Mechanical Apple Watch From Real E-Waste Apple Watch

The pictures and videos speak for themselves, for the last 3 months I have been constructing a fully functional mechanical Apple Watch! A dumb watch if you will! *chuckles at own joke*

Why? Well it all started around last December when I was shopping for a regular Apple Watch and I realised that I would be spending a couple hundred dollars minimum, for features I did not really need, probably would not use, in a package that would be useless in 5 years either from lack of support, or degraded components. At some point, this idea popped into my head and I found it so funny. I just knew I had to make it.

The other thing that I think makes this project so great is that because these watches are so popular, there are a bunch of second hand ones that are near useless and get sold quite cheap. Whether they are just old/slow, they are iCloud locked, the batteries are degraded, or the screens are cracked, there is an abundance of parts to chose from for this project. I, myself, picked up two $1 watches to mess with before grabbing the nicer, stainless one I used in the final project. One of them even worked in the end, it just took a couple hours to update!

Speaking of my donor watch, it was actually working when I bought it but the battery only lasted ~4 hours. And being a series 1 Apple Watch, it has been unsupported by apple since 2020. But, it is still a beautiful watch. And like it or not, the watch's form factor is iconic now. No one was ever going to use this watch again, it was e-waste (still cost me $50 somehow thanks to the premium materials), and hopefully I have given it back at least another couple of years of use!

I have been wearing it on and off for about a month during the build process, and prior to writing this. The crown only works to wind the first ~20% because after that the pulley system I made does not have enough friction to overcome the main spring. It fully winds from the rotor in about 2 days of constant use. It lasts the Seiko rated 40 hours easily after winding the rest of the way with the rotor. And it loses anywhere from 3-20 seconds a day (not sure why such a difference but most days it is a consistent 3-4 seconds then one or two days it jumped up).

I do also have to mention that there is a real watch brand that has made something similar, H. Moser and Cie., with their Swiss Alp watches. I only learned of this while searching to see if anyone had dome something similar and now that I know a little more about watches, it makes total sense that H. Moser and Cie. had already taken the piss out of the apple watch =).

Major Components:

Seiko NH38 watch movement

42mm Series 1 Apple Watch in stainless steel

Other parts:

Back glass: 0.8mm thick, 27.7mm diameter mineral crystal from ebay

Hands: $10 Nautilus style hands from aliexpress/ebay

Rotor: Silver Côtes De Genève rotor from Mod Mode Watches

Nitrile O-ring (3mm I.D., 4mm O.D. 1mm cross section)

0.2mm K&S bronze sheet

Brass tube (0.8mm I.D, 1.6mm O.D)

3mm K&S brass rod

Dial feet kit from aliexpress/ebay

Aluminium NH35/36 movement holder from aliexpress/ebay

Hand pressers from aliexpress/ebay

B7000 adhesive

JB Weld

Rodico and watch screwdrivers from local jewellery store

I chose the NH38 because it is ubiquitous and easy to find for cheap thanks to the large Seiko modding scene. I chose the 38 over the 35/36 because the extra day/date complications would have compounded the issues I had with the stem. And I liked the minimal look. And I quite liked the open heart, especially because I didn't always know I would have a clear back so needed a way to prove it was mechanical.

Similar to the NH38, lots of consideration went into the watch I chose. It had to be from series 1-4 as those are the ones without the full ceramic back and I thought the full back would make my life a lot harder. From that I chose the series 1 because it was thinner but that almost came back to bite me as I had very little clearance.

And I knew I wanted stainless/sapphire glass because my goal was to make this watch last at least twice as long as a regular apple watch (so maybe ~10 years). You can see in the photos that the series 1 with sapphire glass has no visible scratches while this series 3 is unbelievably scratched. It's chalk a cheese. And based on the wear on the cases, I would say the series 3 is much much less used.

