Mini Mint Tin Automaton Pendant/Brooch

by pixelinabitmap in Craft > Jewelry

4607 Views, 109 Favorites, 0 Comments

Mini Mint Tin Automaton Pendant/Brooch

you_make_my_heart_flutter2_cover.png
you_make_my_heart_flutter2.gif

This is my second attempt to build an automaton. It started out as an attempt to compensate myself for having lost the automaton from attempt #1. My goal was to re-use the theme and general design of the first, but make it "way more awesome" (TM). Really, what I was envisioning was 1) a more complex movement of the heart and 2) actually managing to make a collage that used sweet and whimsical scrapbooking papers/stickers to create a convincing backdrop of viscera/entrails to showcase the beating (anatomically-convincing-if-not-correct) heart. I think I may have slightly overshot the mark.

I'm not sure where the fixation on making entrails out of of floral prints originated, but it's been with me since the first attempt, and it was my one big disappointment with that automaton -- I mangled the collage, ripped it out, and ended up making an attractively messy abstract background with what was left. Ironically, this time around, the entrail collage is the only thing that worked on the first attempt -- I was immediately satisfied with how it turned out, and simply tried not to damage it too much as everything around it was repeatedly scrapped and rebuilt.

Because every other element went through its own evolution, and because most of the scrapped attempts weren't scrapped for any potentially educational reasons, I'm going to limit this instructable to explaining only the things that made it to the final incarnation. However, some of the photos will include some early versions of elements which were later replaced. Also, some of the photos may be out of order or contain elements that aren't mentioned or seen again, because all of this prototype churn made for a very meandering path to project completion, and if this instructable was going to make any sense, I had to put it in some kind of order, which might not entirely fit with the actual chronology. Rest assured, however, that I will do my best to make it possible to create a similar automaton from start to finish by following the presented steps in order.

Also, because this is a fairly complex multimedia project, as far as the tools and materials are concerned, every step will have its own suggestions/options.

Step 1: Container Structure

box_003546.jpg
box_010629.jpg
box_011654.jpg
box_135141.jpg
box_135154.jpg
box_135216.jpg
box_bail_191911.jpg
box_bail_192715.jpg
back_pin_193410.jpg
back_pin_193832.jpg
back_pin_195249.jpg
back_pin_195320.jpg

This automaton is housed in an Altoids "smalls" mint tin -- I sanded off the paint before taking any photos, but you can sort of see some remnants on the first few. If you aren't fond of mints, these tins can be purchased, blank and unused, from a variety of vendors (including on ebay and amazon). The dimensions are listed as 60 x 50 x 15 mm/ 2.36" x 1.96" x 0.59". I went with this size because it's readily available, I find the little boxes really cute, and because I wanted this automaton to be at least potentially wearable as jewelry (without being mistaken for a purse, which a "regular" size Altoids tin could be). However, for anyone who isn't hung up on the jewelry idea, the "regular" size might work better: there's more room to maneuver, and, since the intent of the design is a declaration of feelings, it's meant to be given as a gift, and a slightly larger 95 x 58 x 21 mm/3.7" x 2.3" x 0.8" tin might look more substantial (and is less likely to be mistaken for a ring box). So, for this step, you'll need:

Tools/Materials

⦁ a hinged tin of your preferred size
⦁ relatively thin soft wire -- preferably copper or aluminum (picture mechanical pencil leads, you want some wire about that thick -- a little thicker could work, but I wouldn't go thinner than that, as it would break too easily)
⦁ (optionally, if you want it to be wearable as a pendant) a somewhat thicker soft wire to either make or secure a bail (picture the lead of a standard wooden pencil -- you want it to be at least that thick, but not too much thicker). I used a short (scrap) piece of thin aluminum tubing (purchased in a small bag of random aluminum tubing cutoffs on amazon). Other options for a pendant bail include wire rings or actual pendant bails often sold in craft store bead sections.
⦁ (also optionally, if you want it to be wearable as a brooch/pin) a pin finding -- these are pretty common, and commonly found in stores in/near the beading section. I wanted mine to be removable, so, rather than attaching it directly to the tin, I glued a couple of small rare-earth magnets (another optional component) to the finding with superglue, and glued a corresponding wire "template" to the back of the tin for the magnets to fit into, so they don't slide all around the back of the tin when being worn.
⦁ wire cutters
⦁ pliers -- I would suggest at least two pairs: ideally something with thin flat(non-serrated) jaws, but common needlenose ones will work in a pinch.
⦁ a handheld rotary tool (such as a Dremel; although I use an off-brand pistol-grip cordless one, alternated with a variable speed mini-engraver, meant primarily for manicures) with a grinding bit -- I like the aluminum oxide (usually orange, pink, or white) bullet shaped ones; ceramic cutting disk bits would also be useful, but are not strictly necessary. If you don't have one, don't despair, you can do without by using...
⦁ some thin sharp nails, a hammer, and a plastic bottle cap -- these are really great for punching holes in thin sheet metal, whether or not you also use the rotary tool/grinding bit to make the target spot thinner.
⦁ thin metal rasps -- these are also known as needle files -- they are extremely useful for widening holes and removing burrs; if you have a rotary tool with a grinding bit, the rasps are not strictly necessary, although still quite useful.
⦁ a few glass "pony" beads -- they are the slightly larger variety of seed beads (about 3mm OD) and are pretty easy to find in craft stores or just craft sections of regular stores. The color is up to you. They aren't absolutely necessary, but I prefer to use these over simple wire loops because the manufacturing process for these ensures that they are very smooth glass doughnuts with no burrs whatsoever. This is very helpful when they are used as pulley loops in a small mechanism, as they offer little friction and won't quickly shred the cord/thread/monofilament pulled back and forth through them.
⦁ adhesive that works on metal -- preferably more than one -- glue, resin, and/or putty. You should stick with your favorites if you have them -- nothing beats having a feel for your materials. Personally, I like JB Weld SteelStik or Blue Magic QuikSteel putties and "MAX 5 MIN" 5-minute epoxy resin from The Epoxy Experts (as far as I can tell, with epoxy resins, even minor differences in composition can yield very different adhesion/wetting/curing properties, and this particular resin is extremely difficult to remove from steel).
⦁ permanent or dry erase markers, or anything that will create marks on metal, at least temporarily. I used Sharpies in several colors throughout this project.
⦁ masking tape


Cutting The Tin

Now onto the actual work. First, do yourself a favor, and start by loosening the hinges of the box and separating the lid half from the back half. I ended up doing it eventually, and I don't know if it can be entirely avoided, so if you do it right away, it will, at least, make for much easier maneuvering for all subsequent steps.


