How to make precise sheet metal parts (photochemical machining)

today on applied science I'd like to show you how to make these etched metal Parts sometimes you need a piece of sheet metal that has a feature cut into it that is too fine for a water jet or a laser cutter because it would distort the part or sometimes you just have too many holes to cut like in this mesh pattern here so you need a way of making this photochemically basically make a photo mask and then etch the part with an acid so recently I had a project where I needed a bunch of these and I you know looked around and found about five or six vendors photofab is one of them and was surprised at how expensive it was and how long the lead times were it ends up being about three hundred dollars a setup fee and then around six hundred dollars minimum to make you know 10 or 20 parts that are like this size and complexity and so the bummer is that you know if you have five or six designs you want to iterate on and you have two or three iterations each when you add up all the weeks of waiting in the back and forth with the sales reps and the thousand dollars a pop it ends up being a pretty big investment so instead what I did was figure out how to do the whole process in my own shop here partially just because I like tinkering with stuff but I also was curious to see how difficult it might be to start up a side business like this so I ended up building my own etching machine laminated my own photo resist onto the metal and went through the whole process and made these special parts for another person for a secret project and still came out ahead spent less money than I would have with one of the big vendors let me spend just another minute or two talking about the parts themselves and the business and then we'll jump into the production process and I'll show you all the secrets that I learned when I had to figure all this out so these fine sheet metal parts are kind of everywhere these are battery terminals out of this low-cost clock here imagine like the wipers in a potentiometer or in a micro switch if you need like little spring fingers grilles for microphones or speakers in electronics battery doors this little drone here has a sheet metal part and of course industrially all of these are made by stamping you just cut sort of the shape you want out of a hard piece of Steel and just push it through the sheet to Stamp Out the part you want but that doesn't work if you're in low production volumes or prototyping if you're at that phase you kind of need a process that doesn't involve cutting this piece of tool steel to make the parts because that's very expensive and time consuming so even as bad as photofab is still a better deal than paying for a die to be made and then to have to throw that out to iterate on your design however there are some things that you just can't make even with a die like for example this mesh you know custom mesh if you need a if your project requires a custom mesh size or a custom opacity here the only way to make this is pretty much photochemical etching because you can't stamp this out and you can't really laser cut it and you definitely can't water jet cut it it's pretty much going to have to be photochemically etched another aspect of this is that you may want your parts to be free from the panel so this thing has a smooth edge all around but traditionally the way these things are made is there are tabs holding the part into the panel and someone either you or a technician at the company has to cut the individual Parts out of the sheet and then get rid of those tabs and so one of the things I came up with this idea of doing the whole process in a mesh bag so what we do is put the metal in here and develop it and etch it all within this mesh bag in the etching process the parts will actually just fall out of the panel right because there's nothing connecting it there's no tabs and so what you're ending up with is the panel and all the parts free from the panel in the bag and then we strip the resist off while it's still in the bag and rinse it and then take the whole mesh bag and throw it into a shipping box and send it out to the customer and you never even have to handle the parts or even look at them or count them or anything and so part of the idea of getting the price down for this whole process is to automate and eliminate as much a human work as possible another example of a real product that benefits from this process are these flip pins these are dipped pins that you can solder onto your project to make it work like a dip chip package so instead of using 10th inch header pins which are actually not the right size they can damage breadboards these are real dip sized pins that you can add to your project and normally these would be stamped metal in high production but for you know Niche use like this they are made photochemically another really important aspect is to make the artwork generation automated So currently if you want to get a design done with one of these existing photochemical suppliers you have to email them your artwork and there's a bit of back and forth with the layout engineers and it's kind of a slow and sort of unpredictable process but there's plenty of businesses that have automatic artwork upload so you know there's three and five access CNC Machining that's fully automatic or you can just upload your Cad and see the part in real time on the website obviously multi-layer circuit boards laser cutting water jet cutting there's no reason that you can't just upload your parts to a website see a representation of what you're going to get and then know what the price and timeline is going to be so and that would actually save time for both the business and the customer because it's pretty well understood what you're going to end up with so automating that part and then automating the parts finishing part are kind of my proposals for how you would make this process less expensive so anyway let's take let's take a look at the process now the first step is to get some sheet metal and clean it very thoroughly and I'm just using McMaster for the source of sheet metal and even at McMaster prices this ends up not being a significant parts cost in this whole process so most of these you know 12 by 8 inch sheets of sheet metal in the thicknesses that we're talking about of you know one foul up to maybe 10 or 20 thou or you know quarter millimeter or half millimeter at most still end up being ten dollars five or ten dollars and I was imagining that my standard size for this you know potential side business that I was talking about was like a postcard so you'd get four postcards out of this you know ten dollars it ends up being an almost nothing materials cost and the way that you check if this is clean or not is the water break test and so we basically get on here and just spray some water on and you can see that it is beating up quite well which indicates that the surface is completely contaminated with oil and