The first step was to gut the apple watch completely. There are plenty of tutorials how to do this on youtube however most of them do not tear the watch down this far as there is little need to when replacing a battery/screen. I did, however, find one two youtube videos, both of which showed people taking an apple watch all the way apart so they could replace the cases with clear, cast replica cases. Both videos really deserve more views and really helped me understand how to completely gut the watch.

https://youtu.be/c9rJFID30JQ

https://youtu.be/Y78LgqLPanA

I was quite worried that the front glass that it would just shatter when I tried to remove it. So instead of getting a razor under the glass like most tutorials, I just heated the watch up with a hot air gun (quite a few times) until I could just pull the screen off using a suction cup that came in my phone repair tool kit from ebay. The heat eventually softens the glue up enough and we don't have to worry about overheating the battery or any components as they are all junk anyway.

Popping the back glass out was much the same as the front glass, I simply heated the watch quite hot and for a long time until the glue gave way.

The apple watch screen is actually made up of a number of layers. On the top is the glass (in this case, sapphire, which made these next steps much safer as I didn't risk making a bunch of scratches on the inside of the glass), then there is some LOCA holding the digitizer that detects your finger presses, and then the LCD screen.

The LCD screen pulled away fine with a little heat and isopropyl alcohol, but the digitizer was held on a LOT more securely and it seemed that no amount of heat/solvent would touch the LOCA in any reasonable time. So instead I just started going ham at the digitizer with a sharp screwdriver to crack it all off.

When removing the digitizer, I was slightly frustrated at how long everything was taking, especially as I still very unsure if this project would even work, so I rushed and ended up scratching the black border on the glass in many places.

The scratches were very visible but I decided to try just painting over the scratches with black nail polish and this worked way better than I could have expected. When installed on the final watch, the scratches are essentially invisible.

As a slight aside, I think an empty Apple watch with the screen on would also be a cool bit of jewellery.

The next thing to go was the sensor holder that senses the digital crown turning. I didn't really have a good way of getting rid of this. And again, as I still wasn't sure that the project was even feasible, so I just went at it with some side cutters. I think the sensor holder is steel so this ruined my side cutters but it did work.

I wasn't however, able to cut the bottom of the holder so I just bent it down, out of the way.

I left a small bit of metal toward the top of the watch uncut as it wasn't interfering with the movement and in the end, it actually ended up helping a little to hold the o-ring in place.

Now the real fun begins. If I wasn't so set on being able to set the time/wind the watch with the original button and crown, I could have just added a second crown on either side and this project would have taken a weekend or two instead of taking 3 months. It took so much longer because these next steps were all very delicate, confusing to work out, and time consuming to test with such small clearances. As such, every test and assembly process was super slow and methodical in order to make progress without breaking anything.

To start with, here are some photos of the movement to help visualise the problem. On the NH38 movement used, the stem serves 2 functions: 1. when pushed in, it can be turned to wind the watch and 2. the stem can be pulled out and wound to change the time.

The issue is that the stem generally comes out of the centre of the watch movement (reasonable enough) and the apple watch crown is offset toward the top. I really wanted to avoid adding a separate stem coming out the middle of the watch body in order to maintain as close to the original Apple watch look as possible. This would have been a much simpler solution and was an option I was considering for a long portion of the project. I would definitely recommend doing something along those lines if you wanted to attempt a similar project as trying to use the original crown and button was a very difficult.

The first image shows the movement mounted on a paper mock-up dial. And you can see that the original stem threading barely even reaches the side of the dial. I knew that I would need to make some kind of linkage to connect the button to the stem to pull it in and out (to set time and wind) as well as a pulley to connect the original apple watch crown to the watch movement stem.

The plan was to use what little space I had in the unthreaded section (which is actually INSIDE the footprint of the watch movement!) to connect to the button (drawn in blue in my initial conceptual layout). And then I would use small threaded section of the stem to connect to a pulley to the Apple watch crown (drawn in red in my initial conceptual layout)

The final image shows the solution I ended on.

Throughout the making of this project I had 3 large unknowns:

1. How would I make the button linkage strong enough to pull out the stem?
2. Would the crown pulley have enough friction to actually turn the crown and wind the watch?
3. Would the complete movement fit in the watch body as the height I had inside the apple watch was actually about 0.5mm less than the minimum height recommended by Seiko (which already only gives 0.4mm clearance on either side!)?

I was unsure of the answer to any of these unknowns until very far into the project which made the process less enjoyable than it should have been!