Mark a shape to cut an "observation" opening in the lid. I left more material on the bottom to cover the crank mechanism (which can still be examined by opening the box). The factors to keep in mind are as follows:
The more material you leave, the more space you have create attachment spots for decorative elements or hiding spots for potentially unsightly structural/utility elements. You can also take advantage of opening the lid to reveal obscured elements, as a way of making the object more interactive/interesting (that was my intent in leaving the crank covered by the lid, but still able to be revealed).


BUT, keep in mind, that the more material you REMOVE, the more light can enter the box, allowing the viewer to actually see what's going on. I really screwed up here: I was NOT originally planning to have this automaton light up. What ended up happening was: I got to a point where I was happy with the aesthetics and the mechanics of all the elements, put the lid back on and was ready to call it finished, and THEN it became obvious that it was nearly impossible to see the heart through the closed lid/ribcage, even when it was moving, without pointing a light at it and looking very closely. So I had to choose between starting from scratch on the lid (realistically, just taking a fresh lid off another tin rather than grinding everything off to bare metal would have been the easier option), or adding a whole bunch more cumbersome/bulky/unsightly components to light it all up (which is what I chose -- though given how painful that was, in retrospect, I really wish I tried a new lid instead). This is all to say that the best time to decide if you want to bother with wiring an LED (and a switch, and a battery box) is right now, when figuring out how much metal to remove from the lid (and if you do want to have it light up, this is when you have the most options for placement of the necessary hardware -- it's much harder to work around existing elements that you don't want to ruin).


Once you mark out the area you want to cut out, the method depends somewhat on what you are working with. I did have cutting wheels available, but lost the screw that attached them to the stem, and didn't feel like running out to the store or ordering a replacement and waiting for it to arrive. Also, given all the inflection points on the vaguely shield-shaped opening I wanted, cutting wheels would not have saved much effort -- they would probably be most useful with a rectangular opening (though I would still suggest punching out holes in the corners and cutting between them). What I ended up doing was marking a few places around the perimeter of the opening I wanted, using a grinding bit on my rotary tool to thin out the metal, then balancing the tin lid on a plastic bottle lid (with the help of some masking tape to keep it in place), and hammering a sharp nail through the thin spot with a single hammer tap. I guess using a regular drill to drill the holes is an option, but it's not the easier option: first of all, you still need to use a nail or a punch to dimple the surface prior to drilling, or the drill will "walk" all over (and punching into intact, rather than thinned out sheet metal means more deformation); second, these "tin" boxes are actually tin-plated steel, and drilling through steel with anything but a fresh sharp drill bit is an exercise in frustration; third, even if you start with a fresh sharp drill bit, it won't stay sharp for long -- you'll likely end up going through several of them before the end; alternatively, high end drill bits meant for hard metal will last longer, but also cost more; in the meantime the cheapest generic ceramic/oxide/grinding stone bit will happily eat through steel until it wears all the way down to the stem.


However you end up making holes in your tin, when you feel like you made enough of them, you can force a needle file into them to widen them enough to use wire cutters to cut between them. Then smooth out the ragged edge of the opening with the files or the grinding bit, and you are done.

Attachments For Moving Parts

Besides the observation opening in the lid, there are a few more holes to make for the moving parts and/or wiring. Toward the end of my project, I had to enlarge the hinge openings in the lid because all the added material made it necessary for the lid to have much more wiggle room to still open and close. This can be done now (although you may still find that you need to widen them further later). If you want the automaton to double as a pendant, now is also the best time to create/attach a bail. I made a couple of holes in the top of the back half of the box (making sure the added bail still allowed the lid to fully close), then bent a soft wire into a staple shape, passed it through the aluminum tube that makes up the bail and bent the tips outward and flush with the box after inserting them into the punched holes. Yes, simply gluing it on was an option, but I just don't trust adhesives in structural applications -- not when I can use a fastener AND an adhesive, as in this case.


I used a similar technique to attach the beads to the inside of the back half of the tin, except with the staple tips going from the inside out -- and further secured them with some solder on the outside and some resin on the inside. I'll go into more detail regarding placement when we get to the rigging step.


Last but not least, make two holes near the bottom of the back half of the tin for the crank. Keep in mind that the crank will rotate, so you want to position the holes in such a way that it has the same amount of room on all sides (except the top, obviously) and runs as perfectly parallel/perpendicular to all the sides of the box as you can make it.

Step 2: Background Collage

collage_151444.jpg
collage_152407.jpg
collage_154626.jpg
collage_154830.jpg
collage_155536.jpg
collage_155808.jpg
collage_165225.jpg

The visual idea this automaton is attempting to convey is of a heart wildly beating or "fluttering" inside the human body. There is probably a million ways to make that happen, at least half of them good ones. So this step is really the most open ended (despite my personal obsession with floral viscera -- which is amply illustrated by the photos).