in our next step we have to put photoresist on here and this thing must be extremely clean so the way that I was cleaning them was to put them in this tub of water and alcoholics but you could use even sodium hydroxide or something else and I'm using this immersion cooker this is a sous-vide cooker to clean the metal so it'll stay in there for a while and then to wait to know that it's clean or not is to take it out and rinse it thoroughly and do the water break test again so let's give it a few minutes and then come back okay it's been a few minutes and so we're going to take this out and rinse it with this deionized water and I used to have a dunk tank for this but I don't have it set up so I'm just going to spray it with this bottle okay so now we'll test we'll do the water break test again and you can see that the water just completely sheets off there's absolutely no tendency to form droplets like that like we had before and you need to be really harsh with the water break test if it seems like there's any sort of a hint that the water is not grabbing the surface fully then the part isn't really clean so the next problem I ran into is drying it you can't just sort of leave it out in the air for a long time because the air is full of micro droplets of oil and that will contaminate the surface and you can't take your compressed air hose and blow it off either because that's completely full of oil and so it sounds primitive but the best way I've found to deal with this is to just use shop towels and just scrub it dry sounds strange but this is the best method I think really the best one would be having a really clean air source but anyway I'm going to clean this off and then we'll head over to the photoresist laminator now this is just a cheap laminator that I bought off of eBay but I found out that there's a critical difference between these cheap laminators and the ones that are meant for applying Dry Film photo resist to things like metal and circuit boards as it turns out these rolls of Dry Film photo resist are actually three layers it's a soft protective layer then the photoresist itself and then a hard protective layer and what we want to happen is the laminating machine will remove the soft protective layer exposing the photoresist so that when we put the piece of metal in Here It Coats both sides and it's sort of peeling off that soft protective layer as it's unrolling the stuff and this machine was not set up to do that these normal Chief laminators don't have these take up spools so I added on these extra things here and it doesn't work that well but it was good enough to get this project done but the idea is that I added these these spring loaded rollers that act as take up reels so that when the thing is running it's sort of peeling off the soft protective layer and exposing the photo resist then the hard protective layer stays on throughout the exposure process and you only peel it off right before developing so um let me get our piece of metal here and start the thing running I haven't used this in a while so I wouldn't be totally shocked if this doesn't work kind of hard to see how this thing is threaded but in fact I don't even remember exactly it is pretty complicated but basically these take-up spools are just spring-loaded and held against these and the only part that's driven are the two heated rollers and I have it set to about 110 degrees C and it's running really slow it's two notches out of ten or whatever on its arbitrary speed scale we'll stop that and then I'll just Slice It Off there we have it not bad at all um so what we have is the hard protective layer then the photoresist then the metal then on the other side more photoresist and another hard protective layer the next step is to print out the artwork and I found out that Laser Printers don't make good photo masks it's just the way the toner works you'll always end up with pinholes and not enough density what you're really looking for is the absolute highest contrast ratio between the dark areas and the clear areas and Laser Printers as near as I can tell just never get heavy enough so there's this interesting parallel industry for making photo tools for screen printing there's a whole website set up that sells special ink for inkjet printers and special transparencies that are made to soak up tons of ink these transparencies actually have an Emulsion on one side so when a drop of ink hits it it grabs onto the ink and holds it there and doesn't make a mess they're single sided if you put this in upside down it makes a horrible mess because the ink doesn't soak into the transparency and I've dialed the settings up so that the printer is dispensing the maximum amount of ink and the contrast ratio is unbelievable it's very very dark so what I do is print two photo masks we need two because we're etching the front and the back and print them side by side so if there's a tiny amount of stretching they will be the same in both the front and the back photo mask so for perfect dimensional accuracy it's not going to be perfect in that Dimension or really any Dimension but especially the dimension along the rollers this printer is specifically made to use a large tank of ink so you can pour in anything you want and it's a monochrome inkjet printer kind of unusual setup but another Quirk of it is that it will not print on transparencies I couldn't believe it but if you put a transparency in there it will not print and I contacted support and they said yeah it's just not supported media and the reason it's not supported is because the printer uses like a little Optical sensor to sense the edge of the paper so it's clear and it never sees the edge of the paper and it just spits it through and throws an error luckily if you put a plain piece of white paper underneath the transparency it will pull both through at the same time and I found this to be a very consistent way to get it to work but very interesting Quirk there okay so now it is time to align the top and bottom mask and what we want to end up with is something like this that has double-sided tape on both sides of the artwork and what we'll do is put our piece of metal in between right like this and I originally started off just doing it like a book just putting the double-sided tape on one side and opening it up but I found that that would actually lead to very slight registration errors because just like a book imagine tilting a book's spine it can cause the pages to be slightly misaligned because we're going to put you know a quarter millimeter or a half millimeter of thickness inside here and it will actually not be perfect and so you want this thing to be as consistent and symmetric and perfect as possible so double stick on both sides and I'm doing the alignment just by looking at it I wish I had a better way but just looking at it closely is how to do it foreign exposure station and we've got our artwork in our panel here and normally we'd be doing this under the orange safe flight