To work out the button and crown mechanisms, I laser cut a temporary brass dial out of the same thickness material I intended to use in the end. As I wasn't entirely sure how I was going to make these mechanisms, nor how much space I would have, I simply cut away a large section of the dial for prototyping which also made it much easier to see what was going on. You can skip to the dial section to see more about how this was made. The only major difference is that I used medium thickness superglue to hold the dial feet rather than JB weld as it was quicker and to be honest it worked just about as well as JB weld.

I also cut the plastic movement space to allow me to make my modifications more easily. I always intended to get another and modify it better once I knew how everything was going to go together but it worked well enough in the end and those plastic spacers are really overpriced

The first step to connecting the button and the stem was to try connect to the stem in a way that would allow the connection to pull/push the stem in and out without impeding the stems ability to turn.

I had some of this very fine brass tube that is 1.6mm outer diameter and 0.8mm inner diameter. IIRC, the stem where I was hoping to place this tube was ~0.7mm diameter and the thicker sections of the stem were ~0.9mm. Perfect!

I stated bu cutting a slit down the length of the tube with a jewellers saw and then I cut a section of tube off with an old Xacto blade. With a little cleanup of the ends on some 600-1200 grit sand paper, everything fit perfectly and snapped right onto the stem.

Next I mocked up the part I was thinking of making to connect this brass section to the button.

I started with a CAD (Cardboard Aided Design) model cut by hand. Then drew the part up on illustrator and used my Fibre laser to cut it out of some 0.008" (0.2mm) phosphor bronze sheet. I think 0.2mm brass would have worked just as well but I had this phosphor bronze sheet on hand and it soldered well so I went ahead with it.

The part shown in this step was just a start and I needed to make many more changes yet in order to fit and work in the space I had.

After some more tinkering and thinking, I ended up on this shape, with the upright section of bronze closer to the movement rather than the case.

The whole part is soldered together as shown, with a fillet of solder in the bend for extra strength.

I also made a small bend in the upright section to better clear the movement, and when soldered up, it also provided a little extra strength.

And to clear the rotor, the top of the thin brass tube was filed down.

This part still needs to be bent some more, drilled, and modified with JB Weld later on, but this was enough to move on.

Now onto the button, I ran into a delightful coincidence that really made this idea possible: the original hardware for mounting the Apple watch button uses M0.9 threads, the same threads used on most watch stems!

Not all the apple watch hardware was M0.9 so this was very lucky and I can't say that it would be the same for all generations from series 0-4. Without this, I have no idea how I would have made something to securely hold onto the button, especially seeing as it would NEED to be assembled with screws as the button is outside and the rest of the linkage is inside the watch body.

But because of this, I could use these watch stem extenders to connect to the button, and then to the stem linkage from the last few steps could be secured to the stem extenders with M0.9 bolts

The next step was to file a notch in either side of the button to allow my finger nails to grab the button to pull it out. With it being such a small part to hole while filing, I used this brilliant jewellers trick for holding onto parts. I used thermoset plastic beads to hold the part in a way that is easier to manipulate by hand/hold in a vice, in this case by attaching it to a piece of metal.

Filing went quickly with needle files, I then polished it up and soften the edges with 600 and 1200 grit sand paper.

Next step was to work out how to connect the Apple Watch crown to the movement stem. I decided to start by trying to turn the original part of the apple watch crown inside the case into a pulley with a needle file.

I am not sure what material this part was made of. It was quite hard to start with but softened up once I got through the outer layer and this actually worked really well. I figured that if this didnt work, nothing was lost and I could always try replacing this part, or adding a sleeve over the original part made out of some brass tube.

I did all the filing with the crown assembled. The reason I didn't want to remove this part is that I did remove it on a practice watch and it was very difficult (the area to grip the nut is very small and thin) and I damaged the nut on my practice one. On the videos I linked of people tearing the watch all the way down, one does not show this step and the other actually breaks the thread off in the nut.

Now to make the brass pulley for connecting to the movement. This time I chucked a piece of 3mm K&S brass into a handheld drill and filed it in a similar way to the Apple watch crown pulley to make a pulley shape. I had to remake this part quite a few times to get it to work. I found than any smaller than 3mm and there was not enough mechanical advantage to wind the watch (lucky there was space), and it had to be quite a thin pulley to fit in the case.