So, here are some other ideas: google "major organs diagram", select a favorite, crop and print. Or google "trunk x-ray", ditto. No printer? No problem -- run a similar image search, pull up your favorite, and try to draw that (to scale, or on a shrinky-dink) with whatever art supplies you have on hand. There are also some potentially spectacular options in already printed materials: health pamphlets, magazine ads, educational books for kids, etc.


If all else fails, including imagination (like it did for me on the first try) try something abstract: one excellent idea involves Sharpies (or other permanent markers), denatured alcohol (rubbing alcohol also works to make Sharpie ink run wild, but I don't know if it does anything interesting with the other ingredients), baking soda, and superglue. Scribble around with the Sharpies -- black is the most useful color, but some red and possibly blue wouldn't be amiss here; I'd skip green though... Squirt a few drops of alcohol on that and swirl. If you like what you see when the alcohol dries, you can stop there, if not, the next step is to sprinkle a few pinches of baking soda -- not much, you don't really want to add a lot of volume to the background, it might become a problem later on; and make sure you don't clog up the beads -- then squeeze a few drops of superglue onto the baking soda and give it a minute. When the minute is up, squirt another couple of drops of denatured alcohol onto the wet baking soda and watch what happens (what happened for me, quite on accident the first time, was a slowly spreading white film with very fine foam-like texture -- not 3-d foam, just to be clear, but something like an already dry print of a foam). Then maybe try the Sharpies again. One thing to keep in mind: baking soda + superglue makes an exothermic reaction, so don't mix more than a pinch of one and a drop of the other at a time; also, once they are done reacting, the resulting stuff is very hard, and it will be stuck very firmly to your box, so don't overdo it and leave no room for the moving bits to move.


One thing to keep in mind in this step: you are creating the background. The foreground will be the moving heart. You want to make sure the foreground is not overshadowed by the background and fighting to grab the viewer's attention. To that end: the foreground (heart) will need to be brighter in color and either significantly lighter or significantly darker than the background; the various elements of the background should contrast less with each other than the background overall does with the foreground; if you can, make the texture of the background matte and the heart glossy, to give it some extra "pop" (I slathered my collage with several layers of matte Mod Podge to achieve the effect).

Step 3: Movement

crank_135337.jpg
crank_182717.jpg
crank_143246.jpg
crank_220933.jpg
heart_rig_103508.jpg
heart_rig_105306.jpg
heart_rig_223348.jpg
heart_rig_234308.jpg
banner_3.jpg

Tools/Materials:


⦁ Hard wire. In this automaton I used some (relatively thin) spring steel wire in either hard or half-hard temper -- commonly sold as "memory wire" in beading sections of craft stores (it's not actually memory wire, whichi is something else entirely -- that's why the scare quotes). I'd also previously used a straightened out paperclip -- one of those big, uncoated metal ones -- much thicker than the beading wire, but hard and springy enough to do the job.
⦁ Wire cutters and pliers that can handle hard wire. Or just old wire cutters you've already ruined by using them on hard wire.
⦁ Nylon monofilament in 4 to10 lbs strength -- can be found in the beading section or the fishing section of large stores (I'd head straight for the fishing section -- the weights are clearly marked, and you get a decent selection -- I prefer the 4lbs thickness, personally). While you are there, you may also find some crimps -- the thinner the better. If not, try the beading section: you want #0 or #1 size crimps only. Even if the final product ends up containing half a dozen crimps, you might want to have at least several hundred on hand. If you can't get the crimps, you can't use monofilament either -- there won't be room to tie it. Whether or not you can get crimps, you may want to consider braided fishing line instead -- it is more abrasion resistant than monofilament, and flexible enough to be tied around the crank (though I prefer a knot + a crimp).
⦁ A bead cap or a metal or plastic snap to use as a "heart platform" -- basically an element that can serve as an attachment point for the heart later, but is easier to work with and provides plenty of attachment points for the cords to move it around. It is not strictly necessary, but will go a long way toward saving your sanity.

This automaton uses a crank with six arms (bent portions). The arms work in pairs -- arms within each pair lie in the same plane, are offset the same distance from the center of rotation, but in opposite directions. The three pairs of arms are positioned on planes at 120 degrees to each other (just picture an asterisk * -- that's how the arms are arranged, each pair on the same line through the center).

Each pair of arms is used to pull a cord through a pulley (or a pair of pulleys): one cord passes through one of the beads at the top/center of the box and provides up-and-down movement (from the perspective of the viewer); another cord passes through two beads roughly in the middle of the left and right sides of the box, and provides horizontal movement. The third cord (this one is a nylon monofilament) passes through the other bead at the top/center of the box, and through two small beads attached to the back of the "heart platform" (on either side of the top center bead), providing a slight twist forward and backward to its respective side of the heart platform.

If the third one sounds confusing -- it is. It is also frankly unnecessary. I wanted to make this automaton's motion more complex than my first attempt, and it is, but mostly because the first two pairs of arms are out of phase by 120/240 degrees rather than 180/180. The slight twist provided by the third pair is not even very noticeable, I just couldn't help myself and had to give that pair something useful to do. Do yourself a favor, and just give your crank four arms. And if the intersecting planes also sound confusing, instead of an asterisk, go with a plus (+) -- make the crank arms form a cross around the center of rotation.


Once you bend the crank into shape (make sure it can be freely rotated once inserted into place -- if not, make another one, with shorter arms). You will notice there are resin dots in each bend of my crank -- those are there to prevent the cords from slipping off their designated arms. I used a dab of J-B Quick here, but any method of making the bends much thicker than the rest of the wire would work: solder is an option, so are crimps, so is a bit of sewing thread wound tightly onto the wire and soaked with a drop of superglue. Just make sure that when you are done, the crank can STILL freely rotate when placed into its designated holes.