but like you say it's kind of daylight right now so we're going to put the uh the metal in between the two pieces of artwork there and remember that it is emulsion side closest to the panel and then of course we need to really pinch this thing down if there's any Gap in there the shadowing won't be very exact so I built this exposure rig with two of these quote 150 watt 405 nanometer floodlights and the reason I put the lights far away from the artwork that's in the middle there is to make the shadowing better so if the lights are really close to the artwork in the panel the shadow can sort of sneak under the dark areas in the photo mask and ruin your resolution with this with it being far away the shadows will be sharper just like how you know the shadow of an airplane is quite sharp on the ground just because the sun is so far away I ended up using the exposure of 2 minutes and 40 seconds and I measured the intensity of the 405 nanometer light at the site of the exposure to be about half a milliwatt per square centimeter and looking in the data sheet for most types of photoresist like this the dose should be about 60 millijoules per square centimeter which should be an exposure of 100 seconds obviously a few of not idealities are going on here because I noticed I needed quite a bit more exposure but still there's the numbers so you can actually see the pattern in the photo resist even before we've done anything even before we've developed it and what we can do is don't forget to remove the hard plastic protection layers on here so we'll peel this off and it's this thin sort of plastic layer on both sides that we need to get rid of because so now now the photoresist is actually exposed and the next thing we'll do is put this into a mesh bag and then load it into the developer the developer itself is just tap water and sodium carbonate and inside this plastic tub I've got two spinny sprinklers and a pump outside here with a timer and a heater to bring the temperature up to about 28 degrees C so I've loaded the panel to be developed into this little carrier and this was kind of the first idea of that I had for this modular system where you'd be able to slide this thing into each tank in the process and so we'll put this in here and start it up 70 seconds okay and you can already see that it a pretty good development I normally had a rinse tank but I'm just going to take this outside and quickly rinse it with the hose okay so pretty good near the top here it looks like it wasn't fully developed or exposed because this panel is quite a bit bigger than the ones I usually use but down in here it looks like we have pretty good registration and resolution and everything and so we're ready to etch and so now we're going to load it up into this other little holder that I came up with the holder itself has these rollers here and magnets that are encased in plastic and the magnets will interface with magnets on the outside of this etching tank and the idea is that it's kind of a through the wall movement mechanism so that there's nothing that goes through the wall of the tank the magnets will keep it in registration and the other cool thing is that it could be loaded and unloaded pretty easily just by putting it in there and so if this system were modular I would build one of these for the developer and one for the rinser and everything else and you can just reach in and just pull the thing out pulling it off the magnets and put it in the next tank the actuator itself is a linear belt drive actuator that I got from open builds it runs on this aluminum Extrusion and got a stepper motor driver up there with a Teensy and it even has like a little display and a UI and the idea there is that it also has a heater and controls the Heat and the time and everything else a little bit more detail on how this etcher is built the main pump is an Aro pneumatically powered pump the trick is you need a really high pressure and flow rate of etchant to do a good job etching the difference between just submerging the piece of metal into a you know even a stirred Beaker of ferric chloride is nothing it's about 10 times slower than spray etching like this the force of the etch is actually what blows the etch products away and allows it to proceed more quickly so you really need a very powerful nozzle and as it turns out the rotary pumps that I was using for the developer station for example just don't produce enough power you really need something like these these pneumatic pumps to get enough horsepower in there basically and it's uh controlled by a air piloted valve over here so that little tiny valve operates a bigger one and that just sends the full Airstream in and you can see the output here is connected to this tube which is capped at the top it's sort of an air hammer and that uh cushions the blows from the pump since this is an air powered pump it has a piston in there and it Cycles back and forth so the output is very pulsatile I had a major problem here today that's not the first time finding Plastics that are compatible with chloride is more difficult than it seems nylon is completely out and so is Delrin but even other things like polypropylene seem like they're kind of okay and then they turn out not to be and so what ends up happening is the fittings make a little leak luckily a little leak the nightmare scenario here is you've got this super powerful pump you know making 30 or 40 PSI at gallons per minute flow rates of hot ferric chloride and if one of those things bursts I mean the the mess is just going to be Beyond spectacular okay let's press the start button foreign okay and then here's our bag of parts and as you can see they fell out of the panel just like we were talking about and the last step is to put this into a device that will get rid of the photoresist and the way you do that is you dunk it in sodium sodium hydroxide anything really strongly basic potassium hydroxide too and what will happen is it doesn't dissolve the photoresist it comes off in big chunks and so I was experimenting with this wave putting this bag into sort of like a flow cell that would flow the stuff upwards and hold the top of the bag open so that these big chips of photoresist would come out of the bag unfortunately I took that apart and don't have it anymore so today we're just going to leave the photo resist on and see what we've got here yeah as it turns out it's not my best work ever but they did turn out kind of okay I had a major problem with the etching machine and had to pause it Midway which is never good and like I say I haven't used any of this equipment in a while just decided to make a video but anyway I thought you hope I hope you found that interesting and if you have any questions about the process feel free to put it in the comments and I'll I'll do my best to answer all right see you next time bye

Channel: Applied Science

Views: 349,226

Keywords:

Id: bR9EN3kUlfg

Length: 21min 56sec (1316 seconds)

Published: Thu Nov 03 2022