I find that without a lathe, this is by far the best way to drill holes down the center of round stock. By chucking the stock up in the drill and clamping the drill bit, you can essentially create a really unstable, wrong way around lathe. The benefit of this is that because the stock is spinning, there is no way for the bit to wander off diagonally so it goes straight up the part.

With everything being so loose, it is still quite hard to get the drill centred but it works well otherwise.

The drill bit is a cheapy one from those kits that have bits from ~0.3mm-1.6mm. The bits are garbage but worked more than good enough for this with a little sharpening on a fine knife sharpening stone. I held the drill bit in a pin vice to make it more secure and you can see that I am actually getting proper chips.

Next I tapped the hole. I didn't quite feel like ordering a M0.9 tap and having it break on me straight away so instead I modified a watch stem slightly in order to form the threads in the brass. All I did was ground two sides down to form a kind of spade bit top.

In the end, Im not sure that I actually "cut" any threads. I reckon all the threads were essentially roll formed and I wouldn't have wanted to go any deeper than the ~0.5mm I did but it did work really well.

Then I could cut the formed pulley off the stock. A jewellers saw was definitely an option here but I instead decided to cut it off with an Xacto blade again because I thought this would get me closer to the final shape meaning less sanding and less chance of mistakes. It took quite a while and a lot of pressure to get through but this also worked quite well in the end.

To finish off the pulley I gave each side some light passes on 1200 grit sand paper.

Where the pulley was supposed to sit in the case needed some space cleared. I ordered a set of different shaped carbide bits, not expecting much. But even in my really off centered, cheapy rotary tool, they didnt snap, chip, or dull prematurely. And they cut really really well, making nice chips for a hand tool into stainless steel, which I was also surprised at as I think these are generally marketed for wood working. I would definitely recommend some cheap carbide burs (over, say, high speed steel or even diamond) for this application.

At some point, a little metal bracket pinged off that I later realised was just glued into the case, no hard done. I also blew through the button hole at some point which also wasn't that big of an issue but I did fill it with some thin JB weld at some point when I had extra JB weld mixed up.

For the dial I chose a material I was somewhat familiar with from making knives, but it is not one I have seen used in this capacity for making watch dials and I really REALLY cant understand why. Zirconium!

The first step was to thin the original metal stock as I could only acquire zirconium sheet in Australia at 1.4mm thick and I needed it to be 0.5-0.8mm thick.

Using a block of wood and a belt grinder I got the section I needed down to 0.8mm, but after countless hours and $30 in belts, I decided to leave it there.

Then I cut a section off and evened out the thickness by hand sanding with a janky contraption made of a block of wood and two offcuts of 0.8mm brass.

To maintain the best finish, I thoroughly cleaned the metal before heating it and I tried to mainly heat the zirconium from the back side in the hopes that any weird artefacts from the direct flame wouldn't show up on the front side of the dial.

Now this is why I think every watch should come with an option for a zirconium dial. you can see that in different lighting, the dial goes from a bright steely colour, to a really deep, but still lustrous black. If you highly polish the zirconium, the black is even darker and more lustrous but I wanted the dial in this watch to stand out a little, hence why I left it more of a satin finish (and hence a little brighter overall).

A polish is also harder to achieve and risked showing up scratches worse, however, once darkened like this, the zirconium is actually very scratch resistant, I assume because heating it does something to the effect of oxidizing the surface, creating a hard, zirconium ceramic.

Now it was time to cut the dial out of the metal stock. After a couple more test dials from brass I bit the bullet and put the zirconium under the laser. I also did a small test square to see how many passes it took to get all the way through the metal and then worked backward to work out how many passes it would take to get ~0.3mm thickness for the features that weren't cut all the way through.

For most of the time I spent messing with the dial, I covered the front surface in packing tape to protect it. The first thing to do was to remove the small tabs left from the laser cutter, then fit the sides/rounded corners to the watch case by hand. I also drilled the center hole out to exactly 1.9mm as I knew that the laser created sloped cuts and I really wanted the central hole sides to be square.