When the crank is in place, it's time to do the rigging. Select your "heart platform" and secure either two braided fishing lines or two monofilament lines to it in a "+" formation (or use four strings instead of two, so long as they form a cross). One of these lines will be used to pull the heart up and down (y-axis) and the other left and right (x-axis). The tricky bit is that it is a single object being pulled along both axis at the same time, so its movement will be cyclical, with x-axis and y-axis controlled by a separate pair of arms. Which means both lines (the x-axis line and the y-axis line) will need to have some slack in them to allow the heart to be moved diagonally without pulling against and deforming the crank. How much slack you will need will depend on the specific arrangement of elements you are dealing with (also, if you are using monofilament, you will need a lot more slack than if you went with the braided line, because it is much stiffer). I'm afraid this part requires a lot of trial-and-error to work out. I'm sure formulas exist to work it all out on paper, but given the scale vs. likely measurement precision, you'll end up right back with trial-and-error. Just make sure you have LOTS of crimps, and give it a go.

Step 4: the Heart

heart_all_tries_171539.jpg
wings_134841.jpg
wings_135705.jpg
wings_140418.jpg

Again, there are lots of ways to handle this part. A heart-shaped bead or button would do the job, as would a cardboard cutout, a tiny fabric heart stuffed with a cotton ball, a piece of wire twisted into a heart shape, a shrinky-dink, a wine or bracelet charm, etc. I wanted a convincing-looking "anatomical" heart (convincing to an educated layman, not a cardiologist), and I even armed myself with a laminated educational pamphlet, providing several diagrams to illustrate the function of the four chambers and various blood vessels in detail, but most importantly for my purposes both anterior and posterior views. It still took me most of a day and five tries to carve a tiny heart out of balsa wood without mixing up all the major blood vessels. Then, when I couldn't see this tiny heart in the "finished" automaton, I carved two more in two different sizes (the larger of those was too big and could not be moved around with the lid closed), before realizing that the size of the heart alone just wouldn't do the trick and I needed to light it up.

The first lighting trick I tried, however, involved casting the last, "medium" size heart in red-tinted clear resin, after making a mould of the carved version from silicone mould putty. I figured that, by placing a reflecting surface (a bit of holographic mylar film) on the back of a translucent heart I could make it look luminous, which would make it more visible (spoiler: it does look like it has its own, internal, glow, but that didn't make it sufficiently noticeable). While the amount of effort required was definitely an overkill, I do like the look of the result. I would not recommend making this your first try at using resin: there are a few translucent polymer clays available that should achieve a similar effect, with much less effort. And if you are wondering how some of the blood vessels ended up being blue while the rest of the heart is red: I cast the heart with epoxy resin tinted red with alcohol ink; then, after it cured, I took a few drops of UV-activated resin, tinted it blue (also with alcohol ink), and used a tiny paintbrush to apply several very light washes of blue to those blood vessels (fixing each wash with a UV flashlight before applying another).


The heart attaches to the "heart platform" by a clear plastic snap -- the half attached to the platform is stitched on with monofilament and the half attached to the heart was placed into a carved-out spot and secured with UV resin. Had I known what the final version of the heart would be, I would have simply used the half of the snap as the heart platform, instead of cutting one out of a beadcap and later attaching the snap to it, but it doesn't make much difference in the final product.


Another idea I was pretty attached to was making the heart literally flutter by giving it a pair of wings. I made these wings from holographic flower-shaped sequins (I cut a few of the petals off each flower to form something that looked like fly or bee wings -- or fairy wings, given how sparkly they are). I then pierced them with a needle, and threaded/knotted a piece of monofilament through the holes. The monofilament does double duty -- it isn't very visible, so the wings sort of float around the heart, and it's slightly stiff and springy (this is the 4lbs line, anything heavier would be too stiff) so as the heart moves around, the wings shake around (or flutter!) slightly on their monofilament stems. Finally, I snipped pieces of a floral beadcap and used them as crimps to secure the wings in a stable position on the monofilament (otherwise gravity might make them twist and droop in the knot), and, as a free bonus, also add a little bit more weight to the wings, and slightly increase the amplitude of their oscillation.


The first (tiny) version of the heart had the wings secured directly to the heart itself (I tied the monofilament to a piece of wire which I then forced into the back of the heart and secured with a drop of superglue). The final version has the wings tied directly to the heart platform instead -- I didn't want to drill holes in the resin heart, as they would have been potentially visible through the front.

Step 5: Lid Decor

banner_1.jpg
banner_2.jpg
banner_3.jpg
pins_blank.jpg
pins_box_attachment.jpg
pins_tux.jpg
pins_dress.jpg
ribs_161944.jpg
ribs_130027.jpg
ribs_193643.jpg
box_lid_paint1.jpg
box_lid_paint2.jpg
box_lid_banner.jpg
box_lid_ribs_attachment.jpg
box_lid_together.jpg
box_lid_ribs_done.jpg
pins_demo_dress.jpg
pins_demo_tux.jpg

I screwed this one up in one important way: the opening in the lid does not provide enough light to see inside and the ribcage covering the opening is too substantial, further obscuring the action, and making it necessary to have the automaton provide its own light source. Other than that inconsequential detail, it's a glowing success.

I was having a great deal of trouble figuring out how to make the lid work with and further develop the theme of the piece. Nor can I now come up with good alternative suggestions, so I guess I'll settle for illustrating the specific choices I made.


The keystone piece is the title "banner". I was struggling to think of a way to add text to the piece that wasn't just an incongruous little plaque, and remembered that various mottoes used to be a staple of old-timey illustration and heraldry (the latter association is what made me think of cutting out a shield-like shape in the lid), and that they used to appear on stylized banners or scrolls (frequently held up by a couple of cherubs). So I used some clear shrink plastic to make a similar banner and went completely overboard with silly fonts. In case you were wondering, the "e" in "Make" is made up of a couple of foil gears sold as manicure adornments. The back of the banner is painted with pearlescent white paint and the top is laminated with UV resin.


