dries out then the the tin chemical is going to crystallize into the into the printhead so yeah I’d rather stick with wax if I Katniss is the answer okay to put two final questions okay so how long after all the reverse engineering how long does a reverse engineering take you so it’s not quite as easy a her to see so how long did you take for the whole project okay yeah good question it’s I supposed to be fair I haven’t done a huge amount of reverse engineering because it’s documented so well in the patents I pretty much wrote some code based on on what I’d seen in the patient’s use the logic analyzer to work out which signals were which and it more or less work so I quite rapidly got to that video I showed you where the nozzles were being turned on and off so yeah I didn’t have to work nearly as hard as I was expecting it I don’t know only a few days a couple of days so how do you solve the problems with two-sided printing so okay so those problems on the television stuff yeah good question lining up two layers is obviously tricky but I can imagine you know maybe drilling some holes through and then you’ve got an index and then either lining that up with some mechanical fixture or maybe one day kind of gantry

with with computer vision would be absolutely awesome but uh yeah I haven’t really tried I’ve been mostly concentrating on the sort of fundamentals of getting a wax image onto a onto a copper surface so far it doesn’t seem to be too difficult the real the real challenge is electrically connecting the two layers us that’s the killer feature yeah and I don’t have a solution for that thank you thank you will you’ll make your findings available for all the heck of spaces that are around yes absolutely i actually forgot I was gonna stick on this slide my my website warranty void if removed com I have one more thing to add I gotta take this down to the hack center now and just hang out there so if any wants to come and play with it or whatever i’ll be down there thank you thank you thank you so much