To finish off the dial for now (there will be more!) I glued on the dial feet using JB Weld. The seiko specification calls for 0.64mm diameter dial feet but I actually used one size up, 0.69mm, for a more snug fit as I wasn't yet sure how I was actually going to hold the movement to the dial (this is usually done by the back of the case).

with the final dial made, I could now attempt to connect the button to the crown! I put the watch together with some layout fluid on the bronze linkage, and marked the hole positions with a needle (which was actually way too fat for this job but I dont have a thinner scribe!). Then I held the part, again using the thermoloc plastic trick and drilled the holes out by hand with one of those cheapy, mini drill bits in a pin vice. This did take quite a while to get the initial hole but after the initial hole was made it was very quick to make it larger.

Oh man, what a step this was... until this point I was still unsure if anything about this watch would even work. But seeing it all together like this really gave me the motivation to push on.

And yes, I was considering Dauphine hands for a while but they were a little long. And these Nautilus style hands are funnier and match better, being the same style as the real Apple Watch.

One of the issues I realised Upon mocking everything up was that the o-ring on the apple watch crown was rubbing against the dial. I tried to give it just enough clearance with the laser but accidentally ended up blowing right through. This did give me the space I need and it is not such a big deal as with the black border on the glass, you cant even see the hole unless you tilt the watch really far and if you do, its filled with a black o-ring so it is very camouflage.

I also used JB Weld (yes, this watch is half JB Weld) to plug the original speaker and microphone holes in the watch case. I roughed up the are around the holes first and filled them up. I put some tape on the other side so that there would be an air bubble left and the JB weld wouldn't flow all the way to the outside of the case.

Another issue was that one end of the dial was sitting higher than the other. This meant that on the low side, the rotor was hitting the case, and the watch hands fouled the front glass. I always knew this could be an issue and ordered extra adhesive to stack behind the glass to get it further from the hands. But after accidentally blowing through the dial already when making clearance for the pulley, I didn't want to risk the dial again and the 0.3mm thickness I had on the dial where it met the case was already as thin as I wanted to go. So instead, I lasered the shelf down on the actual case which worked about as well as I could have expected. Its not precision machining but it worked.

And finally, during the mock up I ended up breaking the solder joint between the stem and the linkage. You can see that there is solder left on both the brass tube and the bronze so that tells me that the joint was about as strong as I could have made it. The only solution I could see was to make the stem easier to pull out (which I will do in the following steps), resolder the joint exactly as it was before, and pray!

So this is something I definitely didn't expect to be doing in this project but I had to modify the actual watch movement. When I put the watch together temporarily to see if everything was working as expected and it was very very hard to pull out the crown using the button. There was just too much leverage that wanted to bend the small, bronze linkage I made (despite me still being very impressed by how stiff I was able to make 0.2mm bronze). It also put a lot of stress on the solder joint to the brass tube which ended up failing as you can see in the previous step.

After much thinking on potential solutions, I decided that I would have to modify the setting lever spring on the watch's keyless works in order to make the crown easier to pull out/push in. And this entailed tearing down the watch 😨😨.

In the photos above you can see the part in question. In these seiko movements, the keyless works cover plate has this integral setting lever spring and detent. The long thin arm provides the spring force that allows for the clicks when pulling/pushing the crown.. Before taking the watch apart I had planned to either thin the spring arm, or grind down the detents. Unfortunately, you can see how small even this relatively large part of the watch is compared to my smallest needle file, the teeth on the needle file are comparable to the features I was expecting to file!!

I did try to use some 1200 grit sandpaper glued to my finest screwdriver to sand down the hump but I got frustrated with this and just ended up bending the yoke spring down. This worked an absolute treat, I'm not sure how much further I could have pushed it before it snapped however, but I was pretty frustrated by this point and figured if I did snap it, I would be fine to spend another $50 on a new movement as it would allow me to mess with this one more and I didn't really have any other things to try. I wouldn't have tried this if it wasn't a Seiko movement either as these things are built to be workhorse movements and I expected the part would not just snap.

I watched countless movement teardown videos to prepare for this step and was still blown away by how absolutely tiny the features on these parts are. It is very different seeing these parts blown up on a screen as compared to with the naked eye. I also made some mistakes, I slipped a couple times with my screwdrivers, I must have touched one of the winding gears (maybe during this step, maybe later) because it ended up with rust spots, and I didn't add oil to any of the parts I removed and put back. So the movement was undoubtably worse for wear after I messed with it but all in all, I am very glad I did it as the modification was extremely successful and the crown was much easier to pull in/push out.