The lid opening was always meant to be covered by a ribcage. My first automaton had a ribcage carved out of shrink plastic (I actually shrunk a solid piece with a ribcage drawn on it and then carved it out, because trying to shrink a carefully cut-out filigree ended with a snarled mess, as the plastic curled in the heat and bits of it stuck together and tore when I tried to separate them). I wasn't entirely happy with clear plastic ribs, so this time I tried another idea (which didn't work): a front part of a ribcage of a halloween decoration, carved into a finer and more accurate shape. It didn't not work for any mechanical reason -- I just didn't like how it looked -- it didn't aesthetically fit with the rest, so I ripped it out and tried again. The second attempt is made of aluminum wire, which is liberally slathered with the 5-minute resin (mentioned in Step 1) for structural stability. This version I really liked, so much so that I went to great deal of trouble to keep it. Nonetheless, I did have to make one major adjustment to it, which I'm not 100% happy about: the clear resin over silver aluminum wire was very, very shiny and attention-grabbing, and no amount of color, lights, and motion made the heart behind this ribcage grab visual focus, so the ribcage had to be stained with alcohol-diluted-sharpie (and covered by a bit more resin, so the sharpie ink didn't get rubbed off later). This visual adjustment really worked wonders for making the ribcage give up the spotlight, but I liked it better silvery.


The best idea I could come up with for the remainder of the lid, beyond the rib cage, was to make a suggestion of clothes on a torso. Gluing fabric to metal isn't especially difficult, except when it comes to fabric texture -- any glue that saturates the fabric will change its appearance, and, at its most inconspicuous, make the fabric look wet. In case you are wondering if matte Mod Podge would help with the wet look -- I tried it, it doesn't -- even if it's the only adhesive used. A glue roller -- the adhesive dispensed by the thin plastic tape inside is similar to what you'd find on the back of paper stickers, or scotch tape -- it is tacky, but doesn't saturate fabric, so there are no issues with the wet look, but it leaves the fabric not sealed in any way -- so nothing prevents it from picking up grime, but it won't be readily washable either, not without disturbing the adhesive or soaking the whole tin in the process (and seeing how the tin isn't made entirely of tin, it can and will rust). In the end, I scrapped the plan involving the use of actual fabric, and went with a painted simulation, which both looks better and doesn't present any of the same logistical problems.


I wasn't making this automaton for anyone in particular, or with the intention of addressing it from myself (not this time, anyway), so the choice of "clothes" wasn't straight-forward. I also wanted it to wearable as a piece of jewelry -- either a pendant or a brooch/pin, (and, presumably, if it IS given as a gift at some point, the wearer would be the person to whom it is addressed, not the person who gives it). With these considerations in mind, I wanted to make sure that whether it was given by a man or a woman, and received/worn by a man or a woman, it would be sufficiently unisex to work in any context. At the same time, I didn't want it to look stark or unadorned or generic. So I to go with a (simulated) fabric design that suggested a silver-on-black brocade or embroidery (that would look suitably chic as either men's or women's eveningwear), and to top it off with a changeable collar -- either a shirt and bow tie or a blouse and pearls.


The collar swap is accomplished by means of punching an extra hole in the lid, above the rib cage, roughly where a collar might normally be found, and securing the back of a tie-tack directly to the other side it, so that inserting a tie-tack into the hole would allow it to be held securely in place. Then, I made two different collars out of Yupo paper (which is not actually paper but thin, printable polypropylene sheets -- which, unlike paper, are almost impossible to tear, waterproof, non-porous, and quite resistant to picking up dirt), and accessorized one with a black Duck tape bow tie and the other with cream-color, pearlescent-luster glass seed beads (selected specifically for their resemblance to pearls) strung on bits of thin copper wire, and mounted them on two tie-tack tops. The "tuxedo" collar also has a red triangular manicure gem glued to the top of its tie tack, to suggest a button, and the "dress" collar has a red Swarovski crystal for same. In addition, I used an antique post earring (the earring post itself was too thin to serve as a tie-tack stem so I snipped it off) and glued the already-set crystal to the top of another tie tack -- to create a third, altogether more abstract adornment to make the tin look complete while avoiding any direct visual allusions to specific articles of clothing.


Finally, I domed the painted tin lid with resin to make sure the acrylic paint actually stays put, since it is otherwise trivially easy to accidentally scratch off with a fingernail (I actually tried priming the lid with black acrylic gesso, but it made no difference whatsoever -- there might be a better way to get paint to stick and stay put, but I didn't have anything on hand that I could be certain would work). Again, if you've never used resin before, I don't recommend making this your first try -- in this case, because it's WAY too easy to make a sticky mess of everything you've already made. An excellent substitute is to varnish the paint with some polyurethane -- I highly recommend the oil-based kind, which is in no way harder to work with than the water-based kind. Don't be daunted by having to layer it or to sand between layers -- first of all, you won't need many layers, two or three should work fine, and, secondly, "sanding between layers" is your friend, allowing you to nip any bumps or bubbles in the bud before they turn into gloppy mountains -- take it from the person who can't apply a smooth even layer of anything to anything to save her life.

Step 6: Lights

components_LED.jpg
components_snaps_beadalon.jpg
battery_bag_components.jpg
LED_string_init.jpg
LED_string_init_test.jpg
LED_string_soldered_demo.jpg
LED_string_soldered_sheath.jpg
switch_components.jpg
switch_front.jpg
switch_side.jpg
switch_under.jpg
switch_wired_in.jpg
battery_bag_done_back.jpg
battery_bag_done_front.jpg
LED_string_wiring1.jpg
LED_string_wiring2.jpg

Tools and materials:
⦁ Battery and battery holder (see below for a discussion of options). Optionally, for a battery holder, you will need wide diameter electrical shrink tubing, without adhesive -- it needs to be AT LEAST 20mm wide when fully shrunk, but 25-30mm will work better. A heat source will be necessary to shrink the tubing -- a lighter will do, but a heat gun is better, if available.
⦁ Rare-earth magnets, the thinner the better -- think 1-1.5mm at most-- and the ones with a hole in the middle are the most convenient.
⦁ Insulated stranded wire, thin enough for this application. The absolute best option for this, however, is Beadalon or similar beading cord: stranded steel wire covered with a clear plastic sheath. The reason this works so well is because the steel this wire is made of is very resilient and resistant to metal fatigue, and it is both thin and insulated.
⦁ Small metal sew-on snaps for contacts. The bare metal ones are conductive and will accept solder. The ones with powder-coated paint on them will need some paint removed with a rasp, but the remaining paint will provide a convenient resist to stop solder flowing where it shouldn't. Additional contacts (like for the battery) can be made from any conductive flat bits of metal, like bead caps or pieces of the tin cut out of the lid.
⦁ LED(s) and resitors, if necessary.
⦁ A small switch
⦁ Continuity tester/multimeter -- this is absolutely crucial to keeping your sanity, even if you have to make one from scratch with a light bulb+wire+battery, or by hacking a flashlight -- take a break and make one now.
⦁ Wire cutters, pliers, tweezers
⦁ Soldering iron/solder
⦁ Adhesive epoxy putty, hot glue, or resin -- these will do double duty as structural reinfocement and insulation.
⦁ Needle and thread

Battery holder

You will need to pick a battery to use: the size will determine placement, and the voltage will influence your choice of LEDs and resistors, if necessary. I went with CR2032 because, for a coincell, it is very distinct in size and shape, and it has become extremely common, practically a standard for a variety of small electronics. Both these facts mean that a replacement battery will be easy to find, and the user will not find themselves in the unenviable position of standing in a store in front of a selection of coin cell batteries, suddenly remembering they need one for that automaton pendant thingy, and realizing that they can't remember which exact one is needed, so they'll have to remember to take it out and bring it with them next time (rince, repeat). I hate it when it happens to me, so why inflict it on someone else?


Of course, if you went with a "standard" size mint tin, AAAs would be an even more considerate choice.
Now, much as I've belabored my poor planning and having to tack on the electrical bits at the end, I don't know if there IS a better place to stick the CR2032 than to the outside at the back of the tin: despite how flat it is, it would still take up too much space if placed behind the heart, leaving little room for it to move (of course, if you went with a heart-shaped button or a shrink-plastic cutout for the heart, that could work, but you'd need to make sure the battery could be removed/replaced without cutting any of the strings).


If you happen to be reading this in the planning stages, I would recommend that you save yourself a big headache and buy a few (very inexpensive) components from Adafruit: they have breakout boards to hold a CR2032 both with and without an integrated switch. They also offer a selection of LED "sequins" and suchlike, specifically designed to survive being used on wearables. In fact, if my battery holder gives me any more grief, I'll be replacing it with one of Adafruit's: I've tried everything, and theirs are the thinnest.


I'll spare you the descriptions of my numerous failed attempts to either make a battery holder from scratch, or cut an existing one down to size. What finally ended up working was a piece of (wide -- it needs to hold a battery 20mm wide and a few mm thick) shrink tubing -- yes, the kind you'd use on wiring. When shrunk, it is slightly flexible and leather-like, with a pleasant texture. To leave enough room in the coin-cell purse, I put a nickel (a coin slightly larger than the battery, if you are not a denizen of US), together with the battery, inside the shrink tubing section, before shrinking it, and while it was still hot, I used tweezers to firmly pinch one side of it together and bend it back, creating a bottom that didn't need to be sewn together. The contacts are made from pieces of a bead cap, and sewn to the inside of the coin-cell purse. I also added a small ribbon (sewn to the back/inside the purse), which goes around the battery as it is inserted, and sticks out slightly, so that pulling on it will extract the battery with minimum effort. The sequin (domed with resin for durability) on the front of my battery holder is strategically placed to cover the stitches holding the negative electrode in place as well as providing a diagrammatic suggestion of which way the battery is to be placed into the holder. A rare-earth ring-shaped magnet is sewn to the back and further secured with some J-B Quick resin/putty (the kind that comes in squeeze-tubes, not the solid putty roll).


Alternatively, you could make the battery contacts from two rare-earth magnets and place both inside the shrink tubing with it (just make sure the tubing is wide enough to accommodate all that -- and maybe put the magnets around the battery+coin combo before shrinking it, so there's enough room for them inside). The battery holder will still stick to the tin if there's a bit of shrink tubing between the tin and the magnet, but two magnets might make the battery holder too bulky, plus inserting and removing the battery might become tricky, as these magnets are extremely strong. The upside of this option is that there will be NO doubt both electrodes are fully in contact with the battery at all times, which is not as easy to ensure on this scale and with all the sewing thread involved as it may seem.

LED

There's a cornucopia of options for choosing an LED -- I wouldn't even know where to start outlining them, so I'll stick with detailing my final selection. I cannibalized a battery-driven 10-LED light string from a dollar store, which, because it was sold as Halloween decor, came in "orange". The reason for the scare quotes is that the "orange" LEDs actually ended up being more of a coral pink color -- which I though was both pretty, and a good match for the interior of the automaton. The original battery box was supposed to hold two AAs, and there were no resistors in the circuit, so I surmised that the LEDs were supposed to handle the 3V that my CR2032 was putting out. I also traced the circuit polarity to figure out if there was a way to tell which end of the tiny surface-mount LEDs was which once they were snipped off the wires, and it turned out there was: there's a tiny green triangle/arrow printed on the back, with its tip pointing to the NEGATIVE end.