I also found that the screws from the button assembly were fouling on the movement body, making it hard to push the button in still. To fix this, I had to grind some of the watch body for clearance. I actually decided to put the movement back together to grind it because the alternative was taking the entire train of gears off which I was not confident with, especially without the ability to add any new oil.

Once the watch was together, I made a wooden part to rest the watch on and protect the stem, I covered up any holes with Rodico, and then I covered the rest of the movement with tape.

The actual grinding was quite uneventful, I just used a fine finishing belt on my grinder to take the bump down as far as I could manage without hitting anything important.

Yes, so this is a bit of an odd step. This was toward the end of the project, and I simply couldn't find anywhere to order M0.9 bolts in a reasonable time, for a reasonable price. Everything down to M1 seemed to be attainable, with M1.2 and M1.6 being very attainable, but no M0.9.

The ONLY place I knew I could easily find M0.9 threads was watch stems so I ordered a box of watch stems on Amazon and made my own bolts. I think this step would be confusing to explain and is easiest just to follow in photos.

I ended up darkening the bolts with a torch which made them almost invisible as the dial holding screws in the next step.

Now I was ready to figure out how to actually hold the dial to the case. I couldn't think of a great way to do this so I decided to make my own little mounting points, again utilising the stem extenders.

To do so, I drilled holes in two corners of the dial. I had to put the drill bit into a pin vice and then into the drill press to use it. I do actually have some of those ubiquitous carbide bits (for PCB drilling and whatever else) which would have been much better for the zirconium and wouldn't have needed the extra pin vice, but I only had down to 1mm and needed 0.9. I can't believe I used to think 1mm was small before tackling this project!

Then I bolted the stem extender to the dial, put JB Weld on the stem extender/inside the case, then put the dial into the case and let everything set up, leaving the stem extender JB Welded into the case. Doing it this way I was left with a mounting point that had deep threads, was secure enough for what I needed, and I didn't need to perfectly align two <0.9mm drilled holes or tap M0.9 threads into stainless steel (I can't even begin to imagine attempting that...). I roughed up both the case and the stem extender as much as I could before the JB Weld.

It would have been nice to have the foresight to make another test brass dial, exactly as it would be in zirconium, before drilling holes in the button linkage. Because after dropping the dial a little by lasering the case, the linkage holes no longer lined up.

I also found that the o-ring was pulling the bend I made in the bronze flat, so I JB Welded a strip of metal under the original bronze with a large fillet of JB Weld to fill in the gap from where the bronze was bent up. In the end this part looks really slapped together, nothing like the precision parts usually found in a watch! But it did work so what can I say =).

The final part to make was a gasket in an effort to keep dust and water out of the case. I gave up on the idea of making it waterproof very early in the project (I think it would have been possible by filling in the bulk of the button hole and only leaving two holes for the stem extender to go through with o-rings but I REALLY didn't want to complicate this project further at this point) but I hoped that this large silicone gasket would at least help with dust and splash resistance.

Again this was toward the end of the project, and not wanting to order/wait for anything else, I found the thinnest silicone in my house, this old protective sleeve for a power bank, and cut the gasket out of that with a CO2 laser. I ended up using the laser on a low power to thin the silicone down to ~0.2-0.3mm to keep it as thin as I needed.

I actually intended to glue this gasket to the case and have only the stem extenders move through the two holes but the B7000 glue really didn't want to stick to the silicone (fair enough) and the stem extenders kept catching on the holes, pulling the gasket away from the case anyway. So I just left it floating there in the end, again, definitely not waterproof but hopefully splash resistant at the least.

Getting close now! Decided to polish the case. I like the fact that the stainless can be really scratched up and still look good but decided that if I am going to go through all this effort to try give this case a second life, I should try bring it as far back to polished as I can.

I also didn't want to cut away any of the case with heavy abrasives like sandpaper so I knew that some of these scratches and dings would be too deep to remove. Beyond those deep marks, the Cape Cod polishing cloths worked quite well through it did take a lot of elbow grease.