The wires these are mounted on look bare, but they are, in fact, insulated by the silvery color coat (I can't tell if it's paint or plating) -- under which, the wires are actually copper. The reason I ended up cutting them at all is because I wanted several closely-spaced LEDs, and, although the same effect could be achieved by coiling the wires between the lights, the coiled-up wire ended up adding way too much bulk. So I snipped 5 LEDs off the tip (all 5 didn't quite fit in the allotted space, so I eventually cut one off) of the string, leaving about a centimeter of wire still attached at both ends. I sanded off the silver insulation off the wire tips, and twisted them together, connecting the 5 LEDs into a much shorter and more compact string. Two sections of Beadalon wire went to one end of the new light string to make leads to the rest of the circuit. Once I made sure that none of the bulbs had their polarity reversed, and the sanded wire tips made good contact with each other, I soldered them together and snipped off the ends. I was careful to point the wire connections away from each other to avoid a short circuit, and once they were soldered together, I used some hot glue to insulate them and keep them from touching.


The resulting short LED string looked too messy for my taste, so I grabbed a section of shrink tubing, cut out little holes in it that lined up with the spacing of the bulbs, and pulled it on. Not content to stop there, I took some aluminum tape (because it's both shiny and adhesive, so a decorative holographic foil tape would also work), and cut 5 little ovals out of that, which I placed through the holes and inside the shrink tubing behind each LED, to serve as tiny reflectors. Once everything was in place, I shrunk the tubing with a lighter.


When the light string was ready, I taped the stripped wire tips of the Beadalon leads to my test battery, and set about trying to figure out how best to place it inside the tin. Despite every photography tutorial I ever read extolling the virtues of side lighting, it turned out that in this case, the best, and, in fact, the only effective way to light the scene was directly from the front -- every other configuration failed to do the one thing that this whole lighting exercise was all about: make the heart take center stage in the composition. Finally, the way-too-substantial wire rib cage insert in front actually made itself useful: it turned out to be perfectly suited to holding the string of lights in place AND making it all but invisible when viewed from the front. Since I wanted to make sure the LEDs would be replaceable (more on that later), I decided to tie them into place behind the rib cage with a couple of pieces of monofilament (rather than using glue, or putty, or resin), and used my mini-grinder to cut shallow grooves in the "sternum" for said monofilament to fit snugly and all-but-invisibly.


Wiring


Now, I know that all things are destined to end up in a landfill eventually, but I get sentimentally attached to things, and it bothers me when they break -- especially if they can't be repaired. It's worse when an object's sole purpose revolves around evoking feelings: all things decorative and ornamental, all things typically given as gifts are designed to take on symbolic meaning and value in people's lives -- and to be difficult to part with. So when such objects break and can't be easily repaired, they stick around in a sort of half-existence, hidden in drawers and boxes, no longer fit to serve their original purpose, but too precious to throw away, and they make people sad. Design is ultimately all about picturing the future of an object that does not yet exist -- so I tend to think a lot about how an object I'm making will function, in which ways it can fail, and what the result of said failure would be. When designing this little automaton, I thought a lot about how it might break, and how it might be repaired if it does, if its future owner is not inclined to simply toss it in the trash, of course. That's why I made such a fuss about the battery being easily replaceable, and that's why I opted to make the rest of the electrical components as modular as I could manage within the space constraints.


It's not especially difficult to use some wire to connect an LED to a battery and a switch -- if one even bothers with a switch. Obviously, the battery is the weakest link in such a circuit -- once drained, it must be replaced, or the whole thing is now junk. But if we make the battery replaceable, the next thing most likely to fail is the LED -- they do last a long time, but not forever. I was going to stop at making the battery and the LED replaceable, and embed the switch more firmly in the body of the automaton -- and I would have, if the switch I was using didn't give up the ghost just when I was about to call the project a success (I confess, I was really mad about that). That's how I eventually ended up with all of the circuit components (besides the wires leading from the outside/back of the tin to the inside of the lid), and after trying a few ways to connect them firmly, but not permanently, I ended up using metal snaps. It's not critical, or even strictly necessary, to go this far in making everything replaceable, but if that's what you are after, here are a few things I learned along the way.


The humble little sew-on snaps are actually miniature feats of engineering. The "socket" half of each snap contains two tiny wire springs, which coil around the perimeter and converge on the socket in the middle, where they enter through two grooves on opposite sides to grasp the ball on the tip of the "post" half of the snap. This is relevant because, when one solders wire leads to metal snaps, it is extremely important to make sure solder does NOT get on those tiny springs, because if it does, they no longer "spring", and the two halves of the snap can't be connected. Now, if you've ever wrestled with electronic components which are MEANT to be connected by solder, but are for some reason made of/coated with an alloy that absolutely refuses to accept it, you'll be relieved to know that these little snaps, meant to be attached to clothing with sewing thread, LOVE solder -- it flows like a charm, and holds on for dear life, flux or no flux. Unfortunately, those tiny springs are no exception, so watch out for that, and make sure to test the newly-soldered snaps not just for continuity, but for still fitting together. It turned out that the painted snaps are actually easier to deal with, despite having to sand some of the (insulating) paint off the outer ring and the ball/tip of the "post" to make them conductive -- the paint is also an effective solder "mask", so it will only flow onto the sanded portions and leave those all-important springs alone.


Beadalon wire, on the other hand, is not friendly towards solder, but the strands inside are (literally) hair-thin, so as long as each of those steel hairs is wound snugly around a bit of bare metal that does like solder, the solder will flow all around the wire strands, helping them stay in place and increasing number and size of metal areas in contact with each other. This means that each end of the wire needs to have about an inch of its length stripped off, unwound, and each filament individually wrapped around its own section of (snap, or ) whatever component it's being attached to, then twisted/braided back together, to make a "cold" connection secure enough to work even without solder. It's very fiddly work -- but I like doing it, so, to me, the extra effort involved is worth it for the resilience advantage that Beadalon offers over copper stranded wire (which, in my mind, is way too vulnerable to metal fatigue to be reliable). It's a trade-off worth keeping in mind. Whichever you choose, however, copper or steel, should probably be stranded rather than solid to make a good connection with those snaps -- they have four holes around their perimeter, and you'll want to pass at least one strand of wire through each of them, rather than relying on just one.


As mentioned, a battery-switch-LED circuit is very simple, and not a lot can go wrong with it that's not immediately obvious. However, adding a whole bunch of physical components to make it modular changes this rather dramatically: every physical connection in every one of the components can fail, and it won't be obvious which one did; additionally, the tin itself is conductive, so every place where an electircal component touches the tin is an opportunity for a short circuit. This is why a continuity tester is so important: you can test each component as you manufacture it, and test it again after it's in place -- both for continuity within the circuit and for DIScontinuity with the tin, so if/when this really really simple circuit somehow manages NOT to work, you can (hopefully) locate and fix the problem.


I went out of my way to color-code positive and negative polarities of the battery holder and LED snaps (the switch snaps are purple, because it can be inserted anywhere, at least theoretically; it can also be removed entirely -- one of the battery holder's snaps can double as a switch). I'd like to claim this is intended purely as a user guide, but I probably needed it more than anyone, because, when given the opportunity to get distracted and get my wires crossed, I'll do it every time.

Step 7: Putting It All Together

hinges_1.jpg
hinges_2.jpg
hinges_3.jpg
hinges_4.jpg
hinges_5.jpg
lid_wiring_132910.jpg
lid_wiring_142346.jpg
lid_wiring_152323.jpg
box_back_rubber_181422.jpg
box_back_wired_switch_185317.jpg
box_back_final_101343.jpg
box_back_final_101554.jpg
lid_wiring_test.jpg
test_front_lit_101654.jpg

Depending on how much stuff you've added to your mint tin, this could be as simple as putting the lid back on and tightening the hinge tabs back up. I added a whole lot of stuff to almost every surface of the tin, so, while it could be closed and stay closed, it could only open a crack with the original hinge tabs tightened to their original positions. Because the box walls are now effectively much thicker on all sides, the lid needs a lot more wiggle room to fully open.

I tried a few different ways to extend the original hinge tabs enough to provide this wiggle room: wires, ribbon, tape -- they all worked in the purely mechanical sense, but they all looked terrible. Then it finally hit me: I've used sections of metal bead caps in a bunch of places on this automaton, why not also make hinge extensions out of them? Sure enough, a large stamped-metal bead cap provided a couple of ornate elongated-heart-shaped sections that fit the bill. I used a couple of bits of wire to form loops between holes drilled in the original hinge tabs and the rim of the back half of the tin, and to connect the hinge tab extensions to the originals. Once the beadcap sections were in place, I colored in the lower portions of the relief stamped into them with a Shapie, and used some 5-minute resin to coat the beadcaps and bridge the wires holding them in place, and the visible portions of the original hinge tabs into a single surface -- both visually appealing and structurally very strong.

The brooch connector, switch, and battery holder are all attached to the tin by rare-earth magnets, which are actually strong enough to require some effort to pull away. The battery holder and the switch form a circuit by snapping together, but the snaps are used purely to close the electrical circuit, rather than doubling as physical attachments, which should, in theory, improve their reliability by reducing physical strain on the connections -- instead, it's the magnets that hold them in place. This is the main reason the back of the tin has so much material added in some places, while remaining bare in others: the magnets might be difficult to pull away from the box, but they slide around effortlessly, so adding putty and resin keeps components where they belong (while also hiding/protecting the "integrated" wires that lead from the back to the inside of the lid where the LEDs are).


The wiring that had to go inside the tin -- namely the leads connecting the battery/switch at the back to the light in the lid -- gave me a lot of trouble. The Beadalon wire is springy and resists sharp bends, so, in order to pass it from the back to the lid (and not have it float loosely outside the tin), I had to make a small hole through the back, right at the edge, pass the two wires through it to the inside of the tin, and then loop them through the hinge opening (which had to be widened a little to accommodate them) and into the lid. Only then did I attach the snaps for the light to plug into and secured them to the lid with steel putty. Because the lid opening and closing changes the distance the wire has to cross, portions of it had to remain loose -- unfortunately those portions also had to be inside the tin lid, above the opening -- which is where the heart's wings also need to have room to loosely flutter around, without catching on anything and getting yanked at. Getting those two components out of each others way was tricky, and I didn't get it perfect -- the wing on the left side occasionally still bumps into the negative wire, which is closest to the hinge and had to be coiled into a half-hitch knot to give it room to expand and contract. I don't have any better ideas for how to deal with that problem, so it remains only partially solved.


The light string attaches to the ribcage in a way that forces its wire leads to travel from the bottom of the lid to the top, where they plug in -- it seems backwards, but because the wire they are made of is as stiff as it is, the longer leads are actually a lot more cooperative than shorter ones would have been, which is why I attached it this way. I still had to come up with some way to keep the wires close to the lid and out of the way of the crank and the moving heart, but they couldn't be glued or embedded in putty, or the light component would no longer be replaceable -- so I attached a bit of wire to one half of a snap, forming a tight hook, and glued the other half to the lid. The little hook puts a little tension on the wire leads, so they can't simply slide out of it, but it can always be un-snapped from its mate to remove the LEDs, and snapped back into place to hold the wires of its replacement. The rest of the inside of the lid is covered with some black Sugru, which I rubbed with gunmetal-gray mica before it was completely cured, giving it that silvery look -- it's there purely to pretty up an otherwise messy-looking lid.


And that's it, finally -- everything is put together, everything works, and I'm pretty happy with how it turned out. I'm also itching to make another automaton. I hope you enjoyed following along, and, if you made your own version, I hope you had as much fun with it as I did with mine!