At one point I actually thought about using a domed glass back as I wasn't sure if 0.8mm would be too thin and just crack but I couldnt go thinner with the flat glass because it was already sticking out the back of the case. I have included a video of the domed back because I think it is cool, but definitely glad I went with the more traditional back. Its much clearer to see and the watch is already weird enough.

I also lightly sanded the edges of the glass to make it more comfortable on my wrist.

I cut the apple logo out of the rotor with my fiber laser, being careful to keep the part cool with compressed air.

Then the text was engraved, also with the laser.

It turns out the rotor was plated brass so the text already came out a little gold but I wanted it even more gold so I filled the text with some Tamiya gold enamel paint. I was really concerned I wouldn't be able to find a paint that would go on this thin and still be pigmented enough to show up but obviously model paints were always the answer.

I think this step comes across much better in the build video. There are many things to note however.

I glued the movement to the dial. I always wanted to create a better solution to this but I figured that with the way the watch is assembled, this would likely be some sort of bracket JB Welded to the dial anyway. So in an effort to get it done (perfect is the enemy of good enough or something like that), I just used B7000 after being really impressed with it's holding power on the back glass during test assemblies.

The watch hands usually actually got installed once the watch was together, I would place a laser cut block of wood in the cut out to support the movement from the centre when pushing on the hands. This way I would not risk damaging the hands while installing the dial. I just thought it looked better for the video to do it before.

Normally you would have to depress the setting lever dimple to push in the stem but I think because of the angle, it just slides right in. Every time I assembled it, it was very hard to get the stem deep enough to actually engage with the keyless works because everything is so tight and my work is so sloppy. I really needed to push on the stem from the pulley (a very tight fit!) using thin tweezers and push on the movement to get the stem to seat properly.

When putting on the back glass, I always struggled to fill enough B7000 so that there were no bubbles, without a bunch of B7000 spilling out into the case. I always rotated the glass to spread the B7000. In hindsight, I think I should have used more glue and not pressed down so hard but it really came out pretty good so I don't know why I'm complaining.

This step is for me, so that in the future, when I need to service the watch, I will at least have some idea of how!

I needed to disassemble and reassemble the watch many times throughout this process so it is definitely possible to disassemble neatly, but it would be near impossible without knowing the exact steps.

1. Remove back and front glass. To do this, isopropyl alcohol helps to soften the glue but the main thing to use is heat. Heat the watch with a hot air gun until it is barely holdable and hold it at that temperature for several minutes. Try some isopropyl alcohol and if it wants, it may wick its way into the glue bond, speeding up the process. To pull off either side, use a suction cup (currently using one for phone repairs but who know, maybe there are no phones in the future). The back glass has been stuck on with more adhesive than I have every used previously (to fill in the whole joint as the glue joint is visible) and was already normally hard to pull off so it may be really difficult now
2. Remove the hands as you would with a normal watch, being careful to protect the dial.
3. Remove dial screws
4. The dial can be lifted from the top left corner. There is likely some tap under that one corner, that was to give the new rotor a tad more clearance as the dial sagged a bit at that corner.
5. The watch lever dimple will need to be depressed to pull the movement away from the stem. use some very fine tweezers
6. Once it is free, you can pull the movement away from the stem, just clearing the side of the case.
7. The movement is held onto the dial with B7000 glue. Use heat if possible, otherwise, continually wick isopropyl alcohol onto the joint. When prying up the movement, be careful not to rip off the dial feet (their positions are visible as brass rods in the grey plastic movement holder) a watch back pry tool works well for prying here.
8. The pulley should just come off the apple watch crown.
9. Then undo the two bolts connected to the button and everything should be apart.

In the end, I am beyond thrilled with how this turned out. It looks abolutely incredible thanks to the movement, zirconium and original watch body, and I feel really good wearing something I put so much effort into.

I would have liked to experiment with more watches, doing different dials, maybe having the second hand be a loading symbol (like H. Moser & Cie. but I was thinking of using the rainbow pinwheel) but this already took so long, I don't know if another will ever happen.

I also thought it might be fun to make a wall clock out of an ipad, a desk clock out of an iphone, or the dumbest idea yet, a pocket watch out of an iphone. Food for thought ;)

And thank you to my friend Daniel for the brilliant photos. Jesus they really sold this thing but I promise it looks even better in person!